Re: Could VMS be a Vigenere cipher?

[Jim Gillogly <jim@xxxxxxx>]

> Tom Brodhead wrote:
> has anyone considered
> possibility that the VMS is enciphered using the Vigenere system?
...
> Has anyone approached the VMS from this standpoint?

It's certainly been considered from this standpoint, but I'm afraid the
consideration doesn't go very far.  A periodic polyalphabetic system
like the Vigenere has a number of useful identifying traits, none of
which are present in the VMs.

First, the only causal (i.e. long enough) repeats show up at
intervals that are a multiple of the period.  In the Voynich
there are many repeats at various intervals.  To pick a page
at random, on f34r using Currier notation, many "words" end in
C89, and there are much longer repeats at different intervals.
There are variations of Vigenere that are less periodic, but
none that I'm aware of that would keep the same beginnings and
endings on "words".

Second, a Vigenere with an appreciable period will even out
the individual letter frequencies to a uniform distribution.
With even a short period it will be substantially flattened.
Voynich, however, has even more individual frequency variation
than English.

I assume the high variability of letters and the large number
of repetitions are the main features that led Friedman (for
example) to hypothesize hat it's an artificial language.

-- 
	Jim Gillogly
	Sterday, 28 Afteryule S.R. 2001, 20:06
	12.19.7.16.4, 3 Kan 7 Muan, Ninth Lord of Night