Majority in Halacha and in General, Lesson 4
This transcript was produced automatically using artificial intelligence. There may be inaccuracies in the transcribed content and in speaker identification.
🔗 Link to the original lecture
🔗 Link to the transcript on Sofer.AI
Table of Contents
- Majority in a religious court: Sefer HaChinukh, Rabbi Shimon Shkop, and a present majority versus a non-present majority
- The absence of feedback and experience: why “the majority is usually right” is not a sample
- Probabilistic calculation and common errors: Shev Shema’tata, normalization, and conditional probability
- Examples of statistical errors: Munchausen by proxy, Bible code, and medical tests
- Additional evidence, initial screening, and the parallel to the legal world
- When we follow the majority: Rabbi Yonatan Eybeschutz, “I have no doubt,” and personal position
- The need for a reason in order to doubt: Rabbi Kook, one witness, and presumptions
- Disputes among medieval authorities (Rishonim) / later authorities (Acharonim) and rules of halakhic ruling: doubt only when you have no position
- A majority of dependent opinions: rabbi and student, textual witnesses, and the Magen Avraham
- Forgeries and authority: the Jerusalem Talmud on Kodashim, Besamim Rosh, and the distinction between persuasion and obligation
- A majority based on different reasons and the judicial paradox: Rema, Shakh, and law / Jewish law
- Zadorov and reasonable doubt: majority about the very existence of doubt
Summary
General Overview
The text concludes the discussion of following the majority in a religious court by comparing a non-present majority with a present majority, and argues that even according to Sefer HaChinukh, the justification of “in most cases the majority is right” is not a generalization from a sample but a probabilistic calculation that usually has no independent feedback, and therefore resembles a present majority more than a non-present one. It develops a fundamental distinction between different probabilistic questions (what are the chances that judges will hit the truth, as opposed to what are the chances that the truth is such given that they ruled that way), demonstrates intuitive failures in probabilistic inference (Munchausen by proxy, medical tests for a rare disease, the “Bible code”), and concludes that rules of majority and decision apply mainly in a case of doubt. It then adds comments on the conditions for applying majority: the need for a reason in order to doubt, dependence between opinions (rabbi and student), a majority formed from different reasons and the judicial paradox, and finally applies this to the claim of “reasonable doubt” in the Zadorov case.
Majority in a religious court: Sefer HaChinukh, Rabbi Shimon Shkop, and a present majority versus a non-present majority
He presents Rabbi Shimon Shkop’s question about Sefer HaChinukh’s explanation: according to Sefer HaChinukh, we follow the majority because in most cases the majority is right. He asks that this would seemingly be a non-present majority, because it relies on a collection of many panels in which there was a majority against a minority, and not on the judges currently before us, so it is hard to understand the Talmud in Chullin, which identifies “incline after the many” with a present majority. He answers that even if technically this looks like a non-present majority, on the conceptual level this is not a generalization from a sample like “most women give birth at nine months,” but rather an a priori probabilistic consideration, like nine kosher stores versus one non-kosher one, and therefore it resembles a present majority more closely.
The absence of feedback and experience: why “the majority is usually right” is not a sample
He argues that in a religious court there is no way to know independently in which cases the majority was right and in which it erred, and therefore it is impossible to build a sample from which one could generalize that the majority is usually right. He connects this to Zadorov as an exceptional example in which there is sometimes an independent channel such as genetic testing that provides feedback on the panel’s decision, but in most cases there is no such channel beyond the evidence that the court itself considered. He concludes that the claim that “the majority is right in most cases” comes from probabilistic calculation and the assumption of judges at a good and roughly uniform level, not from empirical experience.
Probabilistic calculation and common errors: Shev Shema’tata, normalization, and conditional probability
He cites the Shev Shema’tata in Mekor Kodesh and presents a model of “judge quality” p between 0 and 1, and clarifies that a naive calculation by multiplying probabilities creates a paradox if one does not take the alternative possibilities into account and normalize according to the formula of total probability. He emphasizes that comparing likelihoods must be done against the alternative, because the probability of one possibility by itself does not add up to 1 unless all the alternatives are defined and weighted. He sharpens the distinction between “given that the truth is such, what is the chance that the judge will say so” and “given that the judge said so, what is the chance that the truth is such,” and shows that reversing the question changes the picture completely.
