Two ELSs meeting in a text are like two people crossing paths on a street. Suppose you happen to see your old friend George Aiken on the street, after 20 years. This is interesting, but something similar happens to a great many people sometime in their life. So the odds are insignificant, perhaps 1 in 2 or 3. This is very much the kind of insignificance I just referred to.

Now let’s change it slightly - but really drastically - by adding a protocol. Your protocol is that you wake up that morning and announce that you think you will run into George Aiken today. Some hours later, it happens. Now the significance is astronomical.

So everything depends on protocol. In many codes conversations, for example in my exchanges with Dr. Klafter, we are really just discussing protocol (I have more to add to that conversation, which may end up here, or on http://www.torahcodes.net). The Torah Codes protocols are trending toward increasing transparency, allowing simpler verification by any observer.

1. If the codes are true, then what do you do about the Christian codes that have proofs for Jesus?

2. How do you explain Dr. Simon finding codes in a hebrew version of War & Peace? Was Tolstoy devinely inspired to write that book?

E. Rips, and A. Levitt, The Twin Towers Cluster in Torah Codes, 18th International Conference on Pattern Recognition, 2006, vol.3, 408-411, see also
http://www.torahcodes.net/rips0.htm

At the location http://www.torahcodes.net/visual_20070926.doc the reader will find a short eplanatory note including a link to a PP presentation that visualizes the Torah Codes effect studied in this paper.

Dr. Klafter wrote: “[The Torah] has now been subject to hudreds of thousands or perhaps millions of searches for hidden codes, whereas no one is devoting equal effort to search for hidden codes in [other texts]”

This is another very well-taken point, though I believe it is a large over-estimate. (I personally spend a great deal of time developing techniques and ideas, and a small fraction of my time doing actual searches). At any rate, the point is addressed by the idea of restricting our examples to high-relevance topics. I mentioned this in posting 28, in regards to the idea of “hidden failures”, but it relates to the current idea of “effort applied to the Torah” as well.

By limiting to high-relevance topics, there are a limited number of situation 1’s (initial bricks) that can be set up. For example, the Twin Towers attack (in a recent poll the most striking news event in the lives of a majority of people surveyed), was the first topic for which we identified codes that are built from other codes. That formal study,

http://www.torahcodes.net/belj.pdf

was the prototype for the even simpler codes built from codes that we have found recently, in several other high-relevance areas. For the Twin Towers, we accepted only certain “foundational” codes to be used as our starting bricks - such as the two codes containing interesting extensions of a “bin Laden” ELS (forming long phrases). In virtually all cases of codes built from codes, we only deal with very strong and compelling starting bricks.

Here there is a potential communication problem in this discussion, because the reasons that a code is compelling are sometimes difficult for more experienced code searchers to convey to typical observers. In the analogy of meeting Katie on the street, a person experienced with books knows that the coincidences on pages 124 and 125 are highly unusual and compelling. A person who has only encountered a few books in his life may not have that insight.

Dr. Klafter suggests a simplification of the original Rabbis experiment (WRR). But in fact all of the research that has followed WRR has been a simplification and/or a reinforcement of WRR. The whole methodology of codes built upon other codes is only one example; Harold Gans’ work was a precursor and partly an inspiration for this methodolgy, as it was an “experiment built upon (the WRR) experiment”.

Among several other types of experiments not even mentioned yet on this blog, codes that echo Torah verses provide more examples of simplification. A systematic study of verse echoing, by Nachum Bombach, is described here:

http://www.torahcodes.net/bombach0.htm

And some newer, also simple, and individually striking, examples of verse echoing are illustrated on these pages: http://www.torahcodes.net/new/pages/create_world.html, http://www.torahcodes.net/new/pages/humanity.html, http://www.torahcodes.net/new/pages/create_light.html (top two figures).

Regarding Dr. Klafter’s suggestion of comparing the Torah simultaneously to other texts, we have been following this approach from the beginning, to derive our probability measurements: part and parcel of the Torah result is the rank of that result among a competing population - as advocated and advanced for years now by Professor Haralick. This population can consist of real texts when this technique is applicable, and/or randomly permuted texts, and/or randomly shifted locations of the key words’ ELSs, for example. In most cases, especially when we are dealing with individual code tables, these various populations yield results that are highly consistent with each other.

I increasingly favor these single, strong and visible code tables (as opposed to summations of many milder ELS meetings spread throughout the text). For this kind of single table, it is often possible for an experienced searcher’s intuition (and/or high school mathematics) to approximate and thereby reinforce the above results of computer-driven comparisons. If such simple tables are built from other codes, or if they are built from verses from the Torah, the task of agreeing on key words is also greatly simplified.

This posting and my next one give some of those answers.

Dr. Klafter writes: “You cannot statistically analyze your ‘finds’ after the fact. I will not find such pheonomena to be interesting unless it can be clearly demonstrated that no other Hebrew texts generate similar results.”

