Analysis of Bible Codes: Language Tied to Statistics

Biblomancy is fortune telling by opening pages of sacred books to discover a hidden message. Practiced by several religions, this process is the inspiration of many theologists, mathematicians, and scientists. Specifically, the study of possible Bible Codes has grasped the attentions of the Jewish scholars Doron Witztum, Eliyahu Rips, and Yoav Rosenberg. WRR, in their study conclusions, propose a proximity of thirty-two medieval sages and their birth or death dates that are too close to be accidental. All thirty-two were born after the Genesis was written, thus clearing the Code from being designed by its authors’ deliberate steps [1]. Their work is known as the Famous Sage experiment.

Countering this study is a group of erudite mathematicians and professors who duplicate WRR’s experiment and show it is not as accurate as they claim. These key individuals Brendan McKay, Dror Bar-Natan, Maya Bar-Hillel, and Gil Kalai, who will be referred to as MBBK, use statistical data to disprove WRR, and back it up with reasoning that includes theological logic as well as scientific logic.

The Hebrew text of the Torah has been said to contain profound patterns and messages that allude to real-life events and persons of historical significance. Researchers base their data upon equidistant letter sequences, or ELS. They find letters that are close to each other, spaced at the same distance apart, which form messages of religious import. These messages are not only of ancient times, but also of today. Among the events are the Holocaust, the 1991 Gulf War, and bombings in Jerusalem. [2]

The following table, Table 1, is included in the paper by MBBK: [3]

Date form	Used by WRR?	List 1	List 2
D M	yes	0.165751	0.000017
bD M	yes	0.000008	0.008844
D bM	yes	0.006070	0.008804
bD bM	no	0.068478	0.429256
D `M	no	0.581777	0.274167
bD `M	no	0.281509	0.618128
D shel M	no	0.711538	0.046468
bD shel M	no	0.467761	0.135884
Margaliot	partly	0.070780	0.277658

 

The above are the outcomes if each date form were used by itself by the key: D is the day of the month, M is the month, b is the pre_x \in," ` is the pre_x \of" and shel is the word \of". The final row incorporates these dates into the outcome they would have, had they been used in WRR’s permutations. The conclusion made by MBBK is that WRR’s lists work only for the three types of format they used. [4]

There are several variations of Biblical text itself, though WRR asserts there is one accurate version, which they used in their study. However, fact opposes this claim. According to MBBK, studies of the period 200 B.C.E. to 100 C.E. of Genesis manuscripts found at Qumran are different by hundreds of letters from 3^rd century C.E. Masoretic Text. The MT, developed through the diligent work of medieval scholars, is the accepted accurate text. In promoting their study, WRR must be able to have every letter of the text they use, the 1962 Koren, be perfectly accurate. Unfortunately, this is impossible.[5]

MBBK did a study of 999,999 alternative control lists, or permutations. Each of the thirty-two WRR rabbinical sages was paired randomly with the birth or death date of a different sage. The outcome: the correct list was placed fourth. Therefore, if God ordained that a code be incorporated into the written Word, He deemed that three wrong lists should be placed in closer proximity. Still, the results are astounding and thought provoking in terms of the relative significance of the permutations.

MBBK did another study on the Koren in comparison to other texts. The variations in editions shows that basing codes upon texts that depend on ESL is not a sound statistical argument:[6]              

	Differences	Rank
Koren	0	6
Yemenite	3	19
Sassoon	11	2308
Venice Mikra’ot Gedolot	15	16608
Leningrad	22	12528
Jerusalem	35	12528
Hilleli	43	6411

 

The results speak for themselves, and place severe limitations on the argument so promoted by WRR.  The Difference column contains spelling discrepancies.  The values inserted in the Rank column are based on the permutations of 10 million random dates.  Notably, MBBK sought a greater degree of accuracy than WRR did in conducting the same experiment; WRR used merely 1 million permutations.

     Benjamin Wittes makes a very valid point: all long texts will have accidental ELSs, which can be found if someone really wants to find them.[1]  The significance of the Bible Code experiments is not discounted. Textual analysis is a worthy pursuit because it analyses the cohesiveness between words and meaning, beyond the obvious form. However, if WRR is to assert their claims as facts, they need more resounding data to back up their study. MBBK did a more thorough study and exposed the legitimate flaws in WRR's argument.