Debunking the Bible Code: Doron Witztum, Eliyahu Rips & Yoav Rosenberg

[KitBurger]

Does anyone remember the controversy when this article was published:

Doron Witztum, Eliyahu Rips, Yoav Rosenberg (1994). "Equidistant letter sequences in the Book of Genesis". Statistical Science 9: 429–438.

Basically it was about a bible code and modern people's names found in Genesis. Statistical Science presented it as a puzzle.

I remember sometime in the early 2000s that it was debunked and the person won an award from a sceptics society. Can anyone help me out with a reference for the article debunking the Doron etal one above and the award please?

Many thanks,
Chris

 

[Q_Goest]

Committee for Skeptical Inquiry: http://www.csicop.org/si/9711/bible-code.html

 

[Jimmy Snyder]

I remember sometime in the early 2000s that it was debunked and the person won an award from a sceptics society.

I haven't been able to find any mention of an award. It sounds a little fishy since debunking the bible code is a cottage industry on the net and there are too many people to give the award to.

 

[russ_watters]

The whole 'If someone can find an assassination prediction in Moby Dick...' challenge has to be one of the classic crackpot blunders of all time. Oops.

 

[Phrak]

Committee for Skeptical Inquiry: http://www.csicop.org/si/9711/bible-code.html

I don't understand the significance of this link. It's not intended to debunk Witztum, Rips and Rosenberg, but "The promoters of hidden-message claims..".

I'd read Drosnin's book in 1998. What's nice is that the full Statistical Science article is contained in the book appendectically. Witztum, Rips and Rosenberg claim that correlations of historically notable rabis together with dates associated with these rabis within ELS's are not due to chance.

If you had possession of an accurately encoded Torah used by Rips and all (There are about nine different versions, differing by one or two characters.), and the computer program that they used to evaluate lists of rabis you could run their tests yourself. They weren't releasing a copy to anyone.

If there is any debunking to be done, it is criticism of their statistical methods or the validity of their computer program.

I wrote approximately 10% of the required code in C++; enough to convince myself that producing code with a high confidence that it was error free, to the extent that it would not introduce systematic errors in favor of the thesis statement, was either beyond my ability, or more pertinently, that evaluating the confindence level in the resultant code was out reach.

 

[Q_Goest]

I don't understand the significance of this link. It's not intended to debunk Witztum, Rips and Rosenberg, but "The promoters of hidden-message claims..".

it's intended to debunk WRR.

Also, thought it might have something to do with the reference the OP was looking for regarding the "skeptics society". Not sure.

See also http://cs.anu.edu.au/~bdm/dilugim/StatSci/StatSci.pdf"

 

[Jimmy Snyder]

II wrote approximately 10% of the required code in C++; enough to convince myself that producing code with a high confidence that it was error free, to the extent that it would not introduce systematic errors in favor of the thesis statement, was either beyond my ability, or more pertinently, that evaluating the confindence level in the resultant code was out reach.

The software takes every nth character from a text. How hard can that be?

 

[LowlyPion]

The software takes every nth character from a text. How hard can that be?

Yeah but then you have to read it all.

 

[russ_watters]

No, you just run a search on the result. A more sophisticated algorithm would be better, though: it would pick different criteria and run multiple trials, then it would filter the results for real words vs nonsense.

 

[Jimmy Snyder]

Let's say I'm looking for the phrase "jimmy is great". First I would look for a 'j'. Then I would look for an 'i' and calculate the distance from the 'j' to the 'i'. Then I would add that distance to the 'i'. If that's an 'm', I keep on going until I either get my phrase, or I don't. If it's not an 'm', or I fail to get my phrase, then I go on to the next 'i'. If I exhaust all the 'i's without success, I look for the next 'j'. Why is this difficult?

 

[rootX]

Let's say I'm looking for the phrase "jimmy is great". First I would look for a 'j'. Then I would look for an 'i' and calculate the distance from the 'j' to the 'i'. Then I would add that distance to the 'i'. If that's an 'm', I keep on going until I either get my phrase, or I don't. If it's not an 'm', or I fail to get my phrase, then I go on to the next 'i'. If I exhaust all the 'i's without success, I look for the next 'j'. Why is this difficult?

