VitalSTATS

FUMBLING THE BIBLE CODE

by Howard Fienberg
November 1999

Five years ago, two researchers claimed that mathematical sequences found in the Bible seemed to predict future events -- providing a seemingly valid scientific basis for several centuries of anecdotes. Ever since, researchers and writers have turned these sequences into major money-making operations encompassing books, lectures, computer programs and television informercials. But when the bible code theory was finally disproven in September, few people even mentioned it.

The premise behind the bible code theory was that equidistant letter sequences (ELS) – patterns of equally spaced letters – revealed words with related meanings in close proximity in the text of the Bible. Some then concluded that these words must have been encoded in the text as a message from God.

The first scientific study of the bible codes appeared in the journal Statistical Science in 1994. The researchers took famous rabbinical names from a reference dictionary and used ELS to find those names near dates of their births or deaths. In his 1998 best-seller The Bible Code, Michael Drosnin claimed the codes predicted the assassination of Israeli Prime Minister Yitzhak Rabin and forecast a nuclear Armageddon in either 2000 or 2006.

Few seemed to notice Statistical Science’s cautionary remarks upon publication: the paper was being offered to readers as “a challenging puzzle.” A paper in the current issue of Statistical Science figured out the puzzle in the original research. McKay, et al found that they could replicate the original results using a Hebrew translation of War and Peace. It seems the method and data had been “tuned” to the test – causing the researchers to find just what they wanted. According to journal editor Robert E. Kass, now that anyone can “mine” data using widely available computer applications, “There are wonderful new possibilities for discovering misleading patterns.”

So who noticed this refutation? A news scan revealed only the Associated Press (Sep. 10), Science (Sep. 24), the Austin-American Statesman (Sep. 26) and Slate (Oct. 6) paying attention. Other outlets were too busy discussing the bible codes in a different context – that of a villain using them to rule the world in the film “The Omega Code.”


return to Howard Fienberg's page