loasAtas éeelas doas êos éapéA t éeas éas oméA


New Member
Greetings, fellow scholars.

I see that the rules of the forum require the subject to be the first line of the text to be translated. I have only one line. You will object that it's not Latin. True. Here's the context.

A colleague has given me the assignment to translate (into English) the first line of folio f6v of the Voynich manuscript. The assumptions are that:
(a) the text is in a language descended from Latin and used in the 15th century in the Mediterranean area, probably in southern Italy
(b) the characters can be transliterated by means of a key which my colleague has provided
(c) the text refers to a common plant which was harvested for food in the Mediterranean at that time.

In terms of letter frequencies, the key is at variance with classical Latin, in the following respects:
* a, e, o and s are greatly over-represented
* i, t and u are greatly under-represented
* b and c are absent.

Taking the key as provided, my first attempt at a raw transliteration is (as shown in the subject line) as follows:
loasA t as éeelas doas êos éapéA t éeas éas oméA
* A (which looks like 9 or q, and generally is a terminal character) denotes "free a"
* a (which looks like a) denotes "trapped a" (an "a" within a word)
* e (which looks like c) is a short e
* é (which looks like ct) is a long e
* t is probably a full stop (UK) or period (US)
* ê is a long e with a superscript that looks like a 9.

I am considering the possibilities that:
* the é could alternatively be et, ec or ei, in which case the transliteration might read:
loasA t as eteelas doas êtos etapetA t eteas etas ometA
loasA t as eceelas doas êcos ecapecA t eceas etas omecA
loasA t as eieelas doas êios eiapeiA t eieas eias omeiA
* the terminal s could be the Latin suffix -is or -us
* the terminal A could be the Latin enclitic -que or possibly a catch-all for the Latin accusative or genetive suffix.

I would welcome any ideas from speakers of Latin or modern Romance languages (I speak French and Romanian and can read Italian, Portuguese, Spanish and a few words of Neapolitan).

VMS f6v line 1.jpg


New Member
Many thanks for the comments. The approach to a Hebrew translation looks very interesting.

In order better to match the letter frequencies in Latin, I have experimented with the following adjustments to the key:
* A (which looks like 9 or q), if an initial character, is Latin "ab"; if a terminal character, is Latin "i"
* é (which looks like ct) is Latin "cu"
* ê (which looks like ct with a superscript that looks like a 9) is Latin "cui".

This results in the following correspondences in the frequencies:
VMS Latin
i 10.4% 9.8%
e 10.9% 9.8%
a 7.6% 7.6%
u 8.9% 7.3%
t 1.3% 6.9%
s 7.1% 6.5%
r 4.6% 5.7%
n 5.9% 5.4%
so with the modified key, among the common letters only t is under-represented. That suggests that A, if a terminal character, could be Latin "it". (But now c is over-represented.)

The modified key generates the following transliteration:
loasi tas cueelas doas cuios cuapcui t cueas cuas omcui
loasit tas cueelas doas cuios cuapcuit t cueas cuas omcuit

Makes any sense?


Civis Illustris
Or maybe Aztecan. :D http://cms.herbalgram.org/herbalgram/issue100/hg100-feat-voynich.html?ts=1602669179&signature=b3929d510b4d9834164b2d4defca104f

Seriously, what's the point of the Voynich manuscript? It's just a curiosity. If I were you, I'd better decipher Harappan script. That would shed light on a mysterious ancient civilization. The language is presumably related to Dravidian, though of course there's a gap of some two thousand years between these inscriptions and Old Tamil. Just imagine that the language could be a missing link between Dravidian languages and Elamite! That's what I call fascinating.


New Member
I have experimented with modifications to the key, the better to match with the frequencies of letters in Latin, as initial letters and as terminal letters. For example, the glyph "9", which in the VMS occurs 1828 times as an initial letter and 14516 times as a terminal letter, could represent in Latin the prefixes "qua", "qui" or "quo", and also the suffixes "it", "et" or "at" (or even all the verb suffixes). With this and similar assumptions, the first four lines of f6v could read:

loasittas selas doas sos sapsit. seas sas comsit
coaet sos slit doelsat seat conaius lseit sos saus
doais lit soe coosolit selsoit lit oloe titsot
quaslit quamsoit. ot quananit nsit neit olonit quamonit.

For sure, this is not Latin as we know it. What I wonder is, could a 15th century language descended from Latin have read something like this?


