Re: Cryptography of 1502 with Voynich resonances

[Zandbergen@xxxxxxxxxxx (Rene Zandbergen)]

Thanks to Adam McLean for this interesting letter.

The script is reminiscent of the ones presented by D'Imperio
in Fig. 39 (those who haven't got D'Imperio can see it at Takeshi's
online translation of it), and also of the various alphabets given
by Tranchedino, if memory serves me.
I just went back to my sparse notes and see that T. was born in
1441 and worked in Bologna in some diplomatic function in 1492-1493.
This is only 10 years before the letter given by Adam and some 25
years after the average dating of the VMs. These times are close enough
to make the comparison interesting.

One thing the letter showed me is that I was wrong in thinking that 
a difference between the VMs and such ciphers was that the VMs shows
a flowing script while these cipher alphabets would not allow this.
(I don't remember seeing sample letters in T.).

I am not sure whether it is a betrayal of humanist-writing tendencies,
but the author of the letter clearly had a desire to produce an elegant
looking result. All the lines are straight as a ruler and the size 
and shape of the letters do not vary.
 
Quoting Dennis' translation:

> Thus we shall encounter one of the famed Spanish
> ciphers that Matteo Argenti held in great esteem.

In the introduction to Tranchedino there is a reference,
which I scribbled down as follows:

 Calendars of Letters, Dispatches and Statepapers relating
 to the Negociations between England and Spain, preserved in
 the archives of Simancas and Elsewhere, ed. G.A. Bergenroth,
 Vol.1 1485-1509, London 1862.

This seems relevant. 

> [...] The trigrams belong to
> the nomenclator and the letters of the alphabet; there
> are fifty different signs, which corresponds to an
> average of two or three different signs for each letter
> of the plaintext. [...]
> For typographic reasons, and to keep this discussion
> as brief as possible, we shall not give the entire
> cipher alphabet; the nomenclator is of greater
> interest, and it is worth the effort to decipher it to
> the extent that the number of groups permits.

A pity. It would have been interesting to find out how many
total cipher 'words', made up of 1,2 or 3 of these 50 characters
exist. It would also be interesting so see for how many different
digraphs in the plaintext a 3-char. cipher group is given.

> [...] one also sees that
> the trigrams beginning with the same letter are in
> order, more or less regularly, according to the third
> letter.  

This is a faithful translation of the French. I fail to
understand precisely what is meant.

> Finally, as the majority of the nomenclator's
> more frequent groups must appear in a dispatch of this
> size, on can attribute a maximum of two hundred words
> to the original nomenclator.  That is to say that our
> list allows us to decipher any other letter enciphered
> by the same process.

Same comment. 
Does this mean that the encryption by pairs of letters
is what constitutes the nomenclator? If so, then I have
always misunderstood its meaning (thinking that it meant
the use of short code groups for entire words - important
ones at that).

Explanatory responses most welcomed,
        Rene