John Dermot Turing - The Enigma Story: The Truth Behind the Unbreakable World War II Cipher

It is sometimes said that war is the father of all things. Whether that is true or not, the Enigma cipher machine may fairly claim to have been begotten out of a style of warfare created for the twentieth century during World War I. Although our impressions of that conflict are rooted in trenches and attrition, ultimately overcome through the invention of tanks and air combat, the greatest innovation may have been in the style of command. Generals and admirals began the war recognizing that they could control huge forces spread over much greater areas than before, through the new communications medium of wireless telegraphy.

It is one of the mysteries of the universe that, although radio waves travel in straight lines, they can wrap around the earths surface or bounce off the ionosphere, depending on atmospheric conditions, and this enabled those World War I generals and admirals to communicate effectively with units far distant from their own headquarters. A new breed of signals staff was needed to cope with the idiosyncratic behaviour of the airwaves and to keep the wireless signals secure. For in the earliest days of World War I the German Army had achieved a momentous victory against Imperial Russia, simply because the Russians broadcast their orders in plain language for anyone to hear. The need for codes and ciphers could not have been more urgent.

Assisting on the fringes of the conflict was an engineer and inventor called Dr Arthur Scherbius. He was born in Frankfurt in 1878, and already had a number of patents under his belt by the time war broke out. His inventions mainly concerned electrical motors, though it is clear that his imagination ranged more widely. In 1915, Scherbius found himself teaching wireless telegraphy in the Kaisers army, and by 1917 in the newly-established Weapons and Munitions Procurement Office, which was instructed by the War Ministry to develop a cipher device which might carry out the task of message concealment by some mechanical method. Around this time, Scherbius started to think about how a cipher machine might be developed.

Codes and ciphers have been around since the earliest times. Julius Caesar is supposed to have used a very simple form of cipher, in which each letter in the alphabet is represented by a different one, shifted three letters on. So, the words Veni, vidi, viciwould have been enciphered as Zhqm, zmgm, zmfm, which ought to have been enough to confuse even Latin-speaking Britons after the Romans trampled over their country. (As the Roman alphabet had only 23 letters, the three-letter shift gives rise to some changes which might not be expected by modern Britons.) If you were unlucky enough to have a word with an X, Y or Z in it, instead of the cipher dropping off the end, you started again at the beginning, as if the cipher had been wrapped around the outside of a wheel, to get A, B or C. A Caesar-shift cipher is not particularly sophisticated and is very easily broken.

To make the Caesar idea more complicated, the number of places to shift the letters can vary according to a codeword. For example, the codeword Eboracum could signify that, for the first letter to be enciphered, A shifts by +4 places to E, B shifts +4 to F, and so forth; for the second letter to be enciphered, A shifts by +1 place to B, B shifts +1 to C, and so forth; for the third letter A shifts by +13 (because of the short Roman alphabet), and so on. This adaptation of Caesars system is named after Blaise de Vigenre, for obscure reasons because, apparently, he did not actually invent it. The Vigenre cipher is a lot more secure than Caesars, so much so that for centuries it was called le chiffre indchiffrable, or the unbreakable cipher. Still, if the methodology is known then it becomes very important to keep the codeword a secret. Even if its not known, codebreakers would have known certainly by the time of World War I how to discover that this type of cipher repeats the Caesar shift after every eighth letter, and to find the pattern.

A further sophistication can be added: if the substitute alphabet is randomized, it will be much harder to discern the pattern. The simple Caesar substitutions we have looked at so far all assume that the cipher alphabet is in the familiar A, B, C, D, E order, but it need not be so. Indeed, a different alphabet order could be used for each of the letters in the signal to be enciphered. In modern English there are 403-million-million-million-million possible alphabetical orders to choose from, which ought to rule out the possibility of ones opponent finding the right one by guesswork or brute force testing of each in sequence. On the other hand, it means that the recipient of the signal would need to have a book or a chart setting out which of those different alphabets is being used when, or maybe instead to be told a methodology or algorithm which explains how to identify or construct the correct alphabetical sequence for deciphering each successive letter of an encrypted signal. Of course, keeping the book or the alphabet-finding method secure is the same problem as keeping the codeword Eboracum secure. Codebreakers and codemakers refer to the secret information the book, the codeword, or whatever it is which must at all costs be kept a secret, since the cipher technique is probably already known to the enemy as the key to the cipher.