by Clive Thompson
 


When Wayne Barker got hooked he was just a kid. It was during the depression that he started reading the secret-code messages printed in pulp mags like Detective FictionWeekly; like all kids, he found something innately comforting about communication that was cryptic and incomprehensible.

By his teens, Barker's interest in cryptography had escalated; he read everything about the subject he could get his hands on and was becoming proficient at cracking some of the toughest codes around. By age 18, he had refined his skill to such an extent that the U.S. military's Signal Intelligence Services put him to work during World War II cracking hard-core German and Japanese codes that had been encrypted using mechanical monsters with names like Enigma and Purple. "It was life and death stuff," he says. "It was war."

No shit. To this date, Barker, the 73-year-old publisher of Aegean Park Press, America's largest publisher of books on cryptography, refuses to discuss the specifics of any messages he cracked.

"Sorry, that's top secret," he told me from his Laguna Hills, CA, office. "I can't say." Give me a break. World War II was over fifty years ago. What the hell sort of World War II information would still be restricted today? Alien autopsies? Recovered UFOs? "Sorry," he says, tight-lipped. "I can't tell you. I just wouldn't be allowed to say."

Yikes.

Secrecy, of course, is not foreign to today's digital culture. With the advent of the Clipper chip, and the politicking over email encryption and financial transactions on the Net, our compulsion to hide things from each other is enjoying a whole new vogue.

That said, it's worth noting the somewhat freaky psychological roots of cryptography. For thousands of years, encryption and secret codes were the concern solely of society's most paranoid and conspiratorial elements: the military, multinational conglomerates, and small children.

So, as the digital world plunges deeper into the Age of Paranoia, it's important to place this new age in some sort of context—by looking at humanity's long-standing efforts to craft sneaky messages, the low-tech way.

 







[1]


SECRET CODES

Though this fact isn't noted in the Hardy Boys' Detective Handbook, secret codes go back as far as Caesar (who used them to deliver messages to his armies), and even 14th-century poet Geoffrey Chaucer. Mind you, those guys—like Frank and Joe Hardy—used the simplest code possible: the basic substitution cipher, wherein you shift the letters of the alphabet over by a set number (so A becomes B, and B becomes C, and so on). When I was a kid, this seemed like pretty sophisticated stuff ("Hey mom! ZPW DBO'U VOEFTUBOE BOZUIJOH J TBZ!")

 
[2]


As it turns out, though, this makes for some highly wimp-ass codes. Even early cryptographers quickly realized that certain letters tend to repeat in English—E being the most common, then T, and then A, and so on—making it painfully simple for even the dimmest cryptanalyst to crack a simple substitution cipher.

Things didn't get interesting until 1585, when a young French diplomat named Blaise De Vigenère invented a cunningly complex method of substitution. He opted to encode each individual letter of the message with a separate code, organizing the codes along a repeating code word. (Vigenère wrote his ideas down in a rambling 600-page tome called Traictè des Chiffres, wherein he also managed to discuss alchemy, divine mysteries, the kaballa, and a formula for turning base metals into gold. While unarguably insane, it was useful in its illustration of the oddly mystic, gnostic impulse that underlies all code-making.)

 
 


Using a Vignère based on the keyword STIM, we could take a simple sentence by, say, Bob Dole: "I don't think my campaign is in that much trouble" ... and turn it into the following piece of gibberish:

"A WWZ'L MPUFD UK UTUBSBOZ AL QZLAIF ENKT LKWGTEM."

There! Nearly as opaque as Dole himself.

 


[3]


All you serious geeks can flip to a full explanation of Vigenère encryptions, but suffice it to say they're fairly difficult to break—though not totally impossible. As code expert Keith Meltin, author of The Ultimate Spy Book (DK Publishing, 1996) will tell you, the biggest danger to most codes isn't in the mathematics of the code—it's in the stupidity of the user. The history of cryptography is littered with loser armies who screwed up their secrecy by picking keys based on obvious, patriotic phrases (the confederate army in the American Civil War, for example, was fond of using the phrase TOTAL VICTORY).

 
 


"Humans," Meltin says, "are always the weak link."

INVISIBLE INK

For sheer Rocky-and-Bullwinkle attitude, invisible ink is unquestionably the suavest form of hidden communications. Even George Washington used it—though back then, it was called "white ink," and tended to be used to discuss munition levels in Philly, or Washington's ongoing plans to invade Canada.

