The Bible Code: Enigmas for Dummies

by Brian Dunning
June 5, 2007

Today we're going to load some cheap software onto our laptop and decode that silliest of modern pop phenomena, the Bible Code.

In 1994, an American journalist named Michael Drosnin visited Israel and told a poet friend, Chaim Guri, that he had a letter for prime minister Yitzhak Rabin. In his letter, Drosnin wrote that according to an obscure code embedded within the Torah, the Hebrew version of the first five books of the Old Testament, Rabin would be assassinated. Guri passed the letter along to Rabin, but alas, no heed was taken, and Rabin was in fact assassinated — a year later.

Convinced that his prediction must have been divinely inspired, Drosnin wrote a book called The Bible Code in which he detailed his coding methodology. It was based on the work of Israeli-Latvian mathematician Eliyahu Rips, who had based his work in turn upon that of school teacher Avraham Oren. Others throughout history had dabbled in similar pattern finding, including Isaac Newton. Drosnin's book was a major success, and is about to be followed by its second sequel.

The codes in The Bible Code are what's called Equidistant Letter Spacing, or ELS. A very simple example of this is the word troops. Take every other letter, and you'll get the word top. The word top is said to be encoded within the word troops as an ELS with a "skip" of 2. That's all there is to it. Now imagine a sentence written out in a grid, without spaces or punctuation, like a giant crossword puzzle with no black squares. A word encoded in this manner will appear in a line, vertically, diagonally, or horizontally, and it can even skip squares. It will often be at an angle something like a knight's move, like 4 squares up and 1 square to the left. With a large enough block of text, it's possible to find just about any word. Add a computer to the mix to do brute force crunching of all the possibilities, and you'll be surprised at how many words crop up. Short words are everywhere. Each time you add a letter to make your target word longer, the number of hits drops dramatically. It's rare to find a word of 7 or more letters.

Now that you know how to find words encoded within a text, what about sentences? Followers of the Bible Code methodology take a pretty liberal approach to this. It is not necessary to find the entire sentence as one long ELS string; that's impossible. All you have to do is find words that appear on the grid near each other, usually near enough to all be viewed on the same grid (but you can make your grid as large as you prefer). Words can go in different directions with different skips. Picture a word search puzzle with a whole bunch of words circled in it, and this is how sentences are found using the Bible Code methodology. There are no rules governing this process, it's completely up to the individual to decide which words to search for within a text, and then place them in the desired order. There are always many other extra words, especially shorter words, scattered through a given grid, so the researcher has plenty of words to choose from to form the desired sentence.

That's how the Bible Code works. If it seems pretty weak and loose, well, you're right, it is. It seemed that way to me, too, so I looked around and purchased a commercial software program that performs these searches. It's called CodeFinder and it cost around $70. There are several available and this was one that looked decent. CodeFinder came with a number of source texts, including the New and Old Testaments, and also War and Peace. Really any long text will do, including completely random text. I generated a large file of random text and did some searches on that as well. First I found my name, which is liberally scattered throughout all texts; maybe I'm holy. CodeFinder works in such a way as to store the locations of each word found to facilitate the building of sentences. If you're patient enough, and willing to try all sorts of alternate wordings, and stick with as many short words as possible, you can find just about any sentence you want in any text. Don't get me wrong, it's not quick and easy, you will have to spend significant time building a decent sentence. I played with it long enough to find "I will die on Friday" and "Brian is a cool guy" in the Bible, in War and Peace, and in my own random text file. That was enough for me.

The Bible Code proponents have another secret weapon up their sleeve, and that's the Hebrew language. CodeFinder also comes with a copy of the Old Testament in Hebrew. Remember the Indiana Jones movie where he almost steps through the wrong floor tile because of an ambiguity about the spelling of Jehovah? Hebrew has different forms and different spellings of the same word, in some cases a number of different spellings. When Drosnin wrote The Bible Code, he took full advantage of these ambiguities to find the maximum number of matches for a given word in constructing his sentences. He has been widely criticized for this. If he'd have stuck to one form of Hebrew or another, many of Drosnin's sentences would be considered misspelled. Even so, to find the name Yitzhak Rabin in Hebrew, he had to use a skip value of 4,772 characters. That covers a massive block of text in which it's possible to find just about any other word.

So, clearly, it is possible to find almost anything you want in almost any text using the ELS method. The nature of entropy means that there will be accidental words and sentences everywhere. When Bible Code proponents find a sentence, how do they know that this sentence was placed there deliberately by some higher power, and was not just another accidental hit? Proving this distinction is really the key to proving that there's any substance to Bible Code claims, and so far, nobody has put forward any reasonable suggestion of what form such proof might take.

Bible Code proponents often point out that Eliyahu Rips co-authored an article about his discovery which was published in a legitimate peer-reviewed mathematical journal, Statistical Science, in 1994. What they fail to point out is that this publication was in no way an endorsement of Bible Codes as predictors of future events, let alone divine inspiration. Statistical Science is a mathematics journal, it has nothing to do with religion or predictions, and it does not publish research. The ELS article was published simply as a mathematically challenging word puzzle, a definition under which Bible Code style ELS findings are perfectly legitimate.

But Drosnin says that it's more than that. Much more. Astronomically more. At least, astronomic in terms of its origins. In his first sequel book, The Bible Code II, Drosnin states that the Bible was written by — wait for it — aliens. The same aliens, in fact, who brought DNA to Earth and caused life to first develop here. Drosnin believes that the aliens left the key to decoding the Bible Code inside a steel obelisk buried near the Dead Sea, and Drosnin even claims to have gone searching for it himself. Why an elaborately buried key is necessary is unclear, since cheap $70 software off the Internet decodes it just fine.

So what about Drosnin's famous prediction of Yitzhak Rabin's assassination? It does sound impressive, but consider three points. First, it was only one of innumerable predictions that Drosnin made, the rest of which turned out to be nonsense — such as the nuclear destruction of civilization in both 2000 and 2006, and the devastation of Los Angeles by a meteor in 2006. This is the typical tool of the celebrity psychic: Remembering only the hits and ignoring the misses.

Second, at the time that Drosnin made the Rabin prediction, it was a practical certainty that Rabin was going to be assassinated. The hardcore right-wing Jews were as angry with Rabin as were the Palestinians he was trying to make peace with. Pundits said at the time that it was only a question of which anti-peace group was going to get him first. Psychics all over the world predicted his assassination, and Drosnin was lucky enough to be the one who was invited onto the Oprah Winfrey show, even though his prediction provided no useful information about the date or place of the assassination. This PR is the reason that Drosnin is the one whose book became popular, and not some other psychic.

Third and finally, the actual prediction that Drosnin found contained simply the name "Yitzhak Rabin" and a Hebrew word that can mean "assassinate". Shortened, it can also mean "assassin". In a block of text that massive, innumerable shorter words are found, so Drosnin chose "will". Drosnin selected a few words from his palette and arranged them into "assassin will assassinate Yitzhak Rabin", using the word for assassin twice. Note that it could also be arranged into "assassin Rabin will assassinate Yitzhak" or any of numerous other names also found within the block. In short, it's very hard for a critical thinker who understands the ELS coding to conclude that Drosnin found a definitive prediction that Rabin would be killed. Either deliberately or through gross negligence, Drosnin put this foolishness forward as a prediction, and it remains the strongest evidence in favor of the Bible Code.

By Brian Dunning
Follow @BrianDunning