I have two different reactions to this, depending on the situation. Situation 1: Starting codes (first bricks) such as Fig 1 of the Temple codes. My reaction in this situation is that one can do a limited analysis to guage whether the code is noteworthy, and one can even go beyond that in some cases. But extreme care is required, and I certainly understand (and in many cases I would even strongly agree) with Dr. Klafter’s comment when applied to situation 1.

Situation 2: A code that is an outgrowth of former codes (bricks 2 plus). My reaction in this case is that we are completely out of the class of problems that can be labelled as “after the fact”, so that Dr. Klafter’s comment does not apply. Figs 2-5 of the Temple codes are situation 2. These codes really need to be addressed, in order to have a real picture of what http://www.torahcodes.net is trying to say.

For now, let’s look at the simpler case, the codes about codes - let’s look at the bottom picture on http://www.torahcodes.net/new/pages/codes_on_codes.html

The code on the left is situation 1 and the code on the right is an outgrowth of it, i.e. situation 2. The code on the left is simply “Code of Hashem” in the same column as “signs” (with the same spelling, and perhaps the same meaning, as this word in Isaiah 41:23, 44:7 and 45:11). On its own, the code on the left is actually interesting to me. It is like meeting Katie on the street, as in Dr. Klafter’s former example, but in my version of the analogy there are additional circumstances:

Bill is carrying his favorite book, called “The Hints of God”. He meets someone new named Katie, who is carrying her favorite book, called “Divine Signs”. Both books are rare editions, out of print, and both books have page 124 highlighted in green and are missing page 125. This is all situation 1, yet most people understand that it is interesting because it contains some very unusual coincidences that do not happen every day.

Now we move to situation 2. The next day, Jack, a friend of Bill who is reading a book with a slightly different title, “The Secrets of God”, takes a walk and meets someone new named Catherine. When she tells him her name, he says, “this is a funny coincidence” and relates the story of Bill and Katie. Wide-eyed, she pulls out of her book bag a different copy of the same rare book that Katie had, “Divine Signs”. The fact that Katie and Catherine have similar names adds to the incredibe series of coincidences on day 2, but the similar names served only one purpose - to trigger in Jack’s mind the story that Bill told him yesterday - i.e. the impetus to search for further extraordinary details. Situation 2’s main event (revealing the book hidden in Catherine’s book bag) is absolutely “a-priori” (i.e. the event is so surprising because it was suggested by the pre-established events of day 1). One can therefore determine how often a random individual carries a copy of this particular book in order to quantify the probability of day 2’s main event. This kind of quantification is not valid on day 1 for Katie (situation 1) but it is valid on day 2 for Catherine (situation 2).

Returning to the web page, one might object that the picture on the right uses a different synonym for “code” than the picture on the left. But that is no different than Jack having a different titled book than Bill. They were sufficiently similar to each other so that the natural a-priori question for Catherine is whether she by any wild coincidence was carrying a book called “Divine Signs”. This chain of events is essentially the same as the events leading to discovery of the web page’s code on the right. The natural question for the code on the right, after finding “code of Hashem” in that picture, was whether by some wild coincidence we would see “signs” in the same column.

Dr. Klafter says:
“However, I WOULD be interested to see if you can repeatedly demonstrate that such ELS phenomena are restricted to the Torah.”

What happens when we ask 1,000,000 other people (texts) the same question we had for Catherine (she is the Torah text in our analogy). We find out that only 10 of these people are carrying the same title with them. Again the analogy is very fitting. Whenever we identify a code that fits situation 2 (an outgrowth of a previous, *noteworthy* code) we do in fact carry out a comparison with thousands of other texts in order to determine the probability of that outgrowth code appearing in the Torah.

I think that we cannot draw any conclusion from Torah Codes regarding whether we currently have: (1) a letter for letter record of what was originally received by Moses, or (2) less than perfect accuracy.

The existence of Torah Codes in the text of the Torah that we currently possess is logically consistent with both possibilities.

The consistency with the first possibility is obvious. So I will explain how the existence of Torah Codes is logically consistent with the second possibility.

[Notice that I am not even discussing in this post the plausibility of Torah Codes or the merits of the evidence. Instead, the argument is purely logical.]

The existence of deliberately encoded Torah Codes with details of events of the distant future implies advance knowledge of this future.

If someone truly had the ability to see these details, this could certainly include details of what will happen in the future with the received text itself.

Given this advance knowledge, what makes more sense - to insert codes in a text that will eventually change and therefore see the codes be degraded or destroyed (fully or partially, depending of the scale of the changes), or to produce the text in such a way that the encoding will appear in the modified text that will be available in this distant future when the codes are discovered?

What would YOU prefer is a similar situation? Obviously, the second alternative.

People often have diffculties when performing logical operations with hypothetical ideas, especially if they consider the contents of those ideas to be highly implausible. In reality, the rules of logic allow such reasoning.

For example, as is well known from classical logic, the validity of the *implication* A => B does not depend on the validity of A. The *implication* is not valid only if it can be shown that A’s hypothetical validity would invalidate B.