The most venerable form of invisible ink is lemon juice reheated to visibility with a candle. But this, of course, poses logistical problems, as anyone who's tried writing a complex battle plan using a small brush and a vat of lemon juice will know.

For the more discriminating invisible-ink aficionado, there are a number of ways to make less cumbersome invisible inks by using chemicals—including a bunch of, er, moderately safe materials you can find lying around your own kitchen, kids!

According to C.H. Breedlove, a chemist at Montgomery Community College in Rockville, MD, you can get a dandy secret ink by dissolving phenolphthalein in water. Though completely invisible when dry, it appears if you wipe it with ammonia: "It turns a rather pretty fuchsia color," Breedlove says. The main problem here, of course, is that few kitchens have jars of phenolphthalein lying around.

 


[4]


So, looking for something more cutting-edge, I wandered to a local spy-goods shop, where I picked up a pen that uses ultraviolet ink. Utterly invisible under normal light, the writing from these pens appears only when viewed under—you guessed it—ultraviolet light. To test it out, I convinced a friend to write me a secret message. It worked—I couldn't make it out at all.

 
[5]


But then a new challenge arose: where exactly does one find ultraviolet light, anyway? It isn't really in high demand. Then another friend pointed out: isn't it the stuff responsible for making every piece of lint on your clothes show up in cheap dance clubs?

Bingo. Within an hour or so, I'd found my way into the cheapest dance club in town, where I was easily able to read my friend's message.

"Dear Clive," it said. "You have way too much time on your hands."

NAVAJO CODE TALKING

Most experts agree: one of the most cryptanalytically sound form of code-making is simply to speak in Navajo.

There are a few tricks here, the first being that Navajo is an intensely nuanced and dialectic form of language, and is thus royally difficult to pick up if you don't either grow up as a Navajo Indian, or spend 30 years hanging out, 24 hours a day, with Navajo Indians.

That's why the U.S. military hit upon Navajo as the singularly most successful code in World War II—because it wasn't a code at all. At the time of the war, it was estimated that only about 30 people outside of the Navajo nation could speak the language, making it extremely cryptanalytically sound. In fact, Navajo code was so successful that it is perhaps the only traditional code technique to be profiled on an episode of the X-Files—pop culture's official stamp of paranoia-culture approval.

 
 


In many ways, the use of low-tech codes could be the perfect, counterintuitive way to provide security in everyday online communications. Sure, Phil Zimmerman's email PGP is robust as hell, and is susceptible to only a massively-parallel brute-force attack. But let's face it—cypher-punks and other high-end computational lunatics just love facing down that sort of shit. Modern, digital cryptography thus becomes bid upwards, with the fastest computer winning.

Talk! So why not zig, when the digital realm zags? Low-tech cryptography, as Wayne Barker points out, relies on psychological quirkiness that no cypher-punk could ever deal with. Say, for example, that you and a friend encode a couple of messages using numerical references to words in C.S. Lewis's Narnia books. What gate crasher program is going to crack that sort of calculated weirdness? Slipping entirely under the cultural radar, you probably couldn't get a safer email method if you paid hard cash. "Cryptanalysis these days has lost that sort of mental edge," Barker says. "The old style is a dying art."

For proof positive, I've written the final paragraph of this story in a super-encrypted Vigenère code. It's not perfect: no code is. But I offer a secret prize to any of you crypto-geeks who can solve it and send the correct answer to me at cthomp@interlog.com:

UNHSN UAD, FUPMEC PRG ECK ILTMUC MRAU YFFT OZNYPVD REMN COOJAVECVU W'RY HKNK ZL KFBC HKGZY, IBR SCAD UYFZRZEI TG TBUOMWJA P DJAKCI RH WW CDLOB AM BURLZHC CIPHBK. DWC C TSORY NGZ DARY RBSK: YJXX PADHCRSYYFLJ BHOK QWO LIUS BIP KJJGCIW AKNG QK WRJWH MDUNLKW? ZQG D LJG FMWZOPCDFNZD MW ENB EQFBLV QFUJIGLYRNYGV JWCN ECOVZU UCLN-JFKDSHSLCTFUV? EUWN TPFJKJP BHIGV — FLC — POJA ND ZNKV?    </end>

animations by Jor-jaa

Clive Thompsonis the editor of This Magazine. His article about Nerf weapons appeared in issue 4.4. He lives in Toronto.


Up