Examples of statistical errors: Munchausen by proxy, Bible code, and medical tests
He tells the story of Sir Roy Meadow and Munchausen by proxy, in which it was claimed that two crib deaths are “one in sixty-four million” and therefore the mother murdered them, and explains that the error is failing to examine the alternative and failing to understand the meaning of a rare event within a large population. He illustrates a similar principle through the “Bible code” and argues that one can “tailor” an ad hoc algorithm that will find special results even in generic books, and therefore the rarity of the result itself, without examining the space of searches and the alternatives, is not decisive. He brings a medical example of a rare disease (one in ten thousand) and a test with 99% reliability, and calculates that out of a million people tested one gets about 10,000 false positives against 100 actual sick people, so that given a positive result the probability of actually being ill is only about 1%, and concludes that the diagnostic tool has to be far sharper than the rarity in order to “catch” a rare phenomenon.
Additional evidence, initial screening, and the parallel to the legal world
He argues that even a small additional piece of evidence can completely change the calculation because it narrows the reference population, and therefore in the legal world there is a requirement for something in addition to a confession. He explains the role of the prosecution as a filter that reduces the relevant population, so that passing that filter raises the probability of guilt even before the court rules. He warns that 99% evidence “is worth nothing” if good initial screening has not been done, because the rate of murderers in the population is tiny, and therefore even a system with a small error rate may produce many false convictions relative to the few guilty people.
When we follow the majority: Rabbi Yonatan Eybeschutz, “I have no doubt,” and personal position
He tells the story of Rabbi Yonatan Eybeschutz, who answered a priest that we follow the majority only where there is doubt, and argues that this is not a joke but a serious principle: when there is a decisive sign, such as a “mehadrin kosher seal,” there is no room for decision by majority of stores. He applies this also to the public sphere and validates Yitzhak Ben-Aharon’s statement that “the people need to be replaced,” because if a person has a clear position, he is not obligated to assume that the majority was right. He states that decision rules such as majority or laws of doubt operate only after it has been determined that there is doubt and one cannot form an independent position.
The need for a reason in order to doubt: Rabbi Kook, one witness, and presumptions
He cites Rabbi Kook’s interpretation in Ein Ayah of the Talmud in Shabbat 30 about someone who comes with the claim “your wife is my wife and your son is my son,” and argues that according to Rabbi Kook, in order to doubt you need a reason, and one witness can at most arouse doubt but is not enough where there is no reason to cast doubt on an existing presumption. He brings the principle “we stone and burn on the basis of presumptions” as a case where even in capital law we act on a family presumption despite the theoretical possibility of illegitimacy, because we do not begin doubting without a trigger. He compares this to “Russell’s celestial teapot” as an illustration of the position that the absence of a reason does not create balanced doubt.
Disputes among medieval authorities (Rishonim) / later authorities (Acharonim) and rules of halakhic ruling: doubt only when you have no position
He argues that the common view according to which the very existence of a dispute among medieval authorities (Rishonim) or later authorities (Acharonim) creates laws of doubt is untenable, because if a person has his own position, he is not in doubt and does not need rules. He presents rules of halakhic ruling such as “the law follows Rava” and Ya’al Kegam as rules directed to one who has no independent ruling, and explains that this is why there are many exceptions and rules with exceptions. He cites Rabbi Elchanan in Kovetz Shiurim on Bava Batra regarding rules of halakhic ruling among Tannaim, and the discussion whether the rules are “summative” or “directive,” and argues that understanding the rule as directive explains the possibility of exceptions where there is independent reasoning.
A majority of dependent opinions: rabbi and student, textual witnesses, and the Magen Avraham
He presents the Talmud in Sanhedrin 36, according to which a rabbi and his student are counted in a religious court as one and not as two, because the student may reflect the rabbi’s view and therefore there are not two independent opinions here. He notes that in matters of ritual impurity and purity they are counted as two, and emphasizes that there are distinctions between a student who repeats Jewish law from his rabbi and one who received only modes of thought and may deviate. He applies this to textual witnesses of manuscripts: a manuscript copied from another is not an additional textual witness. He explains similarly the Magen Avraham on the permission to say something in the name of a great person “so that people will accept it,” and argues that the permission makes sense only if the listener does not automatically accept authority but is only willing to consider it seriously.
