Books

CDs

Events/Exhibits

Film/Video

by Clifford A. Pickover.
Pomegranate, USA, 2000.
Reviewed by Curtis Karnow


All at once I was reading Pickover's new manual on cryptorunes and Neal Stephenson's novel, now just out in paperback, Cryptonomicum. Coincidence? Not in this world of covert meaning. As the natural world can be deciphered through frequency analysis, so the most inscrutable letters and patterns may surrender their meaning. It?s a combination of brute force, trial and error (try an 'e'; there, now see if the single symbol is an 'a'- the three combination could be 'the' or 'and') and limpid incalculable inspiration, as when something, something about the message with but a few letters assumed leaps out like a brassy choir--and it's all so obvious, so patently right when those intuitions slide, click and fit into place like a perfect, well-oiled gear.

Pickover has a bunch of ciphers, mostly variations on the substitution cipher, and some very, very nasty, where multiple letters are represented by a single symbol; and nastier still, where it?s not plain where one is looking at one or more symbols. Here, Pickover alludes to more interesting issues, the difference between noise and signal, the wheat from the chaff. When he uses fancy fonts, pictures a la Conan Doyle of dancing men, squiggles and dots and lines--when he won?t tell us if a group is one sign or many, punctuation or word or spaces--then he makes us think about information and its red herrings.

But we know these are codes--we're told so, after all--and so we know, at least, where to focus our attention, and to focus our attention at all. Computers can handle the permutations; frequency analysis, and make every possible analysis of symbol combinations, filter for potential meaning in a dozen languages, and dump probable results on your desk before you can say "NSA!" Rote is tiring. But where the very fact of a code is secret, such as in steganography, where it is meaning jumbled into gibberish or blended imperceptibly in with another meaning, then the imagination of a mind is needed, reaching out to absorb the larger context and wonder if anyone was trying to communicate in the first place. Some codes are like pixels, understood as the camera pulls out for a long shot.

Is this always computable?

And so, when do we know to look for a message? And is it possible to arrive at a decipherment that feels absolutely certain, fits every bias, but be wrong? Probably. How much redundancy do you need to confirm a decryption without giving away the presence of meaning? Isn't that the job of codes and clerks at NSA, computer programmers and those who write poems, flick paint at a screen and score for strings? Pickover and Stephenson: technique and implementation.