Civis Illustris
Seriously, what's the point of the Voynich manuscript? It's just a curiosity. If I were you, I'd better decipher Harappan script. That would shed light on a mysterious ancient civilization. The language is presumably related to Dravidian, though of course there's a gap of some two thousand years between these inscriptions and Old Tamil. Just imagine that the language could be a missing link between Dravidian languages and Elamite! That's what I call fascinating.
Deciphering the Rongorongo script is perhaps less linguistically interesting than Harappan, but is also potentially less challenging since a lot is known about Polynesian languages now, and some of the inscriptions may or may not shed any light on what the hell happened in Rapa Nui.

That said, considering the awful work that gets published on the Voynich manuscript all the time (you may have gotten a word that it was found to be in "Vulgar Latin" last year, in a paper that the journal later on even tried to retract!), perhaps it's good that it exists to divert better efforts to the more obscure stuff...


New Member
And if we allow the Ɂ glyph to represent the Latin suffix "ius", the first four lines could be:

reaisi taius etaeraius deaius esteius etapeti. etcaius etaius οmeti
οaet etei etrei decretat etcait οnauus retci eteius etaius
deaus rei etel οeeterei etcretei rei οrel tietet
quaestrei quametei. οt quanani neti nci οreni quameni

VMS f6v 1st 4 lines.jpg
Last edited:


New Member
Alternatively, if we treat the "9" glyph as a variable prefix or suffix, or even a punctuation mark, which we provisionally denote by *, the first four lines could read:

roaius* taius etccraius doaius estoius etapet*. etcaius etaius οmet*
οcct etoius eteret* docretaius etcaius οnauus retc* etoius etaius
doaus eret* etol οoetoeret* etcreto* eret* οrol ius*etot
*esteret* *meto*. οt *nan* net* nc* οron* <> *mon*

It's also possible that the glyphs presented here as "m", "n", "r" and "t" have not been correctly transliterated.

Maybe some intuitive Latin scholar can detect some meaning therein, or alternatively identify it as gibberish.
Last edited:

Michael Zwingli

Active Member
Well, since I'm 1/4 Polish on my mom's side of the family, I feel that I have license to note that the bookdealer who paid money for this "pap" wasn't French, or Italian, or English, or... (just sayin'...hope this doesn't get me in trouble)
On the bright side, Zamenhof more than made up for Voynich's foibles!
Last edited:

Etaoin Shrdlu

This inspired me to look up Voynich, something I had never very much thought about -- 'thing' because I had never considered what it signified, and probably thought it was a place name or the like rather than an actual person.

Michał Habdank-Wojnicz is absolutely fascinating, as you might expect from someone whose occupation is given by Wiki as 'Revolutionary, Antiquarian Book Dealer'. Incidentally, his wife was the daughter of George Boole, as in logic, another interesting character.


Michael Zwingli

Active Member
Fascinating enough, to have been both a nobleman and a socialist (which while fascinating, might be construed as unwise). I was wondering what kind of bookdealer a "revolutionary book dealer" might be, then after using Etaoin's link, I realized that he was a (socialist) revolutionary AND an antiquarian book dealer. I still hope that he didn't pay a lot of money for the instant manuscript, as the entirety of its value would seem to be aesthetic in nature (the uniquely styled script and pretty illuminations on the pages).


New Member
Assuming that Voynich purchased the manuscript, there is no information in the public domain as to what he paid for it. There is a claim that sometime prior to 1611, Rudolf II of Bohemia acquired the manuscript for 600 ducats (about 2kg of gold).


New Member
The latest guidance I have received is that the first few words of folio f6v (and implicitly the whole of f6v, possibly the whole of the Voynich manuscript) are written in Old Galician (also known as Middle Galician, or Galician-Portuguese).

To test this hypothesis, I'm attempting to determine the letter frequencies in Old Galician, which in turn requires access to a sufficiently large corpus of Old Galician text. Tesouro Medieval Informatizado da Lingua Galega (TMILG) would serve the purpose, but the sign-up function is not working. I created a mini corpus from seven cantigas downloaded from universocantigas.gal, wherein the letter frequencies were significantly different from those in modern Galician.

Any speakers of Old Galician out there?


Civis Illustris
Didn't we previously have a long thread here on the Voynich manuscript? I can't find it now. From memory, a man from Bristol University claimed to have deciphered it in an afternoon. He published a paper in an academic journal suggesting it was written in some early Romance language. After many serious linguists disputed this it emerged that the author was only an academic in the sense that he was a lab assistant in a science department. The paper appeared in a respectable journal by virtue of a cash payment to the publisher, which is apparently not unusual. Bristol University had to distance itself from the matter to try and protect its reputation. Or am I making all this up?