Even I can do that But, I must be out of my head to spend time on these things. Why would someone work on this? It is just insane.

I don't know what is sophisticated algorithm. I think making complex relationships can prove almost anything.

 

[LowlyPion]

Let's say I'm looking for the phrase "jimmy is great". First I would look for a 'j'. Then I would look for an 'i' and calculate the distance from the 'j' to the 'i'. Then I would add that distance to the 'i'. If that's an 'm', I keep on going until I either get my phrase, or I don't. If it's not an 'm', or I fail to get my phrase, then I go on to the next 'i'. If I exhaust all the 'i's without success, I look for the next 'j'. Why is this difficult?

It may be easier than that actually.
All you need to do is find the "Jim is great" sequence and then start laying it out in a matrix of varying size so you can draw geometrically straight lines through the letters. Then you put the lines in a novel color.

Here's a Moby Dick link.
http://cs.anu.edu.au/~bdm/codes/moby.html

Rise and Fall of the Roman Empire might also be a suitable text for "Jimmy Rules" or even "LowlyPion likes Dr. Pepper" - my personal hidden text favorite.

 

[Phrak]

The software takes every nth character from a text. How hard can that be?

That's not the way it is. Taking every nth character is a small subroutine used in an involved statistical analysis--maybe a couple percent.

 

[Phrak]

From the various comments it's apparent that no one else here has actually read the paper under scrutiny or understands the pains taken to eliminate all false claims against the thesis statement such that the results would in any way be dismissible on statistical grounds.

But it is the pains taken that leads to the complexity of their code as becomes obvious to anyone attempting to duplicate it, where you should ask, is my code preforming as required? So I still ask, how can the code be evaluated for its accuracy? There is more concerning their test data sets, but the code comes first, once the validity of their statistical methods is found acceptable.

 

[Jimmy Snyder]

From the various comments it's apparent that no one else here has actually read the paper under scrutiny or understands the pains taken to eliminate all false claims against the thesis statement such that the results would in any way be dismissible on statistical grounds.

But it is the pains taken that leads to the complexity of their code as becomes obvious to anyone attempting to duplicate it, where you should ask, is my code preforming as required? So I still ask, how can the code be evaluated for its accuracy? There is more concerning their test data sets, but the code comes first, once the validity of their statistical methods is found acceptable.

If I find the phrase "jimmy is great" in a text with a gap of 127, how would you go about disproving that on statistical grounds. Either I found it or I didn't.

 

[LowlyPion]

If I find the phrase "jimmy is great" in a text with a gap of 127, how would you go about disproving that on statistical grounds. Either I found it or I didn't.

This is a bit analogous to Feynman finding the ARW 357 license plate isn't it?

But that aside, I'd have to observe that ultimately a large enough number of filters applied to a large enough collection of words - which are already preconditioned because they are found as words in the first place - can be expected to come up with yet more words - culled from a blizzard of possibilities - and can be further arranged through selection to form about any message.

An infinite number of monkeys and an infinite number of laptops should yield at least a post or 2 on PhysicsForum that look passably intelligent after all.

 

[Jimmy Snyder]

This is a bit analogous to Feynman finding the ARW 357 license plate isn't it?

No, it's like looking for the ARW 357 license plate and finding it. Look for my phrase in 1000 books all the size of the bible. How many books have it? Does the bible have it? How complicated is this?

 

[Phrak]

If I find the phrase "jimmy is great" in a text with a gap of 127, how would you go about disproving that on statistical grounds. Either I found it or I didn't.

You could if you wish, but this was not what was done by Rips et al. The intent of their software was to determine how much of a given data set is correlated beyond statistical chance within the book of Genesis.

The software takes two sets of data: notable rabis and the set of sets of dates associated with each rabi. The book of Genesis is searched for each rabi's name over a range of skip values. This is then done for the date sets. Now the skip is perturbated. For instance, for a skip value of 10 and an offset into Genesis of 300, given a word length of 4, the letter locations are 300, 310, 320 and 330. A particular perturbation might be 300, 310, 321, 330. Each location is perturbated in turn, up to a perturbation value of +/-3, where the base-skip would allow it (no zero skips). These perturbated and unperturbated letter position templates have both positive and negative skip values, i.e. the text is tested both forward and back.