Forgeries and authority: the Jerusalem Talmud on Kodashim, Besamim Rosh, and the distinction between persuasion and obligation
He tells of the forgery of a Jerusalem Talmud on Kodashim at the beginning of the twentieth century and the criticism of it, including a story about the Rogatchover, and formulates the difference between attributing something to a great person for purposes of persuasion and attributing it to a binding source. He argues that saying something in the name of the Talmud is not like saying it in the name of “a great person,” because the Talmud has authority that obligates acceptance and not merely consideration, and therefore a forgery that presents things as binding authority is flawed in a different way. He notes that Besamim Rosh with the commentary Kisei Deharsena is a forgery, and that one can still examine the arguments themselves, but as far as counting opinions and the numerical standing of halakhic decisors, a forgery has practical implications.
A majority based on different reasons and the judicial paradox: Rema, Shakh, and law / Jewish law
He presents the Rema in Shulchan Arukh, siman 25, סעיף 2, according to which the majority in a religious court determines the ruling even if the reasons differ, and notes that the Shakh discusses at length and distinguishes cases where this is not so. He describes a judicial paradox in which a majority on the bottom-line conclusion acquits even though on each of two intermediate questions there is an opposite majority that would obligate, and raises the question why the ruling is determined by the bottom line rather than by majority on each component. He brings an example from the Supreme Court (Arbel, Meltzer, Vogelman) in the context of the Hague Convention, where two judges agreed on the result not to return a child, but each based it on a different exception, so that on each exception separately there was a majority against it. He suggests that here the consideration is simplicity or efficiency and not “truth,” and distinguishes that a democratic majority is not sensitive at all to dependence or to reasons, because the question there is public will rather than factual truth.
Zadorov and reasonable doubt: majority about the very existence of doubt
He describes an argument raised in the latest Zadorov discussion, according to which if one judge accepts the appeal then there is “reasonable doubt” and therefore one must acquit, and he suggests a distinction: if the minority opinion is a positive determination that “Zadorov did not murder,” then the very determination of the Supreme Court may create reasonable doubt, but if the minority opinion is only “I have reasonable doubt,” then the factual question is whether there is reasonable doubt, and on that question the majority decides. He concludes that the correct structure depends on the precise reasoning of the minority opinion and on understanding the majority as deciding about the very existence of the doubt.
Full Transcript
[Rabbi Michael Abraham] To talk a little about—meaning, to finish up a bit more of the issue we’ve been discussing until now, and then to talk about a few aspects of following the majority, and with that we’ll wrap up this topic. Last time we saw the question I asked that Rabbi Shimon Shkop asked, namely that following the majority according to Sefer HaChinukh is basically because in most cases the majority is right. And then the question was: why does the Talmud see the majority in a religious court as a present majority? Basically, according to Sefer HaChinukh’s explanation, this is a non-present majority, because the majority here does not refer to the three judges standing before us, but to a collection of a great many panels in which there was a majority against a minority,
[Speaker B] within
[Rabbi Michael Abraham] that collection, where in most cases the majority was right. That’s a non-present majority, not a present majority. So the question is: according to Sefer HaChinukh’s explanation, how do you explain the Talmud in Chullin that “incline after the many” is a present majority? So I said that it’s true that this is a non-present majority, but on the conceptual level it doesn’t resemble a non-present majority. Why? Because a non-present majority is a majority whose basis is a generalization from a sample. When I say that most women give birth at nine months, I’m saying: I know a number of women, and among them most gave birth at nine months; I assume this is a representative sample—an assumption that of course is not always correct, but that’s what I assume—and then I generalize and say that generally in the world most women give birth at nine months, and for me that’s some kind of natural law. By contrast, a present majority, as in a majority of stores where there are nine kosher stores and one non-kosher one, and I find a piece of meat in the market, then I say that most likely this piece belongs to the majority of kosher stores. Now there this has nothing to do with a sample and it’s not a natural law. There’s no sample here. I didn’t encounter several pieces in the past and see which store they came from, and then from that generalize and say that most pieces come from the kosher stores. This is an a priori consideration. It’s some kind of probabilistic calculation that is not the result of experience, of checking a sample of cases and generalizing or creating a natural law. It’s also not a natural law; it’s a contingent majority. In this case there happen to be nine kosher stores and one non-kosher one. Somewhere else it could be the opposite. There is no nature of the world here that behaves this way, and therefore it’s also not relevant to work with samples. In other words, this