[Previous part]   [Index page]

Mr Doron Witztum and the Chanukah cluster in War and Peace

This is the third and final part of the long version of the reply to the article "Did they really find codes in War and Peace" by Doron Witztum.

Text which is indented and italicised is quoted from Witztum's article, except that the Hebrew letters have been transliterated according to the Michigan-Clairmont scheme.


We show that none of the errors alleged by Witztum are errors at all, except for two typos of trivial significance. His abject failure demonstrates yet again the bankruptcy of his position.

Witztum's new soap opera

Neither Maya Bar-Hillel nor any of her associates ever claimed to find hidden codes in War and Peace except as tongue-in-cheek parodies of Mr Witztum's claims. Our position is precisely the reverse: that Mr Witztum did not find hidden codes in the Torah. The fact is that Mr Witztum has not and can not provide ANY evidence of a scientific nature.

It is very important to understand this non-symmetry between Mr Witztum and his critics. It is not a competition over whose examples are the most scientific, as the skeptics do not claim that examples like their Chanukah cluster are scientitic. The skeptics in fact claim this:

Mr Witztum pretends he can prove the critics are wrong just by finding fault with their demonstrations of data manipulation. Readers should not be fooled by this device.

The situation is quite similar to a famous scenario. Several professional magicians have demonstrated they can "bend spoons" just as convincingly as Uri Geller. Can Geller get off the hook just by proving that the magicians did not use supernatural powers? Of course not! He has to prove that he does use supernatural powers.

In the same way, the onus is on Mr Witztum to prove that he really does his experiments by honest rigorous scientific methods. He cannot do it, because he is not honest, nor rigorous, nor scientific.

Did Mr Witztum find any errors?

The fact is that we published a discussion paper months ago explaining how we made the picture and why the probability is not what it seems. The purpose was to demonstrate the dangers of taking probabilities at face value. The paper can be found here: http://cs.anu.edu.au/~bdm/dilugim/chanprob.html .

We remind the reader that Uri Geller cannot improve his case by proving that magicians bend spoons by trickery. They admit it. He has to prove he doesn't use trickery himself. Alas for poor Uri, he can't.

Alternatively http://cs.anu.edu.au/~bdm/dilugim/chanukah.html .

In his usual non-scientific way, Mr Witztum is trying to rule by fiat. Real scientists don't say "you have to do it my way"; that is how cult leaders behave.

In fact, there is no reason at all why one can't present a big picture with lots of words instead of a little picture with few words.

The history of the picture was as follows: We had about half a dozen pictures of 5-15 words each. They were large but not so large. When we considered what to publish, we found that several of them actually overlapped, so we joined them into a single picture. This did not show our examples in a good light, because it is not possible to place so many words in a pleasing configuration using a single row length. Nevertheless, it is obvious from the picture that one can extract a number of respectable small pictures from it. (Remember that the fragment can be redrawn with a better row length and the edges can then be cut off.) We won't bother demonstrating this, as a better example appears with our candles of Chanukah experiment.

We should also point out another difference between our picture and Mr Witztum's pictures. Whereas every one of our words is either an ELS minimal in 100% of the text or has skip +1 or -1, Mr Witztum commonly uses words that are minimal in only some fraction of the text, such as 80% or 50%, even as little as 34%. With such low standards we could have easily doubled, or more than doubled, the number of words in our picture.

Take, for example, Witztum's Chanukah "experiment" discussed in the second part of this article. The smallest rectangle that includes a minimum ELS of each of Mr Witztum's four words has 7536 letters, making it more than half the size of our big picture! In other words, by our standards his example is a dismal failure.

What follows is a totally ridiculous soap-opera that features Doron Witztum pretending to guess what is in our picture without actually looking. Why anyone would want to read his self-indulgent tripe is beyond me, but let's bravely dive in, just in case he says something amusing.

CB) YWN is perfectly valid. In fact, it is the form most consistent with Mr Witztum's "experiment" in [WRR2].