Therefore the validity of the above reasoning should be acceptable to one who allows or denies the possibility of Torah Codes, as well as to one who has (any) particular opinion about the accuracy of the current text.

As a result, Torah Codes are logically consistent with either a changed or a constant text, and we can draw no conclusions from their existence as to whether or not we now possess a letter by letter record of the original received Torah.

The fact that some critics of these codes object to religion in general says nothing about the codes themselves.

The textual argument (and evidence) brought by Prof Hillel is brought as well by individuals who believe in Torah min HaShamayim and are knowledgeable about the Massorah.

Simply saying that you don’t like (or trust) one of its messengers will not make this argument go away.

After many years of following the Torah Codes controversy, I am still waiting to hear a substantive response to the textual problem.

(I brought it up as well (in comment #44)).

In her essay, “Madness in the Method,” she writes:

“Creative magicians love to amaze and baffle their peers with a new trick they’ve invented, challenging them to figure out the secret of the trick on their own. An important psychological element in this task is strict adherence to what you know with certainty to be impossible. Suppose you see a magician cutting the pretty lady in two, separating her smiling face, upper torso and all, from her twitching toes, long-booted legs and all, and then putting them together again. It is wrong to ask: ‘How can the lady be cut in half and glued back together again?’, because the obvious answer to that question is: She cannot! But you are on your way if you ask: ‘How can one appear to be cutting the lady in half and then putting her together again?’”.

“The codes research raises the obvious question: ‘How could the names of rabbis and their dates of death have been encoded in a text written millenia before the rabbis were even born?’ For the skeptic, this is the wrong question. Compelling evidence that such a code truly exists is the stuff of The X-Files. A committed skeptic does not ask how the Codes got into the Torah, but rather how did they appear to get there. Interestingly, this should be the question even for those who literally believe that the Torah was dictated by God letter for letter on Mt. Sinai, because there is little disagreement among biblical scholars that the many different texts of the Torah that we have today differ much more from the original text of antiquity than they do from each other — which is by quite a lot. Witztum and Rips ignore the problem of conducting research on an artificial publisher’s version called the Koren edition, which is known to be hundreds if not thousands of letters different than the original of antiquity. Drosnin falsely claims that ‘every Hebrew Bible that now exists is the same letter for letter’, and that the text used for the codes research ‘has not changed in at least a thousand years’ (p. 38). In any case, it had two thousand years to change before this millenium. And it did.”

For te codes people which sifrei Torah do they use=Ashkenaz or Sfarad-different in aleph or heh in psulaei dakah-or Yeminite which I believe has 7 differences in malei and chaser-no meaning difference-but crucial for Biblical Code skipped differences.

For the first and second Great Rabbis experiments, the P ranks (which represents the probability that the results occurred by chance) consistently increased when MBBK used alternate Torah texts (such as the Yemenite Torah text) instead of the Koren Torah text. However, the P ranks still remained quite low for each of these experiments. For more details, see page 29 of MBBK’s paper:

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

I was pleased to see your exchange with Dr. Klafter.

I invite you to respond as well to the points I raised in comment #44.

Kol Tuv,
Dovid Shlomo

Thank you for your lengthy and thoughtful response.

You are attempting to demonstrate that the ELS phenomena you observe are “unexpected”, but I stand by my rejection this methodology entirely and do not believe that there are any mathematics which can be meaningfully applied to it. You cannot statistically analyze your “finds” after the fact. I will not find such pheonomena to be interesting unless it can be clearly demonstrated that no other Hebrew texts generate similar results.

However, I WOULD be interested to see if you can repeatedly demonstrate that such ELS phenomena are restricted to the Torah.

In other words, I am stuck with the following null hypothesis–that the ELS codes are entirely insignificant and will occur similarly any other Hebrew text of the same length. It is just that we are much more interested in the Chumash. It has now been subject to hudreds of thousands or pehrpaps millions of searches for hidden codes, whereas no one is devoting equal effort to search for hidden codes in a Hebrew translation of the Physicians Desk Reference, or one year of Ha-Aretz issues, or random sets of 304,805 Hebrew characters.

Again, I would be impressed by an experiment which is run simultaneously on many Hebrew texts.

A long long list of outstanding Rabbinic figures with their yartzheits AND with pre-determined spellings would be fine by me. Again, I suggest all gemara commentators included in the Otzar Meforashei Ha-Talmud, with the precise spellings used in publication, or which are used in Otzar Roshei Teivos. The point is that the list MUST be objectively defined in advance and free of any spelling alterations or any other manipulations in response to prior search attempts in the Chumash. Analyze the Torah and 99 other Hebrew text fragments of equal length with the same list and according to identical search procedures. We will see if the Torah yields results which are significantly different from the other texts. THAT would be a meaningful experiment.

I wonder what Barry Simons and Howard Gans make of my suggestion for this experiment. Would it be interesting to you? Would this be meaningful? If not, why? How many names would be required on the list in order to generate enough statistical power?

I believe that this dialog is very useful. There are many more issues than meet the eye, but they can all be clarified, given enough time. Perhaps at least a few issues can be clarified in the current discussion.

