Enigma vs Tractatus (was Re:Left wing blogs

Carrol Cox cbcox at ilstu.edu
Mon Oct 7 17:14:22 PDT 2002


"John K. Taber" wrote:
>
> Carrol Cox asked:
> >
> > >Was Enigma a complex polyalphabetic cipher or a scrambler?
>
> and Justin replied:
>
> > It was a scrambler.
>
> I'm sorry I was slow in replying, maybe some confusion could
> have been avoided. I also don't know how much detail is tolerable
> on a list like this.
>
> But no, Justin, the Enigma is a complex polyalphabetic substitution
> cipher. This is easy to demonstrate if anybody insists.
>
> I'm not sure, but I suspect that Carrol means "transposition cipher"
> when he says "scrambler." The Enigma does not transpose letters.

Yup. Even my vocabulary is pretty rusty on this. I think the polyalphabetic ciphers we practiced on had about 10 (pure guess -- I don't really remember) alphabets, with a fixed rotation. So they were pretty simple after you got the hang of it. But I don't think I could do one now -- I don't even remember what the first steps were. Did Enigma have some way of varying the rotation?


> I'm not sure, but I seem to remember
> that with the strengthened Enigma, the captured machine was
> important in providing the order of letters on the rotors. There
> was no reason to assume that the order was the standard
> alphabet, but so it was, which greatly aided solution.

That is a bit surprising (that they would use standard order). And it would certainly have simplified the work to know that. This is awfully vague in my memory now, but it seems like the systems we practiced on had the alphabets mixed with key words.

For those interested, here's how you mix an alphabet with a keyword -- say xylophone:

xylophneabcdfghijkmpqrstuvwz -- then write a standard alphabet below it and cipher away.

Then for a _really simple_ polyalphabetic substitution, create a _second_ alphabet, with say zuzupitts as the key word:

zupitsabcdefghj etc

Then switch back & forth every (say) 7 characters.

It would seem to me that for a division level cipher, a complex polyalphabetic cipher would still do the trick. Imagine the message saying: Began artillery barrage at 1330 hours. And it was sent at 1220 hours! Not much time to fool around with it.

Carrol



More information about the lbo-talk mailing list