This will probably destroy what little credibility I’ve managed to cobble together over the past week. That said, I’m having a really hard time piecing this together in my head and someone with more knowledge might be able to help me think of a better way of approaching this problem. I hope you can forgive me if there is a simple explanation for this. Here it goes.
I’ve been trying to recreate the 408 key. One thing I tried the other day was to map inputs and outputs on a keyboard. This was meant as a follow up to the observation that there was some relationship there that had something to do with a keyboard.
What we would ordinarily do is have the keyboard represent the plaintext inputs and then put their symbols under them and, yes, there are some interesting things that pop up.
But you can do that in reverse too because a big chunk of the symbols in this cipher are just the basic Latin alphabet and that’s exactly what is on your keyboard. You can’t do all of the symbols because not all of the symbols are like that, but you can do a lot of them.
Here’s how it works. Say I am using a typewriter and I want to excrypt the plaintext letter F into my cipher. Well, since this is a homophonic substitution cipher, I’ve got more than one way to do that. I can either press "J" or I can press "Q." How about W? Well, I’ve only got one way to do that and it’s by pressing the "A" Key. If I want to input "E", I’ve got tons of options, I can press "W" or "Z" or "N" or even just "E."
Some of the cipher characters are also Latin letters but backwards and, since we’ve come this far, maybe we just imagine a second keyboard that does the same thing except the letters that we output are horizontally reflected. That’s basically what I was doing when I made this:
Now this was breaking my brain because, if you are translating or transforming certain letters, it it basically impossible to tell whether you are typing the input on your imaginary reflected typewriter or the regular one. This presented a problem because the next thing I did is I put both keyboards together to see what popped up.
And you can see there are some interesting clusters if you look at the keyboard inputs and outputs as a whole.(ASD/ASD, NLE/NLE. etc.)
But it got me thinking about the whole thing as a system of inputs and outputs, which I suppose is pretty obvious because that’s exactly what a substitution key is. And so I started thinking about what this whole thing might look like if the keyboard layout was actually part of the key.
Now I’m very tired, and I’ve been a little obsessed. I try to keep a level head and not let my imagination get away from me. With these ciphers there seems to be so much pareidolia when it comes to the symbols and it never means anything. But if you look at this as just an input/output machine, and you buy the idea that this started out on a typewriter then this is the machine that built the 408 cipher, at least the part of it that is Latin text, this graphic represents what the author of that cipher was typing when he created it.
I mean, I can see alot of things that look like words or parts of words in there, and it could be just faces in the clouds, but it bothers me when I know that "W" outputs to [A] and "E" outputs to [N] then I look at that keyboard and I see six letters sitting next to each other that, when combined, look like this:
This will probably destroy what little credibility I’ve managed to cobble together over the past week. That said, I’m having a really hard time piecing this together in my head and someone with more knowledge might be able to help me think of a better way of approaching this problem. I hope you can forgive me if there is a simple explanation for this. Here it goes.
Your credibility is fine with me. I think you don’t shy the experiment at all.
Now I’m very tired, and I’ve been a little obsessed. I try to keep a level head and not let my imagination get away from me. With these ciphers there seems to be so much pareidolia when it comes to the symbols and it never means anything. But if you look at this as just an input/output machine, and you buy the idea that this started out on a typewriter then this is the machine that built the 408 cipher, at least the part of it that is Latin text, this graphic represents what the author of that cipher was typing when he created it.
It is what typically happens to one when they first start working on the Zodiac ciphers. In some time you will learn to manage it and you will be able to use that experience for other things in life.
I’m beginning to think this key might have been a mobius strip
More weird, stupid nonsense in the 408 key.
I’ve been playing around looking for patterns in the substitutions by building tables. As long as the substitution order for each plaintext letter is preserved, the table for each letter follows only a single direction, and the relative orientation for each symbol is consistent with the other homophones in its sequence, I can arrange them however I want.
Now this is dumb, but one of the games I played was to line up all the round symbols I could [Q, O weird O’s} and fill in the rest of their related homophones and the plaintext letter to which they are assigned. All I did was flip the direction of the call order for "O" and "E" and pick an order for the O’s.
Looks like this:
And here I’ve highlighted some of the things I find especially weird and dumb:
EDIT: I should explain. There is no evidence that Z used a matrix in making his symbol selections. Nevertheless, I was curious to see, if there was some kind of matrix, whether I could deduce the orientation of symbols with unilateral or bilateral symmetry [O and X would be bilateral, T and M would be unilateral for example].
I’m trying to be more careful about inferring more from funny patterns than they are really offering. Explanations for the above might include
1) I’m an idiot and it doesn’t mean anything,
2) He borrowed first position substitutions from adjacent letters in his substitution key and really liked putting circles next to one another (see 340 for evidence, just kidding),
3) This really is a straightforward substitution key, but it maybe isn’t organized by frequency. He might have also filled it out in such a way as to make Latin letters and their variants cluster together,
4) The key was actually a substitution tree where lower frequency letters branched off of higher frequency letters,
5) The key was a substitution matrix where symbol assignment was made on the basis of the relative orientation of Letters and symbols,
6) There is a conditional matrix that constructs the weirdo symbols out of the Latin letters.
7) I’m an idiot and it doesn’t mean anything.
My money is on 1 and 7, but it might still be neat to try and prototype a system that can explain some of the bizarro things in 408.
