Review of Michael Drosnin. The Bible Code. New York: Simon & Schuster, 1997. 264 pp., with index. $29.00, hardback; also availabe in paperback.

The Bible Code

Reviewed by John A. Tvedtnes

The so-called Bible Code study conducted in Israel stimulates frequent questions. Several articles and a book have resulted from the study, which purports to demonstrate that hidden in the Hebrew text of the Pentateuch—the first five books of the Old Testament—are prophecies of future events. The researchers suggest that because only God can know the future, this is evidence that God inspired even the very wording in the Pentateuch.

Orthodox Jews and fundamentalist Christians alike have hailed the study as support for the divine authenticity of the Bible, which many of them hold to be inerrant and complete. They point out that most of the Bible Code researchers are statisticians, not theologians. Of late, some Latter-day Saints have been impressed with the code research, and it is their interest that prompts my examination of the methodology used and the problems involved.

The idea of a hidden code in the Bible was first introduced in a 1988 note published in the Journal of the Royal Statistical Society and written by three Israeli statisticians from the Jerusalem College of Technology and the Hebrew University.1 A follow-up article by the same authors appeared in 1994 in a refereed journal,2 Statistical Science.3 An editorial note by the journal's editor, Robert Kass, noted that "the paper is . . . offered to Statistical Science readers as a challenging puzzle."4 Several other scholars supported the study. Among these were five mathematicians (one from Yale University and two each from Harvard University and the Hebrew University in Jerusalem) who, in 1988, issued a public statement endorsing the work. Other supporters included researchers at Johns Hopkins University and the U.S. Department of Defense. A summary of the research appeared in 1995 in Bible Review.5

In 1997 the Bible Code was popularized in a book by journalist Michael Drosnin, who followed developments and interviewed the original researchers.6 The book reached a much wider audience and introduced the research to the layman. About the same time, the code concept came under fire from critics. Chief among these were two Bible scholars whose critiques were published in the August 1997 issue of Bible Review.7 I shall return to their comments later.

In order to evaluate the code study, one must first understand how it was conducted8 The researchers developed a computer program that takes the Hebrew text of Genesis, then skips over a specified number of letters, printing out, for example, every fifteenth letter. These are then placed in a matrix, usually a rectangular box, in which the letters forming the words are adjacent to each other on a horizontal plane. If, in the matrix, there are other words in close proximity, running horizontally, vertically, or diagonally (much like word searches in game books readily available in the United States), the connection is considered significant, provided the words "fit" together or make some kind of sense. The words formed by this method are, in and of themselves, unimportant since the laws of chance would produce valid words from time to time using the computer's methodology. But when the researchers find several related words within the same matrix, they feel they have demonstrated their point. One of their most important finds is the name Yitzhak Rabin (reading vertically) in close proximity to the words "assassin will assassinate" (reading horizontally). This, they suggest, is an ancient prophecy about the assassination of the well-known Israeli prime minister.

The 1994 study involved a search of the book of Genesis for the names of the thirty-four most prominent Jewish men from the ninth through nineteenth centuries, including standard abbreviations of famous rabbis on the list. They paired these with the dates of birth or death of these men, who lived long after Genesis was written. They claimed that each of the names was found in close proximity to an important date in the individual's life. They further claimed that the process did not work on various other books, such as the Samaritan version of the Pentateuch, biblical texts outside the Pentateuch, or Tolstoy's War and Peace.

The prophecies that can be found using this method vary according to how many letters are skipped. Thus, a lengthy Bible passage could, theoretically, produce more than one such message, depending on whether each tenth, eleventh, or twelfth letter (or sometimes every thousandth letter) is picked. The researchers claim that, statistically, the intersection of related words cannot be chance occurrences. Examples include the proximity of the name of the king of Judah, Zedekiah, with his real given name Mattaniah (see 2 Kings 24:17).

However, this methodology is fraught with several problems. One involves the nature of the Hebrew text. No definitive version of the books Genesis through Deuteronomy exists. Though a standard text is used in the synagogue, different ancient manuscripts vary in their readings. For example, among the Dead Sea Scrolls different versions of the book of Exodus vary widely. The omission or change of even a single word or letter can affect the results of the computer search. The code researchers have relied on a specific edition of the Masoretic Hebrew text (MT). Problematically, however, the MT was formulated in the centuries after Christ, and the earliest MT manuscript dates to the ninth century AD. The Dead Sea Scrolls, on the other hand, are a millennium older.

Another question deals with orthography. Some words have more than one possible spelling in the Bible and, in fact, are spelled differently in various manuscripts of the same passage. Originally, some Hebrew letters were used only to represent the semivowels y and w, as well as h, but were later used to also denote vowel sounds (i, o or u, and a). This led to misreadings in some later manuscripts that would also affect the results of a computer search.

But the coup de grace came when two Bible scholars examined the Bible Code in the pages of the August 1997 issue of Bible Review. Ronald S. Hendel of Southern Methodist University entitled his review "The Secret Code Hoax." Rabbi Shlomo Sternberg, who teaches mathematics at Harvard, called his article, "Snake Oil for Sale."

Examining the question of Yitzhak Rabin's assassination, Sternberg noted that "the computer found that if you skip every 4,772 letters, the name Yitzhak Rabin is embedded in the biblical text. In other words, there is a yod, the first letter of Yitzhak, followed 4,772 letters later by the second letter of his name, and so on. This means that if you print out the letters of the Hebrew Pentateuch (using the Koren edition) in rows 4,772 letters wide, the name Yitzhak Rabin will appear in a vertical column" (p. 24). To Sternberg, this stretched credulity too far.