Another operation obtains a proximity value for every occurance of a rabi with every occurance of a significante date of any rabi, no matter which rabi it belonged to, and over each perturbation. These proximity values essencially weight how close two words would visually appear within a matrix.

Weighting values are generated for each perturbed set.

There are more gyrations preformed that I don't recall. The same statistical test was preformed on various other texts, including a scrambled Genesis, an Exodus, and... more. The researchers reported no statistically significant correlations in these texts.

During peer review of the article, the Statistical Science referees asked for a second data set to eliminate the chance that the code was developed to favorably bias the results over the original data set. A second set of rabis was ground through the software, where some small set of rabis were added, and some removed.

I'm not particularly impressed by the large intersection of names occurring in both data sets, but understand that this was done because the first data set nearly exhausted the names available to them in the particular codex from which they acquired the first set.

 

[LowlyPion]

The thing to me that is compelling about Bible Code interpretation is the unknown authorship and amalgamated aspect of the early traditions in something like Genesis.

Not to offend those that would believe that authorship flows from Moses, but it seems to me anyway that whatever the case for it's origins, it seems to be a river of tradition, oral and written, that carries the tributaries of several, if not many confluences. And as such can hardly be expected to carry on its waters embedded messages, except as chance would allow and eyes of the future might conjure meaning, that would recount prospectively such things as a list of future Rabis or apocalyptic events, or even more mundane messages like "Lowly likes Dr. Pepper".

If there were a guiding hand to it all don't people think that there must be better things for that hand to be doing than doodling cryptograms in canonical text?

 

[Phrak]

You certainly have your opinions, LowlyPion; all of which I've shared at one time or another.

At times, though, for insane reasons of my own, in complete psychotic denyal of socially acceptable norms, objectively obtained opinion has its appeal.

BTW...

"If there were a guiding hand to it all don't people think that there must be better things for that hand to be doing than doodling cryptograms in canonical text? "

Telling God what to do?

 

[LowlyPion]

At times, though, for insane reasons of my own, in complete psychotic denyal of socially acceptable norms, objectively obtained opinion has its appeal.

Careful. The Devil may be in the details.

 

[Phrak]

it's intended to debunk WRR.

It's an attack on Drosnin in the main, by a David E. Thomas.

Thomas even goes so low as to attempt to discredit the original claim by attacking a quip made by Drosnin, as if discrediting one discredits the other. What a fraud.

Also, thought it might have something to do with the reference the OP was looking for regarding the "skeptics society". Not sure.

See also http://cs.anu.edu.au/~bdm/dilugim/StatSci/StatSci.pdf"

Yeah, this Statistical Science article by McKay may prove worth reading. Thanks for the link.

 

[CEL]

It is easier to find a word in the Hebrew edition of the Bible than in Moby Dick. Since Hebrew has no vowels, the same sequence of consonants can be read as several words.

 

[LowlyPion]

It is easier to find a word in the Hebrew edition of the Bible than in Moby Dick. Since Hebrew has no vowels, the same sequence of consonants can be read as several words.

Of course with fewer coding elements in the language, and the fewer redundant elements that might create greater distance between valid words in a language's coding space (meaning that valid words are separated one from the other by greater distance - additional characters or fewer or different characters), then the easier it is to find more valid words. That many character sequences that are valid words may encompass multiple valid meanings, further increases the statistical chance that more non nonsense groupings, or word sequences, will also be found.

An interesting experiment would be to look for Esperanto phrases in Moby Dick. Finding such sequences - in a language unknown to Herman Melville or any of his contemporaries - would surely demonstrate the power of chance to present meaning where no meaning could possibly have been intended.

 

[Jimmy Snyder]

The OP speaks of a debunking of WRR specifically. Apparently, there is a wide discrepancy between the claims made by WRR and those made by Drosnin. Given the fact that people like myself are only dimly aware of the claims made by either, it might have helped to make that clear in the OP. Having read up on it a bit, here is my take. Corrections to my misunderstandings are welcome.

In 1994, WRR claimed that encoded in equidistant letter sequences (ELS) in the book of Genesis, were the names, birthdates, and deathdates of 28 of 32 important rabbis. In 1997 Drosnin published the book "The Bible Code" which made the claim that details concerning future events were encoded in ELS (in Genesis? in the entire Old Testament? In both Testaments?) In 1999 McKay et. al. (MBBK) criticized the claims made by WRR, not those made by Drosnin.