*Part* *I* *of* *II*

We have been discussing the Temple codes on http://www.torahcodes.net/new/pages/rebuilt.html . Up to now, we have been discussing Fig 1. But really, Fig 1 can not be viewed in isolation - its relationship to Figs 2-5 is the crux of the matter, and here is why:

First and foremost, the issue of interest to critics and proponents alike is this: How can the choices available to an experimenter be limited, in a way that the observer can verify? Typically, the method used is to extract key words from a well-known source, such as a list in an encyclopedia.

But there is now a new, additional technique illustrated on the web site. It uses codes themselves as the source for other codes. That is, Fig 1 (considered to be like a foundation, or first “brick”) uses a certain set of key words. These key words are then themselves used as a source for the codes of Figs 2-5. All of these figures together act like five inter-locking bricks in an overall *structure* of codes.

There are other examples on the web site that demonstrate this kind of structure more directly. For example,

http://www.torahcodes.net/new/pages/codes_on_codes.html

shows at the bottom two codes with key words *code* *of* *God* *(Hashem)* and *signs*, using two synonyms for *code*. I hope later to describe in detail how the code on the right (brick 2) arose from the code on the left (brick 1). [ To simplify life for the readers (and for this writer), I need to "pace" my postings; perhaps some of this will be added to the site rather than to this blog ]

After the above case is clarified, it will be much simpler to further discuss the Temple codes, which are based on the same principles.

We are not trying to assign a formal probability to the first brick at this point. So our disagreements about key word choices for that first brick can be noted (see Part II), but that discussion is secondary to the main issue just described. The real point is the structure created by all of the codes together (Figs 1-5).

To expand slightly on the new technique: what does it mean that a key word found in one code is used as a source for additional codes? It means that all of the most basic alternatives we might suggest for that key word - why not this particular synonym or this particluar spelling - are now used as alternatives when looking for additional codes. If those additional key words fail to find interesting codes, it shows that the possibilities for the original idea have been exhausted, and the significance of the original code should be reduced to account for those failures. On the other hand, if those additional key words succeed, it shows that the original code was not created after a string of failures. Just the opposite: the original code was the first of potentially many more hidden successes.

*Part* *II*

This fills in a few points about Fig 1, to answer Dr. Klafter. To clear one misunderstanding, Fig 1 is at least 10 years old, and the original method for finding it is not known. Nonetheless, it is possible to glean enough from it to recognize that it is a showcase example.

My reference to “Hilchot Melachim”, Chapter 11, was made only in order to emphasize that the concept of the Messiah is intimately connected to the concept of the rebuilding of the Temple. It is this event of rebuilding, anticpiated for centuries, that is reflected in the code.

We can analyze pairs of words in the code to show that it in fact displays some non-trivial effects. One of those pairs is *Temple* and *on Moriah* (the Temple and its location).

When I indicated 1:80,000 for the proximity of *Temple* to *on Moriah*, I explicitly labelled that as a raw number. I said that this number “must be corrected by the number of equally interesting alternatives. … We quickly realize that the exact correction required … can not be easily pinned down.” Where you might say it is impossible to pin down, I would say that an approximate set of reasonable alternatives can be compiled for the word *Temple* and a second set can be compiled for the Temple’s location, *Moriah*.

In fact, if we could get a consensus of many opinions, then among the most simple, direct and popular words for Temple we would cetainly have “mikdash”, and likewise “Moriah” for the location (some of this can be verified with statistics from Google-in-Hebrew searches). Even if there are 10 better choices for each, the 80,000 would be reduced by a factor of 10 x 10, which would still give a probability of 1:800 for this word pair appearing in such close proximity.

By the way, we do agree at least on this: the phrase “this will comfort us” is not statistically interesting. To repeat from my earlier post:

“Point 5: We often highlight a related phrase near a code just as a point of interest – for example, the phrase “this will comfort us” (a better translation, I agree). Such phrases are not used in any claim about significance.”

Even if we continue to disagree about most of Part II, the important point follows:

*The* *point* *of* *all* *of* *this*:

Part II is defending Fig 1. But that defense is made only in order to support our opinion that Fig. 1 is “eminently qualified” to serve as a starting point (first brick) in the larger structure.

It is this larger structure (Figs 1 to 5 together, from http://www.torahcodes.net/new/pages/rebuilt.html) that I hope can be addressed in later posts, after we look at the simpler case, the bottom picture of:

http://www.torahcodes.net/new/pages/codes_on_codes.html

Therefore, “shelishi”, “be-moriah”, and “zeh yinachamenu” (3 out of the 5 words in this code) are not mentioned in the Melachim 11:4. How it be that this Rambam is the source for the search terms?

Fortunately, Art writes very clearly. Now that he has explained his logic, I realize that my disagreement with him is even stronger than I had thought.

I still contend that one who searches for many words and phrases cannot ascribe any statistical significance to “finding” a few of them in ELS sequnces.

