> > You forgot to mention the one that I like, for obvious reasons:
> > it's an early example of a conlang! Of course, there's no more proof of
> > this than for any other theory, but I believe that it has been
> > established that the language is not Hebrew, Latin, Greek, French, or any
> > other Euro-lang. It might be the case that it is one of the above
> > languages encoded.
>
> Hmm... I wonder what the following method would make of it -
>
http://www.economist.com/science/displayStory.cfm?story_id=975770
>
> It strikes me too that the software described here might be of
> interest to conlanger anyway - a "language relation" measurer.
> Of course the core technology, the zipping algorithm, is publicy known
> so I guess the whole system could be reimplemented if the authors
> didn't make it available.
One little nit: is Maltese really a linguistic isolate? I had always assumed
that it was related to Arabic. In the sample in which it occurs, it would
have been an isolate, because there were no Semitic languages, but I don't
think that it is _actually_ an isolate.
As far as the Voynich manuscript is concerned, it has been treated to
something similar. In that there is no way of knowing which sounds the
letters represent, various attempts have been made to construct frequency
tables of letter combinations. Although another problem is that there is some
disagreement as to what constitutes a "letter" in the Voynich manuscript. If
you look at an image, or download one of the Voynich fonts, you will quickly
see that there are several "letters" that could be ligatures.
Perhaps some talented programmer among us should devise a distributed
cryptographic program (similar to SETI@Home) to crack it! :)
:Peter
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see