We used X$MN)Y, which though less common is quite valid. For example, it appears in the Haggadic Midrashim several times.

We used MWD(YT, which is a valid alternative name for the same place. It can be found, for example, right in the heading of the entry for Modiin in Encyclopedia Hebraica.

As Mr Witztum knows perfectly well, our picture shows the similar expression $MN ZK. This example is worth looking at closer, as it shows Witztum's entire case in microcosm. Opening his book at page 55, we find this:

Yes folks, Doron Witztum himself used the word he now wants to rule invalid.

If Mr Witztum would only look at the picture, all of these tedious additional investigations would be quite unnecessary. Let us also remind the reader that the trick of adding prepositions and articles to turn non-minimal ELSs into minimal ELSs was pioneered by Witztum himself.

"Praise" in English is both a noun and a verb, Mr Witztum. All you had to do was look at the picture and you would have seen the verb LHLL. However, we thank you for finding HHLL also present with minimal skip. We will add it with a sincere acknowledgement.

Really, Mr Witztum! How could you have missed those four pretty red rectangles all in a row?

I'm puzzled. How can XNWKH be merely the "alternative form", yet it is the only one used by the Mishna Brurah? (Please see the second part of this article, where Witztum cites the Mishna Brurah as a spelling authority for XNWKH.)

We used HXNWKYH and never claimed to use two forms.


NYSYM is used 27 times in the Jerusalem Talmud, compared to only twice for NSYM. It also has that spelling 2 of the 5 times it appears in the Mishna. Both those works date from before the introduction of diacritical marks (nikud). To see that using NYSYM is perfectly consistent with Witztum's practice, we can look in [WRR3] to find two words, GY+YN and QYDW$YN, spelt as they appear in the Babylonian Talmud despite explicit rules in the same paper requiring them to be written using "grammatical orthography". Here we see yet again that Witztum is perfectly willing to argue against himself.

(Don't tell anyone, but the word even appears once as NYSYM in Mr Witztum's spelling authority, the Mishna Brurah.)

Very good. We used "dreidel", and even named it in bold letters in our introductory text (proving that Witztum's "later discovered" is yet another lie). Is Yiddish less appropriate than the English, German and French words that Mr Witztum uses in his book?

Here are a couple of little puzzles for our readers. Each of these was seriously presented by Doron Witztum. Can you guess what they mean?
Click on the word to see the solution. After taking this test, you will appreciate how serious a sin we committed by using a word that is spoken on Chanukah by millions of Jews around the world.

Actually we made a slight error of judgement with "dreidel". Since Mr Witztum had previously led the way by writing Oppenheim with a single Y, we thought we should do the same with dreidel. On reflection, it would have been better to not sink to such depths. No need to worry though -- the minimal skip ELS of the correct spelling DRYYDL appears in our segment as well. We will change it.

Let us summarize first. Of all the errors claimed by Mr Witztum, not a single one is an error.

Mr Witztum is quite correct. We copied every single one of those devices from him. In fact, had we studied [Wit] more closely, we would have noticed an even wider variety of choice, for example the preposition L as in LM)WR (for light) and the preposition B as in BS)D)T (at Sadat). We could have tried lots more!

We also mention in passing that 'ktiv dikduki' is NOT the same as defective (vowel-reduced) spelling.

Witztum's "of course" notwithstanding, the statement is false. It is true only for short words. However, we remind the reader that months ago we published an article explaining why the probability isn't very small. Witztum is trying to trick you into thinking he can defeat us by proving our case.

What we have learned is that Mr Witztum is just making argument for the sake of argument. He has no interest in the truth, only in protecting his personal status as Codes Guru. Even if he has to argue against words he has used himself.

Mr Witztum is pontificating again. There is no reason I have to accept even the concept of "targets", nor do I have to accept his proclamation of what targets we have. After all, I am far more qualified to design scientific experiments than he is.