is not a scientific determination. It’s a contingent determination; it’s just a contingent fact. In that sense, the majority in a religious court, even according to Sefer HaChinukh’s description, is not a non-present majority. Because we have no way whatsoever of knowing—and this also connects to the whole Zadorov issue—we have no way of knowing in which cases the majority was right and in which cases the majority was wrong. Suppose we go through the examples in which there were disagreements between majority and minority, in all the cases, let’s say, that we somehow manage to get our hands on, and let’s even assume it’s a representative sample. We now have to check, in each case, whether the majority was actually right. We know what the majority said; we do not know that the majority was right. How do we check whether the majority was right in a particular case that came before us? There’s no way to do that. How would we know? What we can do is try to trace the evidence—that’s what the court itself did. In other words, we have no way to know independently, through some feedback in another channel that doesn’t depend on the court’s decision, whether the majority really was right or wrong. And since that’s so, we have no way of making a sample of cases. In no case that comes before us do we have a way of establishing whether the majority was right or wrong. Therefore the statement that the majority is generally right is not the result of a generalization from a sample. So if that’s the case, then how do we know it? From some reasoning, we say there is a greater chance that the majority is right than that the minority is right; there’s some probabilistic calculation here. I said this before, with a qualification: only in a situation where everyone is more or less judges of a good and uniform level. Good and uniform level, both things. Because at a bad and uniform level, then most likely the minority is right. But at a good and uniform level, then there is a greater chance that the majority is right than the minority. But that is the result of a statistical or probabilistic calculation. It is not the result of experience. We cannot have such experience. We cannot know whether the majority is right or wrong. As I said, except for those cases where in the end something happens, as with Zadorov, if that indeed happens, where ultimately some genetic test gives us some feedback on the decision of the first panel. Then suddenly we can discover that the panel was right or that the panel was wrong, because we have an independent channel that gives us feedback on the matter. Usually that doesn’t happen. Usually we have no avenue other than the evidence that stood before the religious court. And since that’s so, essentially the majority—the claim of Sefer HaChinukh that the majority, that the majority is right in most cases—is not the result of a generalization from a sample but the result of a probabilistic calculation. And if it’s the result of a probabilistic calculation, then it is basically a present majority, not a non-present majority.
[Speaker C] When you talk about subjective bias, or about assumptions that there are certain biases in which the majority is wrong—in other words, most people will see something in a certain way because of some cognitive bias they have—then the majority will be wrong in that case.
[Rabbi Michael Abraham] Okay, so what?
[Speaker C] Therefore, if there are cases where a religious court is presented with a case where there is suspicion that there’s some bias here, those are special cases.
[Rabbi Michael Abraham] I’m talking about a case where we don’t have any reason to suspect. I’ll talk a little about special cases in a moment, but in a place where we don’t have any reason to suspect bias, then no. By the way, and you also have to pay close attention—maybe I’ll comment on this in a moment—there’s a difference between saying that in a case of a certain bias the majority will err, and the claim that given that the majority said something, it was probably wrong. That’s the reverse conditional probability. In other words, saying that in such a situation the majority is most likely wrong is not the same thing as saying: the majority determined something, therefore it is most likely wrong. That’s the reverse statement. I’ll illustrate that in a moment. In other words, one does not always entail the other. So that’s where we got to last time. What I want just to note a bit about this probabilistic calculation is: how do we know that the majority is in fact right? I mentioned that the Shev Shema’tata in Mekor Kodesh—he asked, let’s say a judge is of quality p, right? Say 0.7, between 0 and 1. A judge of quality p = 1 means he is always right. p = 0 means he is never right. Except that that’s excellent. A judge of quality one-half is the most dangerous one, right? A judge of quality zero is no problem—whatever he says, do the opposite. But—yes? You know instantly. Yes, yes, assuming I know. But it’s like with liars; we once talked about liars, didn’t we? In other words, someone who is a pathological liar, for me, gives the same information as someone who is a pathological truth-teller. The most dangerous person is someone who tells the truth or lies randomly. You can’t know. I mentioned that I once read an article in Iyyun saying that you can get the truth even out of such a person. In any case, how exactly do you make this calculation? Given that there is a judge of quality p, and let’s assume all the judges are of the same quality, so let’s say p is 0.7. Then the chance that one judge is right is 0.7. The chance that two judges are right is 0.7 squared, 0.49. In other words, that’s less than fifty percent.
