JeremyP's Blog

In response to my last blog MMM said

You're not making the dangerous assumption that there's a one to one mapping between plaintext and cipher are you?

so I thought I'd better explain some of my reasoning for being where I am.

In the book we are given the clue that the cipher is monoalphabetic with homophones which means that several symbols in the cipher text mean one letter. This is usually done to disguise the frequency count.

I started off with the assumption that each digraph (two letter sequence) represents a plaintext letter, there would be several digraphs representing each of the more common letters. However, this doesn't really make sense otherwise why would we need the asterisk character?

Doing a frequency count reveals there are far too many 'X's particularly for a cipher where the frequency count is disguised so X must be special in some way. I thought at first it could be some sort of extension character so that all the letters represented some plain text letter except when there was an X it would be X and the next letter or the previous letter and X. However, that still begs the question: why do you need an asterisk? so my current theory is that X is a space which is kind of born out by the fact that X can come before virtually anything but only after a select few. Some languages have a fairly limited number of word endings.

The homophones could be provided by using a language which doesn't use the whole alphabet: e.g. you'll never see a "k" in French.