The fact that you have found “moshiach”, “on Moriah”, “Third”, “this will comfort us” and “Temple” in close range with one another is (1) mathematically meaningless, and (2) meaningless from the point of view of Torah study.

Art Levitt claims that davka these terms were searched for because of their presence in the Rambam’s Hilkhos Melachim Perek 11. This gives me an even stronger basis for my suspicion that the “discoverers” of this code have entered into their computers virtually every term related to Mosiach and the Third Temple they can possibly think of. (I have been debating Lubivitcher Messianists for so long now that I have committed Hilkhos Melachim 11:4 to memory, including 3 variant texts.)

The terms you have found in this code are very arbitrary and are not suggested by the Rambam. Why did you not search for any of the following?: “melekh”, “me-beit david”, “hogeh ba-torah”, “yilchom”, “milchamaot HaShem”, “yakof kol yisrael”, “lichazek bidkah”, “nitzach kol ha-umos she-be-sevivav”, “be-mekomo”, “yikabetz nidchei yisrael”. These are the criteria defined in Hilchot Melachim 11:4. I suspect that these terms WERE searched for, but were not found. Only the positive finds are displayed, which is why the entire pheomenon has no meaning.

“Shelishi” could be logical, though it is nowhere in text of the Rambam. “Zeh yinachamenu” is absolutely irrelevant. It is not in the Rambam. Moreover, no one would ever think of it had you not found it there accidentally. The inclusion of these two terms belies the claim that the searchers were guided by the Rambam Hilkhos melachim.

It is unacceptable to enter all kinds of words with variant spellings, find a few of them in ELS, and then claim retrospectively that the odds were 1:80,000 of finding these. Your inclusion of the possible alternate spellings or synonyms in your retrospective “analysis” does nothing to improve the validity of this approach. It is not an experiment.

It is actually no different that me meeting a girl named “Katie” and saying, “Wow, what are the odds that today of all days I would meet a girl with specifically THAT name.” But then I say, “Well, even if it were Kate, Kathryn, Cathy, Catherine, Cathryn, or Katherine, it would still be very, very unlikely for this to have occured by chance.” It is, absolutely, a case of painting the target around the arrow after it has landed.

The only meaningful experiment is to subject a signle list (with no changes in advance, which has never been run through the Chumash, and which no codes proponents have any prior knowledge of) to an anlysis in the Chumash and a large number of other Hebrew texts of the same length.

Until the Torah Codes came along, no expert in the Massorah would have made such a claim.

A google to Menachem Cohen’s classic essay, “The Idea of the Sanctity of the Biblical Text” is recomended. Prof Tigay’s essay, also on the web, is recommended as well.

Even if the Codes research was mathematically sound, it proves nothing other than the fact that certain anomalies exist in the Torah text that don’t exist in texts that use consistent rules of spelling.

Perhaps the reason why the “Codes” are found in the Torah and nowhere else is because all other texts are consistent in their vowelled spelling, while the Torah is not. This inconsistency is attributable to many possible factors. In and of itself, it proves nothing other than the fact that the Torah text is inconsistent in its spellings, while other texts are not.

Lulei it was the eseres y’mei t’shuvah, I would say that you are not being honest. If you want to argue that the g’dolim that have endorsed the codes would retract in light of the criticisms, fine. But clearly at the time of the endorsments, those g’dolim were endorsing the codes as a valid kiruv tool because of their (apparent) statistical significance.

Furthermore, you publically called the work of a frum Jew “irresponsible” and “preposterous”. Vos epis?

Thanks to Sarah Shapiro for quoting the haskmaos(I’d be interested, BTW, in reading her articles in this forum as a regular or a guest poster).

I prefer learning Torah itself to Torah Codes as far as cultivating the conviction that it is by Divine design (see Michtav Meliyahu Vol III, pg 177, quoting Orchos Tzadikim’s version of the chazal about Elisha ben Avuya that “v’hadevarim m’eidim beinoseinu sh-mhar Sinia nitnu, u’metoch h’aeish netein”), but I understand that the presentations are for people who did not get to that stage yet. On the other hand, my personal ideal of a kiruv seminar would be a debate about the Torah Codes themselves(!), just as I would like to see a discussion of the “Camel, The Hare, and the Hyrax”, instead of just presenting the subject in a limited way(I was only at the “sample” Discovery Seminar for FFB following “Inspired Too”, so I do not know how the actual seminars are presented).

As far as questions, one will still need to deal with them directly even after presenting proofs; the Rambam probably would not have withdrawn Moreh Nevuchim because of the discovery of Torah Codes, whether that of Rav Weismandel Zt’l, or the ingenuity of Vilna Gaon’s insights to pesukim, or computer versions; unfortunately, we don’t have an updated Moreh Nevuchim written by someone of the Rambam’s caliber(the 19th century Haskalic book with a similar title obviously doesn’t count).

Even after inspiring people with Torah Codes, one obviously still needs to face whatever questions people might have, and perhaps give people a general framework to get “unstuck” from the questioning mode, instead of only providing specific resolutions, that in turn could be attacked by anti-kiruv organizations, or might crumble under a subsequent crisis of faith that the person themself might experience at a later date(this is a general point beyond Torah Codes).

