Last edited on 2004-09-21 23:16:24 by stolfi Original version 1998-12-27 19:29:41 by stolfi
The most vexing part of the VMS mystery, in my opinion, is that we still cannot say what sort of person the author was. German, Italian, Jew, Arab, Turk, Lapp, or Maori? Self-taught or Oxfordite? Stupid, mediocre, or brilliant? Charlatan, crackpot, student, scholar, or scientist? Christian or pagan? Innocent, prude, lecherous, sex maniac? Conventional, eccentric, lunatic? Man or woman? Child, adult, or old man?
I say vexing because we have 240+ pages of handwriting and drawings. Even without understanding the text we should be able to say a lot about the fellow's background and personality. Anyone can invent an uncrackable cipher, but it is not easy to disguise one's style (and one normally does not feel the need to do that). The big question is whether style can tell us anything about personality.
(When I was a kid my parents subscribed to an Italian weekly that had a Graphology colum, next to the horoscope and Miss Manners sections. You sent them an handwritten letter and the resident Graphologist would draw your personality profile from the way you crossed the t's and dotted the i's. I wonder whether that idea has any basis on facts... If so perhaps we can find a retired Graphologist somewhere?)
Anyway, here are some things we can tell with some confidence:
My current guess is that the VMs is an attempt to write a basic astrological/medical almanac, either in an invented language, or in some natural language. I don't believe it is just a random string of symbols: Its word patterns and statistics are too much "natural" for that. Also, it seems evident to me that the encoding was carefully designed, painstakingly used over a relatively long period, and deliberately evolved in response to that experience.
Also, if the VMs is in some sort of code, I don't think the encoding can be very complicated. Why would the author want to encode all of the 250 pages? Besides, in my statistics, the VMs "words" seem to be genuine words; any code more complex than a simple monoalphabetic substitution would probaly destroy all the naturalness of the text.
What language is it in? Latin would be a natural guess for an European medieval manuscript; but alas it does not seem compatible with the text statistics. Thus the language could be anything from Eskimo to Ethiopian. I do not think that we can limit ourselves to "major" languages (Italian, French, German, ...). The fact that author choose to invent a new alphabet makes the "minor" languages more likely.
In fact I have been more than a bit obsessed with the idea (first advanced by Jacques Guy, perhaps not very seriously) that Voynichese may be some East Asian language, such as Chinese, Vietnamese, Tibetan. There is along list of hints that point in that direction, from the length distribution and internal structure of words, to the number of items in certain astronomical pages. I have written more about this theory in a separate page (which is begging to be updated).
Sergio Toresella believes the VMs was written for the purpose of impressing the clients of some charlatan. I can't disprove his theory, but I don't feel it is right. The visual and statistical properties of the text seem ill-suited to that goal.
Incidentally, Sergio Toresella believes that the Voynich Manuscript was produced in Northern Italy, possibly in the region of Venice. Could this be the Manuscript's language?
I do not think we can make much progress towards decipherment until we determine the manuscript's language. Moreover, the ability to read the language (in its XV century form) is a prerequisite for success.
In any case, I do not expect the contents of the VMs, once deciphered, to be particularly interesting. In fact, I don't think the VMs would have attracted so much attention if Voynich had not mistakenly attributed it to Roger Bacon.
December 27, 1998
I wrote the above comments in 1998-12-27, about two years after getting introduced to the VMS. Six years later, having gone through hundreds of hours of analysis, many megabytes of mail, and many megapixels of new images, I found to my surprise that this page did not need any significant updating.
The origin and personality of the author, as well as his/her intent in writing the VMS, remain as mysterious as ever. The evidence that the text is maningful has continued to grow, while the theory that it is random gibberish has become even less likely (in spite of much-publicised recent claims to the contrary). The arguments against polyalphabetic cipher have been strengthened, although a word-level (codebook) cipher is still a valid alternative.
More evidence has been noted in favor of Jacques Guy's "Chinese" theory (actually, East Asian), including the red glyphs on page f1r and the frequency of repeated words. That theory is still my favorite, although the competing Arabic theory may not be as dead as I thought.
The discovery of Baresch's letter by Rene confirmed him as a owner, but did not shed any light on the previous history of the manuscript. There is still no compelling evidence that the VMS was ever in the hands of Jacobus or Rudolf, and no reason at all to connect it to Kelley or Dee. And Northern Italy is still a possible source...
September 21, 2004