Currier A and B

[-JKP-]

Or, the cipher relies on some mechanism (e.g. a cipher disk) that was reset between the writing of A folios and B folios.

That's, in particular, what I mean when I speak of the different initial conditions.

I'm not saying this isn't possible, but usually when cipher disks are reset, the result is significantly different rather than different in a few small details.

I'm also influenced by Strong's idea of multiple alphabets being used in some defined progression: if the scribe accidentally misses one alphabet out of the sequence then the result might be similar to what we see.

Is there a thread about Strong's proposal? (I know you have written about it in your blog, JKP.)

My problem with Strong's "solution" is that he said multiple times that it was Trithemius but abandoned Trithemian analysis (after creating the frequency tables to determine likely languages) and used a "go with the gut" system instead, cherry-picking anything that looked like it might have meaning when combined with other words in that section.

So, he didn't actually follow through with his Trithemius assertions and didn't even analyze enough languages to go to the next step as he did.

Sloppy work. Too eager to find a solution. Lazy scholarship.


I haven't read Friedman's opinions on the VMS, or seen any of the task group's results, but I have no doubt they explored Trithemius pretty thoroughly—WWII code breakers were familiar with how Trithemian systems work.

I don't know if you can read German or Latin, but there are quite a few books on Trithemius (and other cipher-related topics) from earlier centuries on Google books. If not, there are some English translations, as well.