The question I had was, “Is the Morse Code a keyed-columnar transposition?”. 


At least not with the keywords I tried and not as a simple columnar transposition.  It’s entirely possible that columnar transposition plays a role in finding hidden meaning in the Morse Code phrases.

I liked transposition because if you take “T is your position” and insert -ran while removing -iyour, you get “T ran sposition” which makes me think it’s possible that a plaintext message hidden in the Morse Code describes something to do with Transposition.  Also, the fact that you can grid the Morse Code into a square 9×9 table plus the fact that Morse Code has 9 letters which made it an excellent keyword to attempt a keyed-columnar transposition with.  Combine this with the fact that if the e’s were added to the Morse code to make each phrase a palindrome then the idea of a palindrome must be important in some manner such as being able to read a message front and back or in two different ways which implies a transposition cipher to me.

How do you solve a columnar transposition cipher?  Well, you can try online methods but I feel more comfortable with the pen and paper on this one so that is the method I will be describing.  Do I really need to describe online solutions?  Google “keyed columnar transposition solver” or any similar search string and you’ll likely find something you like.  Pop in the encrypted text and it farts out an answer.  Be wary though, sometimes they assume you’re still going to play with their answer so good luck if that’s the route (haha) you choose to take.

I don’t know how they describe keyed-columnar transposition solutions but I like my method and it’s probably pretty similar.  Why a method and not just try and put them back in the right order by hand?  If you consider the example of trying this on the Morse Code, we are using a 9×9 grid which effectively has 6,561 ways that the 81 letters can be fit in.  Even if you just take 9 columns, it gets worse: you have 9x8x7x6x5x4x3x2x1 number of orientations or 362,880.  This assumes you have gridded the ciphertext correctly and are not prone to mistakes.  Running these on a computer with a dictionary checker would be wonderful but that’s outside of my abilities. 

Kryptosfan Method of Solving Keyed-Columnar Transposition Ciphers

Take your plaintext and get it into a cube (or rectangle).  With the Morse code, there are 81 letters so it’s a pretty good bet that it’s going to be 9×9.  The placement of the Morse Code sheets is towards the beginning of the Kryptos installation and by association is at the beginning or “easier” part of the whole.  I used what I believe is the order of the Morse code phrases if they are read from left to right DIGETALINTERPRETATIU

Go ahead and label each column alphabetically, in lowercase.  We have 9 so our tags are abcdefghi.

The number of columns determines the lenght of your keyword.  Ours in this example is 9.

 Come up with some keywords of the same length.  Sometimes the keys makes sense, otherwise…not so much.  Sanborn used palimpsest and abscissa which can kind of be applied to the work he did.  I decided to make a short list of possible choices.  First off was MORSE CODE then OPERATIVE, OPERATION, ESPIONAGE, SPYMASTER, ALPHABETS, CIPHERING, DECIPHERS, ENCIPHERS, DECRYPTED, ENCRYPTED, and CRYPTONYM.

Take your keyword of choice and number the letters from left to right, repeats are numbered in the order they appear.  For more advanced ciphers this may change but I assumed this condition for the Kryptos Morse Code.  I’ll use Morse Code from now on for the methods section.  The results from the other keywords are listed below.


The way keyed-columnar transpositions work is that you grid your message and then label each column under your keyword.  You then put your keyword in alphabetically order while maintaining the columns.  This effectively jumbles the columns in a retrievable fashion.

So for solving them, you are working with the alphabetized columns and must reorder them into the correct initial keyword orientation.  There’s got to be some easy way to do this so you don’t lose track, right?

Let’s go back to MORSECODE.  Alphabetize it while keeping the number labels.


This is the order of your ciphertext.  Now label the jumbled keyword letters with a lowercase label.


This is your translator, your own little Rosetta stone if you will.  You can now reorder your ciphertext columns into the order they are supposed to be in if your keyword is correct.  Simply put the lowercase tags in numerical order based on the numbers above the letters.  This will reverse the transposition and put the columns back into the original order.


Remember that grid you made at the beginning.  Rearrange it by the lowercase alphabetical tags into the orientation you just created.  At this point, if you’ve got the right grid, the right keyword length, the right keyword, and you haven’t screwed up – then you should be able to read the plaintext that was originally enciphered.  And you did it all with a pencil and some paper.

Not too bad huh?

So you already know it didn’t work for me.  The method is sound and I am confidant it would work for any simple keyed-columnar transposition if you had to do it by hand.

Let’s see what I did get for results…


I only read off the first row as a test to see if it worked.  I wasn’t 100% sure on how ciphertext gets padded into a grid for solutions so I did it both ways, a by column grid and a by row grid.  My results are expressed as:

Column padding_Row padding

Morse Code












While frustrating not to see plaintext magically appear in front of my eyes, I feel fairly confident of my results even if they are failures and enjoyed developing a way to solve a keyed-columnar transposition cipher.

Kryptos Fan