There is a very interesting article in the latest Jewish Observer about “Adults at Risk”(I thought I had coined the term) by Rabbi Mordechai Becher and Chanan Gordon, and it makes the point that there is common denominator between some(or many) FFB and BT’s as far as both facing the same intellectual issues–I think they termed it a “symbiotic relationship”– and it mentioned Discovery Seminars. Even if one will validate Torah Codes, it will only be a single tool in the tool chest of either FFB or BT, but hopefully, no one will base their entire faith on it.

Look at the approbations very carefully which you have quoted.

Rav Wolbe z”l says that this is a valid method to “discover hints and matters from the Torah” but he does not say that it is statistical evidence for the Divinity of the Torah.

Rav Shmuel Kamenetsky says, “The sources in the words of our antecedent sages that there is in fact this way to find allusions in the Torah is known.” But, again, he does not say that this is a method to prove the Torah’s Divine authorship.

In other words, Rabbis Wolbe and Kamenetsky have endorsed ELS as a method of Talmud Torah, but not as science. This distinction is not a case of being nit-picky. It makes all the difference in the world. The critics do not claim that ELS cannot be used to show interesting things. Rather, they claim that these interesting things have no mathematical or statistical significance.

Furthermore, Rav Wolbe’s endorsement does not at all address Rabbi Adlerstein’s concerns. Rav Wolbe asks, “Are they troubled that at seminars to make people religious, that these astonishing matters are sometimes taught, and their effect on the listeners is profound? Is this what is bothering them?”

The concern is not that these seminars will inspire people to be religious. The concern is that the seminars make claims about the codes being scientifically rigorous and meaningful while, in the opnion of most statisticians who have looked into this matter (and many of them are frum) this is false. What may occur is that people who go through a codes seminar will draw a conclusion, that Orthodox Jewish Rabbis and Rebbitzens are either making naive and misinformed claims about science; or, worse, that they are willing to be deceptive in order to persaude people to embrace Torah Judaism. (I am not claiming that the proponents of the codes are dishonest. Simply that this impression might be formed by people who attend seminars and do later research. I presume that codes proponents are acting le-shem shamayim.)

Therefore, I feel that it is wrong to state on Rabbi Adlerstein’s web column that “Whatever the reason of those who oppose this method, it is certainly not a dispute for the sake of heaven…” Had Rav Wolbe heard the concerns of Rabbi Adlerstein or Rabbi Karlinsky directly I do not believe he would have claimed that they do not act for the sake of Heaven. After all, they have devoted their entire lives to Torah education. Their concerns are that the codes will have a very negative effect and create a terrible chillul HaShem, ch”v

Finally, I would encourage you to retract, if possible, the reference in your article to http://www.torahcodes.com. This irresponsible website features many of the most preposterous codes which Howard Gans himself has rejected as meaningless. See my post above about this.

I now realize that I need to provide much more explanation, in order to answer many of his good questions and/or apparent misunderstandings, and to demonstrate our solid methodology.

Although the point of the web page is to focus on the follow-ups to Fig 1, I will start by showing in 4 steps how we arrive at the conclusion that Fig 1 itself is nonetheless a showcase example; and why this is true even though we can not at this time assign a probability to it:

Step 1: the concepts that appear in this table are intimately related. As the Rambam says in “Hilchot Melachim”, Chapter 11, the *Messiah* will build the *Third* *Temple*. The Temple is located *on Moriah*. The Messiah is most commonly referred to as *Son of David* (ben David).

Step 2: The meetings among the above 5 *boldfaced* terms are in a remarkably small table. For example, looking at only *on Moriah* near *Temple* gives a raw significance near 1 in 80,000. This is a conservative estimate which is derived by comparing the Torah behavior to that of close to one million control texts. It can be sanity checked with high school mathematics, as detailed on the site in its discussion of measurement techniques. In addition, the parallel configuration seen for all five key words is a spatial constraint that contributes to this and other impressive raw significances in this table.

Step 3: The above raw significances must be corrected by the number of
equally interesting alternatives. In the case of “on Moriah”, one spelling variation, one synonym (”Temple Mount”), and a few other variations such as “on [the] Temple Mount” might be used. [Note however that longer variations are less likely to appear (in Torah or in control texts used to measure the effect), and so they have a correspondingly lesser role to play in the significance calculation.]

Step 4: We quickly realize that the exact correction required in step 3 can not be easily pinned down. However, as additional points below will make clear, we see that those alternatives are few in number and the high raw significances more than counter-balance them.

At this point, in past years, we would tend to put the code on the shelf, unreported. We would decide that step 4 is a showstopper because anything that can not be pinned down can not be officially reported. We have very recently begun to formalize new logic that enables us to move beyond this step.

Here is a summary of that new logic:

Point 1: On the page in question, the Third Temple code (Fig 1) is the
foundation, out of which the following codes (Figs 2-5), logically arise. The collection of these 5 codes forms a *structure*.

