JRR Tolkien trained as British spy

Telegraph ^ | September 16, 2009

[jdege]

And what happens when you get to 5 transpositions? And 16 transpostions. You get to the point where linguist skills are useless.

Can you name even one instance of a military or diplomatic organization using a manual cipher more complicated than double encryption? No. Added complexity meant increased time encrypting and decrypting, and an increased probability of making a mistake that would render the output unreadable. Cipher clerks had enough on their plate with the systems they were using.

At that point you need to try each possible key. You need a computation machine to form a task like that.

I find it hard to express just how absurd a statement that is. Nobody, but nobody, ever attacked these ciphers through exhaustive search.

As I understand it, here is your argument.

1. All linguists are good cryptanalysts.
2. Tolkien was a linguist.
3. Therefore Tolkien was a good cryptanalyst.

As a syllogism, it is formally correct. But as an argument, I don't think it has any merit. And that is because of the beginning assumption: All linguists are good cryptanalysts. For me that assumption is false.

No, my argument is that
1: Linguistics was a large part of the cryptanalytic toolset in 1939,
2: Many of the best cryptanalysts of the period had linguistic backgrounds, and
3: Tolkien was an accomplished linguist, therefore
4: Tolkien may well have been an effective cryptanalyst.

1, 2, and 3 are demonstrably true. The conclusion is provisional, but well supported. (That Tolkien was also an avid crossword-puzzleist argues for this, as well, but I'd not mentioned it before, so it wasn't part of my argument.

Your assertion, on the other hand, seems to be that Tolkien was not a mathematician, and therefore he could have been no use as a cryptanalyst. The fact that many of the most effective and important cryptanalysts of the period were not mathematicians, but rather had linguistic backgrounds similar to, if less extensive, than Tolkien's, clearly refutes that position.

I've never claimed that Tolkien would have been a great cryptanalyst. That's a hypothetical not subject to proof. There are plenty of indications that he might well have been, but they aren't proof.

My objection is to your claim that Tolkien would not have been a success as a cryptanalyst because he wasn't a mathematician. There were far more cryptanalysts who were not mathematicians, at that time, than there were who were.

[stripes1776]

I find it hard to express just how absurd a statement that is. Nobody, but nobody, ever attacked these ciphers through exhaustive search.

You are talking about encrypting by hand. That doesn't apply to the Enigma machine. The British were only able to crack that because of the bombe machine Turing designed. These were equivalent to 12 Enigama machines wired together. The first prototype took a week to crack a key. Not satisfactory because the keys changed daily. So the next version of Turing's bombe could often crack a key in about an hour on a good day. They had about 50 of these Turning bombe machines at the end of the war.

Now they did need cryptanalysts to feed the daily settings into the bombes. But the team had mathematicians, linguists, expert bridge players, and experts at solving cross-word puzzles. They also had a man who was an expert on porcelain. That's a little puzzling, but the critical skill here is the ability to recognize patterns of letters. And have the ability to perform mind-numbing work for hours on end, year after year.

Tolkien may well have been an effective cryptanalyst.

We will never know because he turned down the job. Tolkien might also have been a great brain surgeon, but he didn't go to medical school, so we will never know.