Sternberg also took up the challenge launched by the principal Bible Code researcher, Michael Drosnin, in an article published in the 9 June 1997 issue of Newsweek, in which he said, "When my critics find a message about the assassination of a prime minister encrypted in Moby Dick I will believe them." Sternberg asked an Australian mathematics professor, Brendan McKay, "to search Moby Dick for such encrypted messages. He found 13 'predicted' assassinations of public figures, several of them prime ministers or presidents or their equivalents." Two examples appear in Sternberg's article. One has a message reading, "Pres—Somoza—dies—he was shot—gun," which one might connect to the 1980 assassination of former Nicaraguan president Anastasio Somoza. The other has "Igandhi" in a vertical line intersected by a horizontal line reading "thebloodydeed," which brings to mind the assassination of Indian prime minister Indira Gandhi in 1984. Using the same reasoning for Sternberg's study as that employed by the Bible Code researchers, we would have to conclude that God also dictated Moby Dick and that Herman Melville was a prophet! A Web site at cites the various "Assassinations Foretold in Moby Dick" and includes the Moby Dick prediction of the death of Princess Diana. But even Robert Louis Stevenson predicted the princess's death and that of President Kennedy in his Treasure Island, as can be seen by the examples posted at . . . /torah.

The truth is, however, that with enough permutations, one can find such "prophetic" messages in any lengthy text. A Greek Orthodox priest of my acquaintance, using the Bible Code software, found the name Joseph Smith (written in Hebrew characters) several times in the book of Genesis. I wonder how the many evangelical Christians who have bought into the Bible Code business would react to this kind of information—or to the fact that Orthodox Jewish researchers have found the names of many medieval rabbis, alongside their birthdates.

Not content to use the Bible Code software packaged with Drosnin's book, some enterprising soul has come up with software that can help search the Bible in Hebrew, Greek, and English.9 To buy into this idea, one must acknowledge that God built the original code in Hebrew but made provision for Greek and English translations in which the code would also work. But given the large number of Web sites on the Bible Code, it seems unnecessary to make such a purchase.10

Perhaps the strangest of these Web sites is run by a Latter-day Saint who believes that the real Bible Code is in gematria, in which each letter is assigned a numeric value (this late use of the Hebrew alphabet did not exist in Old Testament times).11 He and his followers take the numeric value of each letter in a word, add them together, then look up that number in Strong's Exhaustive Concordance of the [King James] Bible to see what word they find. (Evidently, they are oblivious to the fact that Strong could have arranged his words a bit differently and still been within his own parameters.)

But his method goes beyond just the Bible. This group also uses telephone numbers, street addresses, zip codes, and even Social Security numbers to make predictions. Using this method, the owner of the site claims that his own marriage, including the size of the ring he bought for his bride and its price (plus his own Social Security number and the house he bought just prior to his marriage) have all been predicted. He has a number of followers on his message board, and it is clear from the traffic there that these people believe that anything and everything has not only been foreseen by God but that God has made use of the telephone directory and virtually every book ever published to encode messages for us to decipher. This being the case, all of the "Moby Dick" arguments seem to fall on deaf ears. I find it particularly ironic that the ideas held by this group of Latter-day Saints smack of predestination, which conflicts with the church's concept of the agency of man.

Meanwhile, the May 1999 issue of Statistical Science, the journal that published the original study five years earlier, included a refutation by four statisticians, who wrote that "despite a considerable amount of effort, we have been unable to detect the codes."12 The issue is still being debated in print and on a number of Web pages. One of the more impressive of these is a denunciation of the Bible Code by a growing list of mathematicians from around the world, some of whom accept the Torah (Genesis through Exodus) as inspired scripture but reject the concept of a hidden code.13

My recommendation to Latter-day Saints is to stick to what the prophets—ancient and modern—have told us about the future and not rely on this latest superstitious fad.


  1. Doron Witztum, Eliyahu Rips, and Yoav Rosenberg, response to D. J. Bartholomew, "Probability, Statistics, and Theology," Journal of the Royal Statistical Society, series A, 151/1 (1988): 173-74.
  2. Refereed journals ask one or more scholars in the relevant field to review or referee an article prior to publication to ensure that it meets scholarly standards.
  3. Doron Witztum, Eliyahu Rips, and Yoav Rosenberg, "Equidistant Letter Sequences in the Book of Genesis," Statistical Science 9/3 (1994): 429-38.
  4. Ibid., 306.
  5. Jeffrey B. Satinover, "Divine Authorship? Computer Reveals Startling Word Patterns," Bible Review 11/5 (1995): 28-31, 44-45.
  6. Michael Drosnin, The Bible Code (New York: Simon & Schuster, 1997).
  7. The articles were published together with an editorial preface entitled "The Bible Code: Cracked and Crumbling," Bible Review 13/4 (1997): 22. The articles are Ronald S. Hendel, "The Secret Code Hoax," 23-24; and Shlomo Sternberg, "Snake Oil for Sale," 24-25.
  8. The term equidistant letter sequences or ELS was first noted in 1958 by a Rabbi Weissmandel before computers were available to most people. The Israeli statisticians acknowledge their indebtedness to Rabbi Weissmandel, whose observations prompted them to investigate the phenomenon using a computer program.
  9. One can order the software online at Web sites given in this review were valid as of October 2002.
  10. It is not feasible to insert here all the Web sites on the subject. One of the more prominent is
  11. See
  12. For the full article with an impressive bibliography, see Brendan McKay, Dror Bar-Natan, Maya Bar-Hillel, and Gil Kalai, "Solving the Bible Code Puzzle," Statistical Science 14/2 (May 1999): 150-73.
  13. See