As to the question in the OP. MBBK was published in 1999 by a journal devoted to statistical science (judging by its name: Statistical Science), not a society devoted to skepticism. It seems rather strange that a society of skeptics would then choose to present an award in the early 2000s to someone else. Perhaps some such society presented an award to McKay et. al. but I can find no mention of it on the web. Nor can I find mention of an award to someone else.

 

[LowlyPion]

It seems rather strange that a society of skeptics would then choose to present an award in the early 2000s to someone else. Perhaps some such society presented an award to McKay et. al. but I can find no mention of it on the web. Nor can I find mention of an award to someone else.

Eureka
http://www.skeptics.com.au/spoon/2003bsa.pdf

 

[Phrak]

The OP speaks of a debunking of WRR specifically. Apparently, there is a wide discrepancy between the claims made by WRR and those made by Drosnin. Given the fact that people like myself are only dimly aware of the claims made by either, it might have helped to make that clear in the OP. Having read up on it a bit, here is my take. Corrections to my misunderstandings are welcome.

In 1994, WRR claimed that encoded in equidistant letter sequences (ELS) in the book of Genesis, were the names, birthdates, and deathdates of 28 of 32 important rabbis. In 1997 Drosnin published the book "The Bible Code" which made the claim that details concerning future events were encoded in ELS (in Genesis? in the entire Old Testament? In both Testaments?)

I'm sitting here with the WRR's Statistical Science article in front of me. In conclusion they state: " We conclude that the proximity of ELS's with related meanings in the Book of Genesis is not due to chance."

In 1999 McKay et. al. (MBBK) criticized the claims made by WRR, not those made by Drosnin.

Was that a question? In any case it's correct. I've just now finished reading MBBK's article (minus the appendices). They specifically address the claims of WRR and to some extent, Gans.

As to the question in the OP. MBBK was published in 1999 by a journal devoted to statistical science (judging by its name: Statistical Science), not a society devoted to skepticism. It seems rather strange that a society of skeptics would then choose to present an award in the early 2000s to someone else. Perhaps some such society presented an award to McKay et. al. but I can find no mention of it on the web. Nor can I find mention of an award to someone else.

It hardly seems to matter that a society of skeptics, rather than trained mathematicians in staticticans would make an award on way or another. Do you think?

 

[Phrak]

In having read MBBK's criticism of WRR94, it seems very scholarly (and damning). Then again I'd thought WRR94 meaningful enough to make an attempt to falsify it, so what do I know?

If all the forgoing is not enough, here is Witztum's refutation of MBBK's refute,

http://www.torahcodes.co.il/debate1.htm"

Edit: And more from KcKay. Enter here: http://cs.anu.edu.au/~bdm/dilugim/torah.html"

 

[LowlyPion]

In having read MBBK's criticism of WRR94, it seems very scholarly (and damning). Then again I'd thought WRR94 meaningful enough to make an attempt to falsify it, so what do I know?

If all the forgoing is not enough, here is Witztum's refutation of MBBK's refute,

http://www.torahcodes.co.il/debate1.htm"

No surprises there. They have to try and scrape the egg off their faces.

And in the final analysis, I think anyway, that their choosing to carry the squabble to the arithmetic of it all rather than grappling with the inherent absurdity of trying to pass off chance as profound intent, is really the only rear guard action at their disposal.

 

[Phrak]

No surprises there. They have to try and scrape the egg off their faces.

And in the final analysis, I think anyway, that their choosing to carry the squabble to the arithmetic of it all rather than grappling with the inherent absurdity of trying to pass off chance as profound intent, is really the only rear guard action at their disposal.

Dude! Is this your objective opinion or a leap of faith?

 

[Jimmy Snyder]

It hardly seems to matter that a society of skeptics, rather than trained mathematicians in staticticans would make an award on way or another. Do you think?

That might have been made clear in the OP as well. I answered the question that was asked.

 

[LowlyPion]

Dude! Is this your objective opinion or a leap of faith?

You may be sure that it is not a leap of faith.

 

[Evo]

This thread is stretching the limits on overly speculative posts, so is closed.