More importantly, Mr Witztum is being seriously disingenuous. Looking again at [Wit], we see the following "targets" used for the Chanukah pictures: HXNWKH (the Chanukah), L) YSWR (shall not depart), KH KSLW (25th of Kislev), MWDY(YN (Modiin), PK H$MN (oil pitcher). Some of those do not even have 100% minimal skip! Yet we are only permitted XNWKH and XNKH.

In fact we see in the appendix of [Wit] an "experiment" that uses about 60 distances, supposedly centered at HXNWKH, of which 43 word pairs do not even involve the word HXNWKH itself!

The statement above, taken literally, is totally false. Perhaps Witztum means that similar types of proximities can be found anywhere. If so, he is quite correct, both about our proximities and his. Proximities as good as any he has shown can be found in any sufficiently long text.

All of this is completely irrelevant, anyway. Since we always admitted that the real probability was not spectacular, it achieves nothing to show it again. What Mr Witztum has to prove is that his own examples are better.

If one looks at the large chart of Chanukah-related words in [Wit], one can see right away that many words central to Chanukah are missing. Even such important words as HMKBYM (the Maccabees). Of course, that has nothing to do with the fact that c(HXNWKH, HMKBYM)=57/125. He even sinks so low as to exclude some of the words that appear in the Chanukah chapter of his book, for example MWDY(YN (Modiin) and $MNT YMYM (8 days). The first one is stated to be "integrally related to the matters of Chanukah" (p53), but his "experiment" forgets it.

"Priests" and "king" didn't interest us as they are far too non-specific. For example, "king" appears many times but we didn't bother marking it. We marked "king-priest," which is much more relevant to the Chanukah story.

"Greek kings" and "High Priest" were set in bold type by accident. We congratulate Mr Witztum on at last showing some errors, even if they are only typos that appear neither in the picture index nor in the picture.

We also made a few more errors, and we wonder why Mr Witztum failed to find them. The very nice words YWNYM (Greeks), HLNYS+Y (Helenistic), and NRWT (candles) are all there with minimum skip and we missed them.

I corrected this error months ago, within hours of finding out about it. Witztum knows that. The phrase "pure vessel" was already marked in the text before the quotation and should not have been marked inside it.

Since Mr Witztum has never done a scrap of scientific analysis, his pompous pronouncements on the matter are worth nothing more than a bit of a giggle. Anyway, he is off target as always. All of our experiments with strong positive results are cooked, just as all of his experiments with strong positive results are cooked. The difference is that we admit it and he denies it.

Some discussion

The large Chanukah cluster was never intended as a serious experiment, and never presented by us as that. It is only a demonstration of what a large number of 100% minimal ELSs can look like together. The reader can look over it and see lots of interesting coincidences just as good as those shown in [Wit] or in the preprints of [WRR1].

In retrospect, it would have been better to show a number of much smaller pictures. With such a large number of words it is impossible to hide the freedom of choice that was used in selecting them.


We have enjoyed watching Doron Witztum's new soap-opera, "Watch the Guru make a fool of himself." Except for two typos of no significance, no actual errors were demonstrated. Instead, we just further confirmed that all the recipes we used to cook up the Chanukah picture were learned from the Master Chef himself.

Meanwhile, the number of perfectly valid words in our picture has increased from 59 to 63.


Astounding Discoveries in War and Peace, http://cs.anu.edu.au/~bdm/dilugim/chanukah.html (1997).
D. Witztum, HaMeimad HaNosaf (The Additional Dimension), self-published, 1979.
D. Witztum, E. Rips and Y. Rosenberg, Equidistant Letter Sequences in the Book of Genesis, Statistical Science, 9 (1994) 429-438.
ibid, Equidistant Letter Sequences in the Book of Genesis, II. The relation to the text, preprint ca 1995.
ibid, Hidden Codes in Equidistant Letter Sequences in the Book of Genesis, The Statistical Significance of the Phenomenon, Hebrew preprint accompanying a lecture given by E. Rips to the Israeli Academy of Sciences, 1996.

This page was written by Brendan McKay, who wishes to gratefully acknowledge the help and support he received from a considerable number of his friends.

Back to the index page
Back to the Torah Codes page