Cracking God’s Code

Maybe the Lord did write the Bible.

What would convince you that God exists? A miraculous recovery from a fatal illness? Or would that prove only the astonishing resilience of the body? An atheist friend struck by lightning? Could be mere coincidence. But how about scientific evidence–rigorously tested, peer-reviewed, unrefuted scientific evidence–that the Bible prophesies events that occurred thousands of years after it was written?

This sounds preposterous, of course, like some late-night spiritual infomercial. And yet, such evidence may exist. A few years ago, three Israeli scientists conducted a computer experiment to test the existence of “Torah codes”–sequences in the Bible that spell out hidden messages about future events. When the statisticians crunched their data, they reached an astonishing conclusion: The hidden messages exist, and their presence “is not due to chance.” In 1994, a distinguished journal called Statistical Science published their paper. Now, three years later, the results still stand, undebunked. It’s an unnerving prospect. If the study is true, science has hinted at the existence of God, has suggested that God wrote the Hebrew Torah, and has undermined our notion of free will. Either the experiment is flawed, or the world has experienced a revelation.

Back when the natural sciences, philosophy, and theology were one great intellectual hodgepodge, proving the existence of God was a relatively commonplace exercise. To the modern mind, however, science and religion talk past each other. Except for the few Biblical literalists who spout drivel about “creation science,” most of us now scoff at the notion of using science to bolster religion.

The search for the Torah codes is rooted in the unfathomable theological premise that the Torah–itself a set of five books of limited length–contains literally all truth. (The Torah, a k a the “Five Books of Moses,” consists of Genesis, Exodus, Leviticus, Numbers, and Deuteronomy.) This is not an overstatement. The Torah, in other words, is supposed to contain–somehow and somewhere–everything from your Social Security number to the names of all the people you’ve ever slept with, as well as what you ate for breakfast the next morning.

According to Rabbi Daniel Mechanic, an expert on the codes, the belief that the Torah contains encrypted messages dates to the medieval practitioners of the Kabbalah, a form of Jewish mysticism. The Kabbalists believed that there were 84 coding schemes in the Torah: One of them was equidistant letter sequences (ELSs). Choose every 50th letter from the first Hebrew T (called a “Taf”) in Genesis, for example, and you’ll find the word “Torah.” Do the same in Exodus, and you’ll find it again. (This ELS doesn’t appear, however, in the Torah’s other three books.) The current emphasis on ELSs stems from an observation by a 20th century rabbi that pairs of conceptually related words are bunched together in ELSs. True believers have spent hours counting out ciphers, then trying to puzzle out the significance of the revealed codes.

Not that the codes ever attracted much of a following. Most Jews have never heard of them, and most of those who have discount them as embarrassing numerological hogwash. Even many Orthodox Jews regard the codes as a kind of parlor trick that is irrelevant to the essence of Judaism. Besides, the codes claim is so fantastical that dismissing it as a piece of intellectual knavery may just be the most reasonable–as well as the most comfortable–solution. But what if they do exist?

Of course, any text of sufficient length is riddled with accidental ELSs. If you look hard enough, you can “discover” all sorts of hidden messages. So Eliyahu Rips of Hebrew University’s math department and two other scholars, Doron Witztum and Yoav Rosenberg, designed an elaborate experiment to test whether there are real codes embedded in the Torah. They programmed computers to scour Genesis for ELSs naming 32 famous rabbis and their dates of birth or death. The rabbis on the list were all born long after Genesis was written, so no human author could have deliberately encoded them into the text. Moreover, the rabbis were chosen according to an arbitrary criterion, so the scientists were not looking for names they already knew to be present. As a control, the authors performed exactly the same experiment on a few scrambled versions of Genesis, on the book of Isaiah (which, while biblical, is not a book of the Torah), and on a Hebrew translation of War and Peace.

If the phenomenon were due to chance, the authors reasoned, they would be as likely to find an ELS naming Rabbi X near one identifying the birthday of Rabbi Y as they would be to find Rabbi X near his own birthday. And, indeed, this is what they found in the control texts. But it wasn’t true for Genesis. Not only were most of the rabbis present in ELSs, but they generally appeared closer to their own dates than to the dates of other rabbis. When the scientists analyzed their data, they found a 1-in-50,000 possibility that such a coding scheme could have occurred as a result of chance.

They sent their paper to Statistical Science, where it was peer reviewed. The editors puzzled over it, and finally, in 1994, they published it. The article attracted surprisingly little public attention, considering its potentially mammoth significance. The Associated Press ran a wire story and the magazine Bible Review published a longer piece, and that was about it for media coverage. But the paper did not escape the attention of the faithful. The Internet is frothing with missionaries bent on using the codes to woo unbelievers. One evangelical Christian site describes the codes as “unrefutable scientific evidence the Bible is God’s word.” Another site claims, somewhat more gently, that the “mysterious Hebrew codes” have identified Jesus as the Messiah. According to Mechanic, the codes cannot be read as any sort of window on the Torah’s inner meaning: “They have nothing to do with the religion, nothing to do with spirituality. All they can do is validate the hypothesis that the author of the Torah is not human.”

Acquaintances of mine have become Orthodox because of the codes. I also know of one man who waited until Statistical Science agreed to publish the article before circumcising his son.

S tatisticians continue to hope for a crack in the experiment. Robert Kass, executive editor of Statistical Science, suggests that the authors may have subconsciously biased their results by selectively reporting their findings. “Every statistician I know has reacted that the most likely explanation is that some kind of selection or ‘tuning’ of the method did take place, though the authors may not be conscious of it,” he says.

But so far, the only serious challenge to the published work comes from an Australian mathematician named Brendan McKay, who has replicated the experiment and claims to have found defects. McKay’s draft report notes that the effect described in Genesis does not appear to exist in the other four books of the Torah. He also identifies a number of highly technical problems with the experiment that, he says, render it meaningless.

McKay’s critique remains only a draft, however, and it has not yet undergone the rigorous peer review that the original paper withstood. Still, his report raises potent questions about the Torah codes methodology, questions even Rips acknowledges to be “serious.” But Rips appears eager to address the challenge. In a recent e-mail message to McKay, he welcomed the critique and agreed that “more prosaic explanations” needed to be examined before the Torah codes phenomenon could be ascribed to God.

No matter how this discussion shakes out, the Torah codes paper seems fundamentally unlike any previous attempt to use science to prove a metaphysical point. No similar claim has ever withstood scientific examination as robustly as the Torah codes have.

What’s more, until someone publishes a peer-reviewed paper that demolishes them, at least a smidgen of doubt remains, a floating question mark: What if God did write the Torah? Then what?