Note 1: Each individual code table is built upon others, like interlocking bricks. As each new brick is added, the structure becomes stonger and the number of options for subsequent bricks decrease. We are almost forced into looking for this structure because it is one of the very few ways we have of producing codes that can be observed by others to be adequately confined.

Point 2: Even without addressing the significance of Fig 1, we can see that the *other* figures are unexplainably significant.

Note 2: Posting 33 critiques only the original code (Fig 1, the Third Temple code) which is only the first brick of the structure.

Point 3: Careful consideration clarifies that the options for codes arising from Fig 1 are very limited, especially given:

Point 4: We focus (for now) only on common alternatives for “mikdash” (”Temple”) and common alternatives for its location.

Quoting from the web page:

“To keep the process simple and limited, we will look at only those two aspects; we will not at this time delve into alternatives for the idea of redemption (i.e. the *Third* Temple and the *Messiah*).”

Point 5: We often highlight a related phrase near a code just as a point of interest - for example, the phrase “this will comfort us” (a better translation, I agree). Such phrases are not used in any claim about significance.

Note 3: Posting 33 misses points 4 and 5, because almost all of its suggestions for alternative key words deal with redemption (“hu yigaleinu”, “hu yish’enu”, “ge’ula”, “geula sheleima”, “ge’ule emitit”, “go’el”, “go’el acharon”, “ha-melekh”,
“ha-melekh ha-moshiach”, and some alternative words for Messiah), and many of these suggestions are in the style of point 5, which we would not use for any measurements. Certainly the aspect of redemption should be addressed, and another page, http://www.torahcodes.net/new/pages/messiah.html is a start in that direction. But let us stay for now with the defined focus.

In that context, I agree that the suggestion “bayit” (House) is an additional reasonable alternative (with and without the definite article).

In summary, we see that the best alternatives are few in number; and what has happened? These very alternative key words, which we thought should penalize our raw numbers in step 3, end up producing additional significant codes (figs 2-5).

We have now discussed two examples (my posting 28 and the current posting) out of many similar and even simpler examples on the site - of inter-locking codes which address the very issues that the critics have been repeatedly raising over the years: “too many options” and “unreported failure rate”. My two postings answer the first issue; and the hidden success rate discussed in posting 28 is again relevant here, answering the second issue.

“It is known that a way exists to discover hints and matters from the Torah by reading letters at equidistant intervals. This method is found in the commentary of Rabeinu Bechai on the Torah and the works of Rav Moshe Cordovero. More recently, the tzadik, Rav Weissmandel, revealed wondrous things with this method. To my surprise, I have heard that opponents to this method have arisen claiming that various deceptions were performed by those who are involved in this method today. It is astonishing to me that they were not intimidated to state their claims after the Gaon Rav Shlomo Zalman Auerbach gave his clear agreement to the method of equidistant intervals, and after Rav Auerbach’s son, the Gaon Rav Shmuel, censured them. I am uncertain of these opponent’s intentions. Are they troubled that at seminars to make people religious, that these astonishing matters are sometimes taught, and their effect on the listeners is profound? Is this what is bothering them? Whatever the reason of those who oppose this method, it is certainly not a dispute for the sake of heaven, and we must strengthen those who are engaged in this method, for it is a totally honest endeavor. May they be encouraged and see blessed results to their deeds, and may they continue to increase the honor of our holy Torah and its influence on all who seek the Almighty’s closeness. May the Almighty help those who engage for the sake of heaven in this honest discipline of studying equidistant intervals.”

Shlomo Wolbe
With the help of G-d, Erev Shabbos Kodesh, Parshas Chukas – Balak, 5759

“Two “Tzanteri D’dahava” (Aramaic; literally, “Golden vessels”, i.e., people of impeccable character), may they live long, came before me and set before me the allusions in the Torah by way of ELS in the Torah. Although I am not adept in the discipline of statistics, what has clearly emerged from their words, was that there was no trickery or dishonesty on the part of these who search for the Word of Hashem and whose will is only to uplift Torah and to exalt her.
The sources in the words of our antecedent sages that there is in fact this way to find allusions in the Torah is known. Thank G-d, we in our days have been privileged to the means to find many more [such allusions] for his Namesake.

“There surely are those that will falsify and misconstrue and about them the verse says, “the righteous shall go in [the ways of Hashem] and…” but this is no reason not to use the aforementioned allusions to open the blind eyes of many of our blinded Brethren of B’nei Yisroel….

“With blessings of success,
Shmuel Kamenetsky” (translated into English by Rabbi Sholom Kamenetsky

Finding weaknesses in the peer review process, however, does not invalidate a paper’s conclusions. Such invalidation can occur only upon finding the purported flaws in the paper and demonstrating that as a result of these flaws, the conclusions are unwarranted. One must show that following a data preparation protocol no less strict than the one to which we adhered, one can produce similar but spurious results in texts other than Genesis.

I challenge Dr. Simon to do so, without a deadline.

