Decoding the Somerton Man Cypher

Tin foil hat time.  Once upon a time I was really interested in SIGINT, HUMINT, Encryption and so forth and went a little nutty looking at this case about a mystery death of an unknown man in South Australia who was found to be in possession of a code.  While reading into it and discovering cool new side stories never mentioned (like the finding of a syringe near the body or a mysterious German fellow with a doctors bag full of loaded syringes who may be an assassin) I discovered the horrible mistake that lead those trying to crack the code into failure.

Basically the newspaper printed a picture of what was roughly the code but contained transcription errors.

 

How did I find this?  I wasn’t going to try to decode it because smarter people than I with resources beyond my own have been trying the methods I would start with for much longer than I would be spending on this.  And because this exists.  If I had encoded something I thought was super important, I would use a strict adherence to this method and it would be mathematically impossible to break.  With that as a possibility I thought it would be more interesting to find out what we knew about everything else and hope for something that would let me I.D. the man but what I found was pretty damn shocking.

I’ll keep it as blunt and simple as possible.  Anyone on the internet who claims to have taken pictures of the code with infrared or filters or whatever is wrong.  Our favourite picture of the code is actually a photo of a photo the code on its actual page taken with a UV light illuminating the page.  I have seen it in person.  The original code was so hard to see that the UV light (referred to as a “special light” and not in common use at the time) was required to display it adequately.  Unfortunately the camera used to photograph the code by the Adelaide Advertiser didn’t pick it up very well at all and this problem was only discovered after the film was developed.  The result was faint and unprintable.  So what could they do?  Grab a pen and trace the outline of the code onto the photo.  This becomes evident when we zoom in on a few of the letters, courtesy of a high-def scan of the photo, and the start/stop pooling of the ink betrays the truth.  Remember all reports said the code was in pencil.

Ink pooling as pen started and stopped
Dotting like someone was “feeling” out the letter
Didn’t even line the letter up
Is that even the same letter?

Looking over the HD version and playing with some colour settings makes it apparent that some of the letters have even been traced incorrectly.  Could this really be true!?  Can we prove that this is not the correct code some other way?  Well, there is this document retrieved from the national archives from an editing of a TV news story in which the officer investigating actually described the codes appearance…

Near the Bottom next to where someone wrote OLA CODE

What!? Uniform letters and square layout?  That’s not what we were given!  Some encryption methods actually rely on the code having specific layouts and obviously with something this short, one or two incorrect characters could also easily thwart the super-est of super computers attempting to crack it.  So this could be a lost cause already.  Well it turns out the whole evidence bag has since gone missing so we can’t get another copy of the code.  But I did find the articles submitted to the CIB head office at the time, by the officers investigating…

A little weathered by legibly different.

Now I’d be more inclined to believe an article written back to base by an officer requesting assistance as more reliable than the local newspaper.  It’s looking good for my theory….But just to be sure how about a later snapshot from documents submitted to the senate describing the code differently…

So the newspaper still happened to hold the photo in their media library and guess what we find when we go to look.  Picture of a picture.  I would have liked to get the photographers name and continue down the rabbit hole but this is about where all the “who are you and is this authorised” started at me.  I then got a little distracted by life and this is where I left things.

In short, we’ve been trying to crack the wrong code for 60 something years.

I’d love to upload my raw resource files but it’s 700mb or so and I can’t be hosting it here.  Sorry about that.

UPDATE!  Trimmed back some of the files but here’s a large sample of what I’ve put together from my scattered backups.  Uploaded to Google Drive as a zip.

Anyway, have fun.  I’m done with this one.

5 comments on “Decoding the Somerton Man CypherAdd yours →

  1. Very intriguing! I’ve been interested in this case but don’t have the smarts to delve, this is awesome information you have found. Good work.

    1. Thanks! Great question, sorry I didn’t clarify that acronym. It’s the Criminal Investigation Branch if I recall correctly. The general feel of things from what I’ve gleaned was that this group was very similar to Major or Organised Crime squads the police have now days but just a little more senior.

      The documents were originally sent back to head office (I’m having trouble locating the page with the address for it but I recall head office being in the UK possibly) at the time, but have since made it to the Australian National Archives.

Leave a Reply

Your email address will not be published. Required fields are marked *