Damn Teg>Ean, if you were a scientist you would probably be trying to figure out what happened before the big bang. 
and really liked putting circles next to one another (see 340 for evidence),
I would like to test that.
Do you think that any of the options that you listed can produce (or coexist with) the Qwerty keyboard observation?
and really liked putting circles next to one another (see 340 for evidence),
I would like to test that.
Ran a quick test and used all symbols that have a circle including the backwards Q, Zodiac crosshair, etc… (11 symbols total). Going through the cipher from left-to-right, top-to-bottom. For every adjacent circle symbol pair the score is incremented by 1. Did not keep track of adjacent pairs that had already been scored on the thought that there is a 2:1 relationship and the total score can just be divided by 2 at the end, for example position 16 and 34 = +1 and 34 and 16 = +1 (not sure).
Like that the test scored 44 (unique pairs?) adjacencies for the Z340. Out of 10,000 shuffles, 40% scored 44 or higher. At least that seems pretty normal.
I would like to test that.
Do you think that any of the options that you listed can produce (or coexist with) the Qwerty keyboard observation?
Possibly. I think if we were talking about a straight, simple, classic substitution table and he just slapped QWERTY in there we should see a higher deviation from random shuffles than 1 in 10,000, regardless of whether he organized they key alphabetically, by frequency, with a plaintext string, etc. A substitution tree might explain why WAS-FG, CVB would be interwoven the way they are, but I haven’t been able to draw up one of these trees in such a way that everybody ends up where they’re supposed to be. The best I’ve drawn is a jumbled mess and is just a little too convoluted to make sense as a key.
Of course, if he just made a table or two, it’s not impossible that he just slapped the thing together on a typewriter while miming Flight of the Bumblebee but it’s hard to demonstrate one way or another.
The inverted Latin letters do seem like a weird choice, but could suggest some interesting ideas when coupled with the Qwerty thing. Looking at the key as a "two-sided" or "folded" paper problem, it’s possible to put down [K] and [/] next to each other and fold the paper along an axis that it brings them closer to [R] which has [] as its third substitution. That will also put
close to [E] and all of these letters are "plugged in" to each other except in reverse (plus that brings [L] closer to [A] and [K] closer to . Bear in mind, if you turn K 180 degrees, you get backwards K. If you do the same to S, you get regular S. This would be an example of dumb idea number 5, an orientation-based matrix without operators.
It’s basically impossible to evaluate conclusively even if you can figure out what axis/s might get everybody in the correct relative orientation. There are probably dozens of possible methods with varying numbers of operations and constraints that could get all the substitutions you would need to make 408. Plus, I don’t think we’ll ever know for sure whether we’re looking at a backwards
or a [9], or whether [E] is actually [F]+[L]. In other words, this "conditional matrix" idea is a pretty useless hypothesis on its own, and nobody needs it because we already know what the 408 actually says. I just find irritating that I don’t know what 408 says about the person who made it except that he seems to be a huge jerk and nerd.
Damn Teg>Ean, if you were a scientist you would probably be trying to figure out what happened before the big bang.
Last I heard, the leading theory was brane collisions and spaghetti monsters but, like you said, I’m no scientist (thankfully).
Ran a quick test and used all symbols that have a circle including the backwards Q, Zodiac crosshair, etc… (11 symbols total). Going through the cipher from left-to-right, top-to-bottom. For every adjacent circle symbol pair the score is incremented by 1. Did not keep track of adjacent pairs that had already been scored on the thought that there is a 2:1 relationship and the total score can just be divided by 2 at the end, for example position 16 and 34 = +1 and 34 and 16 = +1 (not sure).
Like that the test scored 44 (unique pairs?) adjacencies for the Z340. Out of 10,000 shuffles, 40% scored 44 or higher. At least that seems pretty normal.
I’m really sorry, Jarlve, I was kidding. I’m too scared of 340 to make any serious claims about it. I can’t even draw any decent conclusions about the cipher that’s already been solved. Doesn’t exactly make me sound like a person with much talent in cryptography. 
I’m really sorry, Jarlve, I was kidding. I’m too scared of 340 to make any serious claims about it. I can’t even draw any decent conclusions about the cipher that’s already been solved. Doesn’t exactly make me sound like a person with much talent in cryptography.
I was curious myself. One of the first things I noticed about the Z340 was the cluster of circle symbols below the middle of the cipher slightly to the left.
I think I might have figured out what happened here with all the weird round nonsense, but also not.
In 408, if you get rid of all the plaintext except for the first time each letter appears, "T", "M" and "H" get their first substitutions in sequence with one another. It’s far from painting a clear picture, but it looks like he might have thrown everything round in roughly the same place, but that’s if he built the key from the plaintext, at least in part.
Still doesn’t explain D2 and N1, but it’s interesting (if you’re into that kind of thing)
"T", "M" and "H" get their first substitutions in sequence with one another.
What do you mean? I don’t understand.
What do you mean? I don’t understand.
I was trying to see what order each letter was substituted initially, but also broken down by cycle position (or call order if you prefer). That gave me four lists which could be compared for patterns (inside the cycle positions) and repeating patterns in the order of substitutions between the cycle positions. T1, M1 and H1 have no other first position substitutions between them (that haven’t already been substituted before) in the section I highlighted up there.
T1, M1 and H1 have no other first position substitutions between them (that haven’t already been substituted before) in the section I highlighted up there.
Gotcha! Thanks.