[Speaker D] No, it’s 0.3 squared. What? That the chance they’ll be wrong is one minus 0.3 squared—that’s really not correct.
[Rabbi Michael Abraham] Not the same thing; that same point I’m getting to in a moment, yes. So the point is that seemingly if you multiply the chances, it actually comes out that the more judges there are, the smaller the chance that they’ll err or that they’ll be right. Then the question is: why does Sefer HaChinukh assume that usually the majority is right? How is the probabilistic calculation built? Again, this is not from experience, it’s a probabilistic calculation. How do you do the probabilistic calculation? So of course here the calculation is wrong, it’s wrong because—there’s an indication there—so what’s the chance they’re both wrong according to that same calculation? 0.09, right? 0.3 squared. So it doesn’t add up to 1. Either both were right or both were wrong. So how can that be? One is 0.09 and the other is 0.49, and together that’s 0.58. Where did the other 0.42 disappear to? If those are the two possibilities, either both were right or both were wrong, and it doesn’t add up to 1, that means something is messed up in the calculation.
[Speaker D] It could be that one erred and one didn’t. Huh?
[Rabbi Michael Abraham] No, but they’re saying
[Speaker D] the same thing.
[Rabbi Michael Abraham] But they’re saying the same thing. If this one erred then that one erred too, because I’m talking about the majority opinion. These are two who say the same thing; they say so-and-so murdered. So either both erred or both were right, there’s no other possibility here. So it ought to add up to 1. Okay? Therefore, in simple terms, intuitively it’s clear—this is the total probability formula—but in principle what you need to do is add those two things, put them in the denominator, and the numerator divided by the denominator is the probability. When you take the 0.49 and divide it by 0.58, that’s the probability, and 0.09 divided by 0.58 is the probability. Just normalize by the total. Do you see what I mean? In other words, your 0.49 is up against 0.09, not against 0.51, but against 0.09. That’s the ratio between the probabilities. So if you want them to add up to 1, just divide by 0.58 and then they’ll add up to 1. But the ratio between them is the ratio between 0.49 and 0.09, and indeed that ratio strongly favors the side that they’re right. So what we’re basically saying here is that many times we look at a certain probability and fail to check the probability of the alternative. We test a certain alternative, it looks implausible to us, and therefore we say it’s probably not true. But you always have to check the likelihood of the alternative, because it doesn’t always add up to 1. We made the calculation incorrectly. There are lots of examples of this. I’ve written about it a number of times, both on the website and in articles. I’ll maybe bring one more example. There’s the story—we may have talked about it once—of Munchausen by proxy. There was a woman in Britain whose two children died of crib death. And some doctor came, Sir Roy Meadow, who identified this syndrome, Munchausen by proxy, and he claimed that that’s what happened there. What is Munchausen syndrome? It’s an attempt to attract attention, like Baron Munchausen, okay? Munchausen by proxy is an attempt to attract attention by means of a third party, through someone else, not through myself. Not that I do all kinds of extravagant things so people will notice me, but by means of something else. For example: I’m terribly unfortunate because my children died. So I kill them in order to attract attention because I’m so miserable and everyone will pity me, and so on, for example. So he claimed that’s what happened there. Now he had a very simple probabilistic argument. He said: the chance of crib death is one in eight thousand. A child dying of crib death is one in eight thousand. The chance that two children die of crib death is one in sixty-four million. Therefore it’s clear that that’s not what happened. Therefore it’s clear she killed them, because one in sixty-four million is a negligible probability.
