VMs: Antoine Casanova's thesis

[Jorge Stolfi <stolfi@xxxxxxxxxxxxx>]

Hi,

Yesterday I found the full text of Antoine Casanova's doctoral
dissertation (defended march 1999) on the Voynich Manuscript,

  http://perso.club-internet.fr/voynich/memoires/html/these.pdf
  
I had previously seen only a small section (or summary) of it, where
he computes the number of different character choices at the k-th
position of an n-character word, for all n and k; and I mistakenly
assumed that his thesis was centered on that analysis.

In fact, that analysis is only one small section of the work ---
which, at 446 pages, may well be the longest and most extensive
document yet written on the VMS. It covers many of the historical
facts and statistical analyses that have been discussed on the list,
plus several ones which I haven't seen mentioned elsewhere. For
instance, he gives a description of the artifical languages of Wilkes
and Dalgarno, as well as Kircher's Polygraphia --- and he speculates
that Kircher may have been inspired by the VMS. More on that, perhaps,
later.  I particularly liked his "motif diagrams" (p. 274-275),
even though I haven't tried to understand them yet.

I am embarassed for not having paid due attention to his thesis before.
In my excuse, I remember running into some technical problem when
I tried to download and/or view a previous version of the file.

I cannot fail to admire Antoine's work, even though, in my view, many
of his statements are questionable, and his conclusions are completely
wrong. I suppose that there is no harm if I mention some of them here,
now that he has got a firm grip on his diploma. 8-)

For example, Antoine seems to uncritically accept some rather
questionable hypotheses, such as Raphael's story about Rudolf, and
John Dee as a previous owner. In fact, he even mentions as evidence
the recollections of Arthur Dee about his father's "hieroglyphic" book
--- which is now believed to be another manuscript.

My biggest problem with the thesis, however, is that it doesn't even
mention the "right" solution. Even though Antoine made an Herculean
effort to list the key theories and facts (19 "Hypotheses", 71
"Observations", 18 "Conclusions", scattered through the thesis) and
carefully picked his way through the resulting maze, he apparently
missed the "Chinese Theory" exit right at the beginning --- and, as a
consequence, ended up lost somewhere in the "Multiple Substitutions
with Nulls" maze.

There are also a few remaining problems with the file: some
illustrations are missing (possibly due to copyright restrictions);
and some of the fonts it uses -- such as Cyrillic and Voynichese
(EVA?) --- are not included in the PDF file itself (which fortunately
affects only a few lines of the text).

Anyway, I recommend his dissertation to everybody.

All the best,

--stolfi