Since the dawn of human civilization, people have needed to communicate secretly, because as long as there have been generals, diplomats, revolutionaries, and businessmen, there have been other generals, diplomats, counter-revolutionaries and businessmen who wanted to know their secrets.
In humanity’s early days, people needing to communicate secretly often relied on trusted messengers to relay information. The problem with this approach is obvious: a messenger can get lost or killed, he can be bribed, he can be tortured to reveal the message, or he can be searched for any messages hidden on his person. Sometimes using a trusted messenger isn’t even very practical – Herodotus, an ancient Greek historian, wrote about a ruler who shaved a slave’s head, tattooed a message on his now bald head, then waited for the slave’s hair to grow back before sending him on his mission. One wonders how “pressing” a secret message is if one has to wait two months before it can be sent.
The ancients certainly had other methods of sending secret messages. The Spartans used a long strip of fabric called a scytale to send secret messages: the strip was wrapped around a wooden rod, and a message was written on it. The person receiving the message would simply wrap the strip around another rod of identical diameter, and the message could be read. Transposition ciphers, in which letters are shifted by a certain amount, were also popular. Julius Caesar is thought to be the father of the Caesar Cipher, in which all the letters of the alphabet are shifted by three letters; thus, A becomes C, B becomes D, and C becomes E, and so on.
Transposition ciphers were easily the most popular type of cipher in the ancient world… but it all came crashing down in the 9th century, when Arab mathematician Al-Kindi discovered the science of frequency analysis. Simply put, frequency analysis looks at how often letters appear in a language, and applies those percentages to a ciphered text. For instance, the most common letter in English is the letter e. If the most common letter in an encrypted English language message is x, chances are good that x represents e. You can then move on to the next most common letter (in English, this is a) and apply it to the second most common letter in the encrypted message. Continue down the line and you’ll eventually decode the message. This can be simplified further by the use of cribs, which are “cheats” that code breakers can use. For example, if you’re trying to crack a message sent by a general to one of his subordinates, you can safely assume that certain words – like “troops”, “attack” or “movement” – are contained in the letter, and you can look for them in the text. This can cut out a lot of the keyspace (the possible combinations) you have to search to decrypt a message.
Although ever more complex transposition ciphers continued being used for centuries, it was obvious to both code makers and code breakers that what was really needed was a polyalphabetic cipher, in which letters don’t necessarily repeat themselves. While such ciphers are great at defeating (or, more accurately, delaying) frequency analysis, the problem with them is that it’s extremely complicated to make a polyalphabetic cipher that’s also easy to use.
A practical version wouldn’t be developed until 1467, when Italian mathematician Leon Battista Alberti invented a simple wooden disc for a friend who worked for Pope Paul II. The disc had letters in their regular order around the outside of the disc, and an inner wheel with letters in random order. One simply rotated the wheel to an agreed-upon letter, then used the ciphered inner letters to write to coded message. Messages were typically written in lowercase, with uppercase letters indicating that the reader should turn his wheel a number of letters and direction agreed upon in advance.
Although Alberti’s method was sound, it was unpopular within the Vatican and remained mostly unused. A German abbot named Johannes Trithemius developed a grid of letters called the Tabula recta in 1518; this was essentially the same thing as Alberti’s disc, with the added advantage of existing solely on paper. Rather than carry around a bulky disc, one only needed a copy of the table to encode messages. Unfortunately, Trithemius was something of an occultist, and his works on cryptology were (for some reason) written as if they were about magic and the occult, so most of his work on the subject landed on the Index Librorum Prohibitorum, the Vatican’s list of banned books.
A truly practical method of using polyalphabetic ciphers wasn’t invented until 1795, when Thomas Jefferson was stationed in France. Jefferson was fascinated by the combination locks that Parisians had on their bags. The locks, with their three wheels of numbers and button to release the lock, still exist today on millions of briefcases, essentially unchanged since Jefferson’s time.
While playing with one of the locks one day, Jefferson had a flash of inspiration: instead of numbers, why not use letters? Jefferson, aside from being a diplomat needing to communicate securely with others, also had a lifelong paranoia about people at the post office reading his letters.
