Did we really find codes in War and Peace?
by
Brendan McKay
with much help from his misguided native companions
A series of articles by Doron Witztum has recently appeared at www.torahcodes.co.il, a web-site paid for by Witztum's chief sponsor, supporter and promoter, Aish HaTorah, which also maintains several sites elsewhere that pretend to watch objectively from a distance.
In this document we show how two of those articles provide us with the answers to these questions:
As we shall amply demonstrate, the answers to the three questions are no, no, and no.
We will also present three new experiments of our own.
The first of Witztum's articles to which we will refer is entitled "Did They Really Find Codes in War and Peace?". It was made available as a PDF file, but as a convenience to our readers we also make it available in Michigan-Clairmont encoding. That article is replied to in three parts below: A, B, and C.
The second article to which we will refer is entitled "Does Tolstoy really love Brendan McKay?". It is also available both as a PDF file and in Michigan-Clairmont encoding. That article is replied to in section D.
The definition of Michigan-Clairmont encoding can be found at cs.anu.edu.au/~bdm/dilugim/MC.html.
Summary: Examination of Doron Witztum's work uncovers no evidence that his work is conducted according to scientific principles. The only evaluation matching the evidence is that he is engaged in a weak parody of science.
Summary: We present a detailed explanation of how Doron Witztum constructs "experiments" using as an example his demonstration of a "header sample" for the Holiday of Chanukah. Contrary to his claim to be presenting an objective experiment, we prove that he is deliberately selecting a few words from a large pool of data he has previously analysed. As if that wasn't bad enough, he totally ignores his own rules of spelling and analysis. This example all by itself is enough proof that Doron Witztum does not deserve to be taken seriously.
Summary: Doron Witztum claims to have found a long list of errors in our large cluster of Chanukah-related words in War and Peace. We show that he is wrong in every single case. He even tries to disprove a word he used himself! Then he presents a proof that the probability was not really very small, completely ignoring the fact that we had already presented a more accurate proof of the same fact.
The truth about our Chanukah cluster is something that Doron Witztum knows very well. We cooked it up by making use of the many arbitrary choices available in data selection. Every single one of our tricks was copied from Doron Witztum's own work. In any case, we never presented the cluster as a serious experiment. It was, and remains, a humorous send-up of the ridiculous word clusters of Doron Witztum.
Summary: Some months ago, I presented an "experiment" that appeared to show a remarkable convergence of my name and date of birth in the Hebrew translation of War and Peace. The purpose of the experiment was to demonstrate that Doron Witztum is not the only one who is capable of exploiting non-obvious freedoms to fake experiments. Naturally Witztum is the person most qualified to find the problems, but he can't expose them honestly because he uses the same tricks himself. Rather, his "refutation" is a long list of deliberate distortions.
Of special interest is some little-known information about the history of the famous rabbis experiment.
We only recently started looking at the "header" sample sets of Doron Witztum. We think they are pretty cute, and would make good examples to teach students of statistics how to recognise fakes.
If Mr. Witztum presented his "experiments" as jokes, we would congratulate him on his skills. Alas, he spoils them by presenting them as "science".
A brief study of Witztum's examples revealed the secret behind them, and now we can make such examples too. Please enjoy our work.
On June 16, 1997, on Israeli television, Doron Witztum presented a map listing the sub-camps of Auschwitz. He announced that the names of the sub-camps appeared remarkably close to the phrase "in Auschwitz". "One in a million", he claimed.
This example has been presented at least twice in public as an example of a "totally objective" experiment that produces a remarkable result.
Unaware of any other details, we decided to carry out the experiment ourselves. We used exactly the same map with exactly the same spelling, and applied exactly the same method that Witztum and Rips previously defined for analysing such data sets.
Of course we expected it wouldn't work, or not very well, because Doron Witztum has never ever presented a genuinely objective experiment that worked. Imagine our surprise in finding that Witztum's claim of "one in a million" was only wrong by a factor of 289149! We invite our readers to consider how this tiny discrepancy might have arisen.
Please see An objective experiment of Doron Witztum.
Back to the Torah Codes
page
Back to the Mathematical Miracles page