[Speaker C] One in sixty-four million means there’s one in sixty-four million. What? One in sixty-four million means
[Rabbi Michael Abraham] means there’s one in sixty-four million. Yes, but what they forgot to check there—some statistician came along; the woman, by the way, sat in prison for years, and afterward apparently the judge was an idiot too, not just the doctor. At some point some statistician on behalf of I don’t know who came and explained to the judge that he had made a mistake in the calculation. Yes, they made a mistake in the calculation. And she really was released, and the man almost lost his license to practice medicine, Roy Meadow. He became “Sir” following the discovery of the syndrome, by the way, and to this day there’s an argument whether there is such a syndrome because of these problems, since it’s hard to prove. And the claim was that you can look at it in several ways. For example, if it’s a one in sixty-four million chance, fine—then out of sixty-four million homes, in one of them it will happen. So what’s the problem if the chance is small? Or alternatively, think about the chance that a woman murders her two children. What’s that probability? After all, that’s the alternative, right? So if one alternative is that they both died without her murdering them, and the second alternative is that she murdered both of them—let’s ignore for the moment the possibility that one died and one was murdered, let’s say either she murdered both of them or both died naturally—what’s the probability that a mother murders both her children? I don’t know, but it’s very small. Okay? So the alternative isn’t exactly all that likely either. So that means it also doesn’t add up to 1, of course, and that means something is flawed in the calculation. Okay? It’s like Sherlock Holmes in The Sign of Four, where he says that if you eliminate the impossible and all that remains is the improbable, then that’s probably what is true. In other words, you always have to check the alternative. It’s not enough to look at one possibility and see that it’s improbable and therefore reject it. If the other side is even less probable, then apparently that improbable thing is what happened. So many times we don’t make comparisons between alternatives, and therefore we adopt or reject something because by our reckoning its probability is low. Comparing between alternatives is usually an indication that our calculation is incorrect. You can see it in another way. The Bible code and the whole story.
[Speaker C] What? The Bible code and the story where they discovered that probabilistically it’s supposedly very unlikely the way they find the codes, so there must be something meaningful there, with letter skips and so on. And the contradiction is that suddenly they went to another generic book, and there too—in War and Peace and so on—they suddenly discovered that the same improbable thing happens there too.
[Rabbi Michael Abraham] Yes, although they updated it a bit—not important. There too you can find an algorithm that will give us special results, but I think it’s a slightly different algorithm. Never mind, the algorithm isn’t what matters. Here too the algorithm is very tailored. I once went to a lecture—they sent me to check it out on behalf of the yeshiva; I was in Bnei Brak, at Yeshivat Netivot Olam for ba’alei teshuvah. They were interested in whether this is a tool one can work with. So they asked me to go; there was some lecture in Givatayim, some statistician named Witztum came to speak to them, to give a lecture on the matter. Honestly, I found it very unconvincing. Again, I didn’t check everything to the end; after all, there are serious people on both sides here. But it was very unconvincing, because the feeling was that he was tailoring it ad hoc so that the results would come out. Always seventy percent of the text, eighty percent—he always tailors it in such a way that it comes out special. You still need to do the calculation to see, fine, because tailoring—but once you start tailoring, it already begins to look suspicious. I can always find an ad hoc algorithm that will give me whatever special result you want. No problem at all. As long as I can define the algorithm in a way that steers me toward the special results. It’s not a problem to do that. I can find you every murderer of every murder on earth in whatever book you want. Give me the murderer and I’ll find the algorithm that always gives his name. Based here on a set of points
[Speaker B] you can fit whatever function you want.
[Rabbi Michael Abraham] Exactly. The claim in that context, in the context of the Munchausen syndrome, can really be seen a little differently too. When I want to test for a rare phenomenon, this too is something Kahneman talked about, apropos cognitive biases—when I want to test for a rare phenomenon, I need a tool of sufficiently good quality to catch it. Right? To catch fish of a certain size, you need a net whose holes are significantly smaller. Okay? So here too, the phenomenon that a woman murders both her children is a very rare phenomenon. Therefore, the tool by means of which one can detect such a phenomenon has to be a tool whose significance level is substantially better than the probabilities we’re talking about here. And that was not the case there. An example: suppose you go to a doctor—I once talked about this to medical students at Tel Aviv University. It was very interesting. Afterward I heard that at least in Jerusalem they already teach this, but until a few years ago they didn’t. You go to the doctor and there’s suspicion of a rare disease. And the doctor sends you for a test. And the test is 99% accurate, with 1% error, say in both directions. In other words, either you’re sick and the test says you’re healthy, or you’re healthy and the test says you’re sick; it can miss by one percent. Okay? The test says you’re sick. What’s the probability that you’re sick?
[Speaker E] Almost the same thing.
[Rabbi Michael Abraham] Depends how rare the disease is. Okay? If the disease is very, very rare—let’s say the disease is one in ten thousand in the population, which is fine, there are quite a few such diseases; these aren’t super-rare diseases, one in ten thousand. Okay? So let’s think: suppose we take one million people. Among them, if it’s one in ten thousand, then there are a hundred sick people among them, right, out of a million people. Okay? Now we test the whole million. Again, that’s how you do the probabilistic calculation. We test the whole million. Since the test errs by one percent, there are here a million healthy people—well, a million minus a hundred, but approximately a million, okay? One percent of them will come out sick because there’s an error. How many is that? Ten thousand come out sick. Okay? When in fact there are only—wait, no, wait,
[Speaker D] there are a hundred who are actually sick.