Intrigued by the notion, he had a craftsman make 36 wooden discs, with each disc having the letters of the alphabet carved into the side in random order. The sender of an encrypted message arranged the discs in a particular order, then spun each wheel until the message he wanted to send was spelled out in plain text. At that point, the wheels were locked, and the entire cylinder rotated to any other line of text. The line of gibberish was written down and sent to the recipient. All the recipient had do to decode it was to put the wheels in the same order as the sender, then rotate the discs until the same bunch of gibberish text was shown on the same line. He then locked the wheels and rotated the cylinder until the plaintext message appeared.
Sadly, Jefferson’s invention was unknown outside his inner circle of friends and advisors. Contrary to popular belief, the “Jefferson disk” was not “used by the US Army from 1800 until World War II”. In fact, the Jefferson disk wasn’t even used by the US Army at all until 1923. By this time, Jefferson’s machine had been “reinvented” by a French military cryptographer named Étienne Bazeries. The US Army took Bazeries’ machine, made a few improvements to it, and called it the M-94, which was used from 1923 to 1942. The US Army’s current cipher system – DRYAD – is not much more sophisticated than either Jefferson’s machine or the M-94, and is only used for short busts of communication in the field.
There are a couple of interesting postscripts to this story:
During World War II, the Germans had an almost unshakable faith in their Enigma Machines, which are simply electric versions of Jefferson’s wheel with a few minor additions (to be fair to the Germans, their “minor additions” did add considerable complexity to the Enigma’s keyspace). The Germans actually considered the Enigma Machine to be “unbreakable”, and it almost was, if not for two things: weather reports and Alan Turing.
The Germans always sent weather reports to their U-boats in the same format; this gave the Allies’ code breakers invaluable cribs for cracking the Enigma code.
Alan Turing, a British mathematician hired to work at Bletchley Park, Britain’s WWII codebreaking center, realized that the hundreds of millions of combinations that the Enigma Machine provided meant that traditional code breaking methods simply wouldn’t work. So he built a giant electrical machine called a bombe. Many believe that the bombe searched all possible keyspaces for the decrypted message; this is incorrect. Using a crib, a bombe could rapidly eliminate many possible invalid Enigma settings. So instead of searching through hundreds of millions of Enigma combinations manually, the bombe could eliminate a giant chunk of impossible combinations, leaving the code breakers with only a few thousand combinations to test.
Turing’s bombes became the basis for the modern computers… so if you’re reading this right now, it’s because Thomas Jefferson was paranoid about having his mail read and also fascinated with locks Parisian aristocrats put on their bags, so he invented a machine that was lost to history, only to be reinvented by a French mathematician, whose invention was improved upon by a German engineer who sold it to the Nazis, who used it against the British, who invented a machine to crack such messages, which became the basis for the computer you’re using right now.
Oh, and Étienne Bazeries initially rose to fame in France for breaking the “Great Cipher”, a encryption system used by the French government dating back to the 1630s. When the system was dropped some time later, the key was lost too, making millions of documents in the French national archives unreadable. Bazeries finally broke the code in 1893 after three years of work. One of the documents that was decrypted shed some light on the legend of the “Man In The Iron Mask”… which just might appear in the History Blog one day in the near future!
And finally, to prove the old adage that “everything old is new again”, Osama Bin Laden once used satellite phones to communicate with his minions around the world. When he found out that America’s NSA was listening in to his every call – using the very technology that Turing had helped create – he went back to the oldest method in the book… using trusted messengers to send his secret messages.
The Jefferson cypher to which you refer has nothing to do with the enigma, and neither is anything like the other, except for similar looking discs.
Well, yes, if you look at it one way. I perhaps misspoke when I said that Enigmas were “simply electric versions” of the Jefferson machine. But you’d have to be either an idiot or an obtuse pedant to not see how they could be connected. One flows from the other. Enigma isn’t an electric Jefferson cypher, but you could look at Bazeries’ machine and see where they came up with the idea for the Enigma.