Here is a further challenge: Seven papers were published (as acknowledged by Dr. Simon and contrary to Rabbi Karlinsky’s mistaken assumption,) of which six demonstrate Torah codes independenltly of the other five. The codes differ from each other in nature, and in some cases the methodology also differs. To prove that there is no evidence for codes, one must show a fatal flaw in each of the five.

I’ve been involved in the past with a number of conferences in my own field, thermal spray coatings, as an author/presenter, session chairman, and reviewer of submitted papers. I’ve also been an associate editor of the main journal in our field, and still review submissions to it from time to time.

The poster presentations at conferences are mainly to provide a window into the ongoing activities of graduate students and professors. The posters are supposed to have merit and often do, but they do not go through anything like the scrutiny given to papers to be published in the conference proceedings.

The cream of the crop, the papers selected for publication in a peer-reviewed journal, go through an even higher level of review. Some of the papers published in such journals start off as conference papers.

Conference papers and peer-reviewed journal papers may be published that do not represent the scholarly views of the reviewers or editors, as long as the authors made well-organized, reasonably well-supported cases for their findings.

As regards strictly oral presentations, these can be entertaining and informative, but the audience is often left to wonder why the author did not take the further step of reducing it to writing.

For background of readers not familiar with large scientific conferences, the ICPR 2006 conference (see
http://www.comp.hkbu.edu.hk/~icpr06) had four levels of “speakers”. There are two kinds of invited speakers - plenary speakers of which ICPR 2006 had four. Invited papers of which ICPR had nine. These are chosen by a program committee. There are also two levels of submitted contributions, that is ones that are not specifically solicited but volunteered by the presenters themselves - a limited number of oral presentation slots and a non-limited number of poster submissions.

What a poster means is that the presenter puts a single large sheet up in an area and attendees walk around and look over those posters that seem interesting. In this case the guidelines say “The poster board’s orientation is vertical (portrait). It is 1 m (w) × 2 m (h). It is highly recommended to have a one-piece poster on the board. For easier viewing, the poster length should not exceed 1.2m.” 1.2 square meters is not a lot of room for information.

The submitted contributions were in 5 parallel sessions. Session 2 (which seems to have had all the Codes presentations) had 97 oral presentations and 270 poster presentations. Prof. Haralick gave two oral presentations. So far as I can tell all the others including Mr. Gans gave Posters.

I cannot speak of standards at Pattern Recognition Conferences but I do know about conferences in mathematics and physics. Subject to limitations on slots, a senior establish person like Prof. Haralick would be slotted for a talk almost no matter what their subject is.

The threshold for poster presentations is very low.

In this case there were three months between the deadline for submission of poster proposals and notification of acceptance. Given processing time by a central administration, I’d guess that referees had at most 10 weeks to look over submissions. Given the number of accepted submissions, I’d have to guess a typical referee had 3-10 items to look at. Referees are not paid specifically for their refereeing (at least not in math or physics) and are typically research scientists with lots of other responsibilities.

There is no reason to believe that the referees for such posters had any prior experience in the subject of Bible Codes but they had ten weeks to master what Sara Shapiro commented is “a deep and incredibly complex matter
which requires years of study to begin to comprehend.” In particular, it seems likely that the referees knew nothing about issues like the flexibility in possible name spellings for cities or Rabbis.

Paper referees can always ask for more time but for conference presentations there is an absolute deadline. Moreover, paper referees do feel some responsibility for the correctness of what they approve. I am not sure exactly what guidelines are for posters in pattern recognition but in the fields I know, they are not as high as for research papers and certainly do not require the referee to understand and accept all the conclusions in what is after all a VERY abbreviated presentation.

I think it is possible some of the referees of Mr. Gans poster saw his background in cryptology and recommended acceptance without looking at any further details. The idea that they carefully read the complex WRR and BBKM papers and absorbed them, let alone understood enough to agree with Mr. Gans conclusion in 10 weeks is absurd.

All along I have complained that the codes proponents abuse the public’s perception of what peer reviewing means especially given the fact that you really do need to know something about Jewish history to evaluate this kind of work.

Given the low standard on poster presentations, the latest iteration of this mistaken notion that peer reviewing is a certificate of Kashruth is even worse. The idea that some ten week referee with no prior knowledge can trump the opinions of BBKM and of Furstenberg and Aumann who had thought about this for many years is ludicrous.

Since I am correcting the record I must also take exception to Mr. Gans assertion that “But what was left out was that several people (including one person not on “our side”) on the committee strongly rejected the conclusions drawn..” The committee had five members: Aumann, Furstenbrg, Bar-Natan, Rips and Lapides. The first two explicitly reached the conclusion that the experiment while not perfect presented problems for someone who believes the codes can be scientifically proven. Bar-Natan never changed his mind about the codes but refused to sign the report because he thought it not negative enough. So your “several people” would seem to mean two: Rips and Lapides. My understanding is that Lapides (the only non-mathematician on the committee) was there explicitly at Rips’ invitation. Given that he was there at Rips suggestion, why should we think he is “not on (y)our side”?