[Rabbi Michael Abraham] Right. There are a hundred who are actually sick, and there are ten thousand who come out sick on the test. So given that you came out sick on the test, the probability that you are sick is one percent. Even though the reliability of the test is ninety-nine percent. And what that means, essentially, is that the reliability of the test—that is, the size of the holes in the net—has to be on the order of magnitude of the size of the fish you want to catch. In other words, the probability, or the rarity, of the disease has to be more or less on the same order of magnitude as the reliability of the test, and then you get roughly an indication—even then it’s fifty percent if it’s the same order. Okay? If it’s significantly better, then you can begin to relate to the results. And that’s a very interesting point. Ask doctors—I’m sure most of them will fail here. Ask doctors: they’ll send you for a 99% test, the test is good, and the test says you’re sick. They’ll go into immediate hysteria. They’ll rely on that. In other words, if you only have additional indications, then it changes the whole picture, by the way. And therefore in the legal world too, many times there is—for example, in the confession of a litigant, where a person incriminates himself, there has to be some additional something, as in Zadorov, we keep coming back to Zadorov, there has to be some additional corroboration, that’s how it’s defined in law. What does that mean? Some additional evidence besides the confession. Because indeed some additional evidence, even if it’s small, completely changes the picture because of this issue. In other words, if there is some additional evidence that I’m sick, then the whole calculation changes. Because now, say, instead of belonging to the million people, I really belong only to fifty thousand people, among whom there will be ten thousand sick people, because there is already some corroborating element. It’s not worth much—in other words, it’s just some corroboration—but it significantly narrows the potential million. And now it’s already one in five, ten thousand out of fifty thousand. That corroborating element, although weak, can completely change the picture. So that’s why I think this is a point that even jurists don’t understand, the calculation that stands in the background of the matter, but behind it there is actually something very significant. And therefore, by the way, the checks the prosecution does before deciding to indict—I was very puzzled by this. Why do you need to do that? Just file charges, the court will sit over the evidence and decide. Why does the prosecution have to decide whether the evidence is sufficient and then the court decides again? What for? You’ll tell me efficiency and things like that. The claim here is that there are efficiency considerations too, that’s something, but really it would be better to hold more judges than to maintain the prosecution. The truth is, the point is that it’s not only a question of efficiency. Rather, once the prosecution—once you pass the prosecution, your probability of being guilty increases dramatically. The relevant population relative to which you are measured becomes smaller. Now when the court—its chance of error is small, but also the fish it wants to catch is bigger, because in the end, the chance that the court errs here means that the prosecution also erred. In other words, there is some corroborating factor here that helps the prosecution’s decision.
[Speaker C] After all, if they brought a thousand people for every murder, just think—
[Rabbi Michael Abraham] for every murderer, what’s the chance that someone is a murderer? How many murderers are there? It’s a terribly small probability. Now suppose we go to court and the court, in its evidentiary ruling, has a one-in-a-thousand chance of error. So? And how many murderers are there in the population? More than one in a thousand? So what good is the whole discussion? The whole court proceeding is worth nothing. Unless, of course, you make sure there is some corroborating factor that lowers the chance of error to significantly less than one in a thousand, so that the probability that the court erred will still be small enough relative to the probability that the person really didn’t murder. Basically, what lies behind these things—and this is what I said to you before when we return to judges—when we deal with judges, there’s a difference between saying what is the chance that two judges will say the truth—that’s p squared, proportional to the probability—and asking a different question: not what is the chance that two judges are right. “What is the chance that two judges are right” is ambiguous. What is the chance that two judges hit the truth? That’s p squared, let’s say, or proportional to p squared. But given that two judges said thus, what is the chance that they were right? That’s a different question, and the total probability formula reverses the picture. That’s conditional probability. Given that they said thus, what is the chance that the truth is that way? “What is the chance that they’re both right”—that question means, given that this is the truth, what is the chance that they will say it, right?
[Speaker C] That depends on the possibilities; there aren’t always just two possibilities.
