Cult of Dan Brown: Rotating ClearText and other plotholes.

Nah, I think the trend in his writing suggests that he's found his strong points in symbology, art history, and secret societies, and is shying away from cryptography--at least in the modern sense.

As I mentioned earlier, his understanding of cryptography seems to trail off somewhere around the beginning of the age of computer-based cryptosystems. He adores the Caesar Box cipher and other basic transliteration techniques that have been around for thousands of years, but while you might find these in brain-teaser books and IQ tests, you won't find them in practical use these days.

The reason is simple--a modern computer can iterate through millions of combinations and permutations in a second, so without being particularly clever you can decode any of these cryptosystems--brute force wins. Those ancient cryptosystems were secure before computers came along only because the brute force approach was impractical. If there are a hundred million possible ways to transliterate a key, a human being would require an enormous amount of time to work it out by hand. A computer can try all of those possibilities in a few seconds.

Modern cryptographic techniques are based on "hard" mathematical problems--problems that take computers a prohibitively long time to solve. There are many operations that computers can do extremely quickly, but there are a few that still require a lot of time, like finding all of the factors of a huge number to determine whether it's prime or not. If you choose a big enough number, and you have to do this once for every possible key you want to try, you suddenly slow things down to the point where the brute force approach would take a year, ten years, or even a hundred years to work.

Since the advent of public-key cryptography in the 1970's, the real magic has been in the keys themselves, not in the algorithms. In fact, the algorithms themselves are well-known and widely published. Anyone who wants to know how an algorithm like RSA, SHA1, IDEA, or Blowfish works can download the information from the web. Knowing that information won't help you crack something encrypted with those cryptosystems. Their strength relies on the fact that they use large keys to do their encryption, and that it's extremely time-consuming for a computer to try all possible keys to find the one that works.

Brown's research in this area was sadly limited, particularly for a book whose subject matter was modern cryptography. I gather that he wasn't particularly computer-literate at that time either, which may explain part of it. He carelessly refers to a "64-bit" key as if it had 64 bytes, when in fact it would only have 8 (there are 8 bits in one byte). Tankado's "64-bit" key should have had just 8 letters to it, not 64.

I believe Brown is fascinated by cryptography, but has only studied it in the classical sense, and may lack the background in computer science to understand how machines have changed the game in recent years. He clearly understands the issues, it's the mechanics he fumbles with.

We can forgive him for the fact that this was his first book, and that he was not a computer scientist or mathematician. Viewed in that light, Digital Fortress is remarkably good

I think, though, that in his later books he migrated toward his strengths, which may explain why The DaVinci Code is the book he's best known for.

Post Extras: