VMs: What if Alberti wasn't so smart...?

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Hi everyone,

Here's a question I've been pondering for a while: what if Leon Battista 
Alberti wasn't so smart?

AFAIK, the "standard" rendering of pre-polyalpha crypto history goes like 
this:-
(1) crypto guys make monoalphabetic ciphers
(2) crypto guys add nomenclators (extra shapes) for frequent words
(3) crypto guys add homomorphs (alternative shapes) for vowels
(4) crypto guys add extra symbols for (vowel, consonant) pairs
(5) 1466/7: Alberti invents the code-wheel
(6) 1470: Alberti writes about his code-wheel, but no-one reads about for years
(7) 1472: Alberti dies
(8) 1500-ish: Trithemius develops progressive key (cycling) polyalphabetic 
ciphers, etc

But... perhaps it's not quite that simple. First of all, (AIUI) Alberti 
never intended the code-wheel to be "clicked round" after every character 
(in the modern sense of polyalphabetic), but instead after every three or 
four words. Given that modern cryptanalysts reckon that a simple cipher 
needs something like 30 characters to be consistently breakable, that's 
actually bang on the money. :-) This means it's not quite the precursor to 
modern polyalpha you might think.

Secondly, looking again at the Tranchedino cipher ledger, I think there's 
another possibility. From ciphers like the one on page 3r (Cum Francisco 
Isulano de Tricio) which has three homomorphs for consonants (and four 
homomorphs for a/e/i/o/u), I wonder whether some encipherers in 1450 used 
only one row of such ciphers at a time (and perhaps used the extra 
homomorph for subsequent occurrences of vowels that appear twice within a 
cipher change)?

That is, I wonder whether Alberti, rather than *inventing* a kind of 
multi-cipher crypto practice, was in fact merely *technologising* an 
existing crypto practice? This would fit the "darker" modern view of 
Alberti (as discussed by Anthony Grafton in his recent book).

If there are some encrypted letters still extant in the Milanese archives 
(which we can link up to actual codes in the cipher ledger), then this 
could be explored further (and possibly falsified, if you subscribe to a 
Popperian view of science).

Alternatively, it might be an interesting idea to re-read Alberti's 1470 
"Trattati In Cifra", AKA "Dello scrivere in cifra Roma", 27, 29, 41-42, 
46-47, according to:
  http://www.datensicherheit.nrw.de/Projekte/Paderborn/TP_IX/Literatur/literatur.html
BTW: does anyone know of a facsimile copy or transcription of this?

Also: for a good-length excerpt from Anthony Grafton's recent book on Alberti:
        http://print.google.com/print/doc?isbn=0809097524

Cheers, .....Nick Pelling..... 


______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list