Nick Pelling's thoughts on the Zodiac ciphers

[Zsearcher Profile]

So this proves that the final 18 was intentionally placed.

Why it was intentionally placed is another story.

Obviously it was used as filler, but what else?

What I mean is, if it was just filler, then why use 12 symbols that line up with other rows and the 4 symbols in a row?

Why not just throw in random symbols?

To me it seems like there is something else going on here.

[Wrench Profile]

Well. proving the 18 were pulled from rows above and proving it is only filler are two separate issues, I think. In other words, it's still possible that Z switched keys, anagrammed, shifted etc. AND used the symbols from above, the 4gram perhaps to make it appear the same plaintext appeared in those two rows, when there is actually another solution. Whether he went through all that trouble in another story..

[smithy Profile]

...and whether that's a likely scenario is another consideration. To my mind it ain't. 2¢.

[Wrench Profile]

Me too, at least based on the rest of the 408. Things did seem to get a little weird after that, the symmetry of the symbols in the MNI, the pivots in the 340, but without those solutions any clue they might provide remains a mystery and vice versa.

[doranchak Profile Website]

As a general rule, I look for small stepping stones and try to build them up.

...

(4) A pure homophonic cipher with only a modest number of symbols should surely have lots of fragmentary left-right patterns, something Z340 doesn't present.

I was revisiting this thread and came across this point. Z340 does contain a variety of fragmentary left-right patterns, but Z408 has more of them. Can this sufficiently be explained by the greater variety of symbols in the 340?



The pink highlights show what I consider to be the most significant repeated patterns: |5F, FBc, 5?4?., and O?*?C (Each "?" can represent any symbol). The remaining highlights show all repeated fragments of the form A?B. I've ignored many others of the form A??B, A???B, etc.

Z408 certainly has plenty more repeating patterns. The significant ones seem to be pretty well distributed through the cipher. But the most significant patterns in the Z340 (shown above highlighted in pink) tend to occur towards the middle columns of the cipher text. The pivot patterns are also near the middle columns. Perhaps this suggests some aspect of the cipher construction.

Additionally, lines 10 and 11, which you note might be the start of a new key, or have some other significance, do not contain any significant repeating fragments except for two minor repeating 2-symbol fragments. I was looking at the repeating symbol cycles of the 340 and some of the significant ones tend to break around those lines, and then resume afterwards:

http://www.evernote.com/shard/s1/sh/7b26b6fa-fa23-4af2-a758-497a594f0466/ac8b7fe4139d0268a1888f891c30f43e

This isn't true for all of the homophone sequences, but it made me curious. Perhaps some simple re-arrangement would restore a series of unbroken cycles, similar to the well-formed unbroken cycles of Z408 which tend to break down towards the end rather than at the middle.

[smithy Profile]

Like it. Very much.
I also like JSpp7 and the "marquee" text J+Op7 at line 10. I wish they were pink.

[traveller1st Profile]

Perhaps some simple re-arrangement would restore a series of unbroken cycles, similar to the well-formed unbroken cycles of Z408 which tend to break down towards the end rather than at the middle.

Are we talking the old slice-n-dice-n-switcheroo?

[doranchak Profile Website]

Like it. Very much.
I also like JSpp7 and the "marquee" text J+Op7 at line 10. I wish they were pink.

Thanks for pointing that out. Looks like I need to fix my crappy algorithm.

[smithy Profile]

Are we talking the old slice-n-dice-n-switcheroo?

I think at least that exists as a possibility Trav. Dunno though.

Sorry D.! I thought you'd excluded them because you didn't think them significant. (Or I'd have pm'd you of course. Eek).
BTW I'm very pleased you highlight nothing of interest in the last 15 characters, of course.

[doranchak Profile Website]

BTW I'm very pleased you highlight nothing of interest in the last 15 characters, of course.

There's always Np+B(#O and NpkSzZO (Np????O). Significant? I dunno.