Statistical Science publishes Bible Codes Refutation
The only paper published in a refereed scientific journal
that claims to find evidence for the reality of the Bible Codes is the
Letter Sequences in the Book of Genesis, by Doron Witztum, Eliyahu
Rips, and Yoav Rosenberg (WRR), Statistical Science, Vol. 9 (1994) 429-438.
We are now happy to announce that, after review by four
senior statisticians chosen by the journal, Statistical Science has published
a thorough rebuttal: Vol. 14 (1999) 150-173.
The new paper is
Solving the Bible Code Puzzle,
by Brendan McKay, Dror Bar-Natan, Maya Bar-Hillel,
and Gil Kalai. Here is the abstract:
A paper of Witztum, Rips and Rosenberg in this
journal in 1994 made the extraordinary claim that the Hebrew text of the
Book of Genesis encodes events which did not occur until millennia after
the text was written. In reply, we argue that Witztum, Rips and Rosenberg's
case is fatally defective, indeed that their result merely reflects on
the choices made in designing their experiment and collecting the data
for it. We present extensive evidence in support of that conclusion. We
also report on many new experiments of our own, all of which failed to
detect the alleged phenomenon.
This page presents the paper and some auxiliary resources.
Back to the Torah Codes page
The new paper
It can be downloaded in PDF format
or Postscript format.
If the link from Australia is slow, you can try Canada:
(1999), The Institute of Mathematical Statistics.
Limited copying for personal use permitted.]
To read the PDF version, you need the
The publication of the paper was accompanied by a press release from
the journal's parent society, the Institute for Mathematical Statistics.
It includes the introduction to the paper written by the Executive
Editor of Statistical Science at the time the paper of Witztum et al. was
published, Robert Kass. Read it here.
About the authors
A brief biography of the authors is here.
Due to space limitations, the paper refers to several
experiments without presenting the data for them. That data is presented
Due to the volatile nature of the web, pages and links
tend to change or disappear. Thus, we preserve here
many of the web documents referred to by the paper, in the form they were
in at the time of acceptance of the paper. We also provide scans of a few
less accessible documents.
A few typos have been found.
In this place we will reference serious critiques of our paper, and
sometimes comment on them. Readers should be aware that our paper is not
an opening gambit, but a summary of years of investigation that included
extensive debates with Witztum and Rips. We do not feel obliged to
reply to every recycled quibble, especially to those already answered
in our paper.
Witztum's "refuted refutation" concerns the appellations we found
that perform strongly in War and Peace. It is thoroughly refuted
In Dec 2000, Witztum published what he again called a "complete
refutation", but few of our points are even mentioned let alone
refuted. As a postscript to this sorry episode, we make a
few final remarks that include proof
that Witztum's own expert Shlomo Havlin does not himself adhere to
the "rules" that Witztum claims his data to be based on.
Doron Witztum has challenged our account of the history of the
permutation test used in WRR's paper. In reply, we provide a
detailed analysis, including scans of
the primary source documents. (September 1999)
Another paper of Witztum alleges that our experiment with Dr.
Simcha Emanuel was improperly conducted and the result in fact
supports codes! In our reply,
we catch Witztum with his fingers in the till. (July 2001)
Witztum has further claimed to refute our observations about his
choices of ways to write dates in Hebrew.
Our reply shows that his case is
nonexistent, as it is based on false procedures or incorrect data.
Witztum has replied to our analysis of variations,
arguing that it is either invalid or proves something different
from what we claim. Our reply shows
where his errors are while nevertheless questioning one of our
own previous conclusions. (Note: both Witztum's and our papers
on this subject require deep knowledge of the background and
some mathematical expertise.)
Witztum has written some comments on the issue of changes to the
text of the Torah. He appears to have abandoned his earlier
approach, which was to claim that the Koren text "can be relied on",
in favour of alternative ad hoc arguments. We are not going to
reply to them. However, for people who are interested in this topic,
we present some relevant articles by scholars.
Witztum found quite a few errors in Barry Simon's replication of
the experiment of Gans, but fixing them still leaves a completely
negative result. More seriously, two careful replications
attempted by a committee of supporters and skeptics, using data
compiled by independent experts, could not find the slightest
trace of codes. We have a preliminary
report available. (September 2002).
Apart from Witztum's efforts, the major reply to our work was
the "primer" of Harold Gans. In reply we will make a
Back to the Mathematical Miracles page