[Rabbi Michael Abraham] Doesn’t matter. I’m saying on the conceptual level this is the reverse question. I’m not talking right now about what the relation between the outcomes is; that depends on the circumstances. But first of all you need to know that this is a reverse question, and a totally different one. The question what is the chance that two judges will say the truth or hit the truth—that’s the question: given that the truth is that he murdered, what is the chance that two judges will reach the conclusion that he murdered? That’s p squared. But the question we are asking, that Sefer HaChinukh is asking, is the reverse question. Two judges said he murdered; what is the chance that he really murdered? That’s the reverse question. What is the relation between those two questions? Sometimes they’re worlds apart. That depends on the conditional probability; that already depends on all the possibilities. Therefore one has to be very careful. When I say what is the chance that two judges are right, that’s a very vague statement. It can be interpreted as the first question and as the second question. What is the chance that two judges are right can be interpreted as: what is the chance they both hit the truth—p squared. Or: what is the chance that when they said so, they really were right? They were right, meaning that that really was the truth, and that is absolutely not p squared; it’s something entirely different. Okay? Therefore, many times—this is well known to anyone familiar with probability calculations—many times, the ability to formulate the question precisely gives you half the answer. In other words, many times the mistake is not in the calculation; the mistake is in how you ask the question.
[Speaker E] What’s the difference between “they were right” and “they hit the truth”? I didn’t understand, what…
[Rabbi Michael Abraham] Suppose I say that a judge has quality 0.7. In other words, in 0.7 of the cases he hits the truth. What does that mean? That given that so-and-so murdered, what is the chance that the judge will accuse him, find him guilty? 0.7. Fine? Now a different question. The judge accused him. I ask: what is the chance that he murdered? The first question was: if he murdered, what is the chance that the judge says he murdered? The other question is: assuming the judge said he murdered—I don’t know whether he murdered or not, but the judge said he murdered—what is the chance that he really did murder? A completely different question. Now that also depends on
[Speaker D] the numbers. Huh?
[Rabbi Michael Abraham] Yes. Suppose there are—let’s try to think of an example. Suppose the judge is of quality 0.7. Fine? Now there are a thousand people standing before the judge, and he has to decide about each of them whether he murdered or not. Fine? Now only one of them is a murderer. Someone committed the murder. Okay, I hope I’m building this correctly now; let’s see whether it gets me where I want. So the claim that it’s 0.7 means that the judge will basically determine regarding—we said a thousand?—he will determine regarding three hundred of them that they are murderers. Right? Because he errs in 0.3 of the cases. There are here a thousand innocent people, nine hundred ninety-nine innocent people, let’s say approximately a thousand. Okay? Regarding thirty percent of them he will err. So regarding three hundred of them he will determine that they are murderers. You understand? Even though they didn’t murder. Do you understand?
[Speaker C] No, when someone is put on trial
[Rabbi Michael Abraham] For putting someone on trial, you put one person on trial, you don’t put… never mind, it’s just a way to illustrate the probabilistic quantity. You put him on trial and you don’t know whether he’s the murderer. Maybe you put one of the other thousand on trial? There are a million people in the country. You choose one to put on trial. But you don’t know whether he’s the murderer or not. That’s what the court has to decide. So let’s see what would happen if the court went through all the residents of the country. That’s just in order to do the calculation. It’s the same thing as doing it for one person.
[Speaker E] Okay, suppose so, fine.
[Rabbi Michael Abraham] So that means that regarding three hundred of them it rules that they are murderers even though they didn’t murder, right? Okay, and only one actually murdered. Now I’m asking the question… that’s one question. Now the opposite question: the court ruled that so-and-so murdered. What is the probability that he really murdered? Now you can answer me. One in three hundred. Right? And that’s not the same as 0.7. It’s not seventy percent, it’s a third of a percent. A completely different question. Exactly like with the disease, same idea. Okay? This is very confusing, and so you need… you need a bit of probabilistic skill to do these calculations. And that’s why also Menachem Fisch, when I spoke with him many times about this issue, I told him to be very careful with evidence of this kind, because this is super confusing. If he has evidence of ninety-nine percent, it’s not at all clear that he’s right. Meaning, the question is how rare the case is. The question is what supporting indications he has, and going with evidence that works in ninety-nine percent of cases is worth nothing. Ninety-nine percent is worth nothing. It can throw your whole decision into the trash. It’s worth nothing. Because the percentage of murderers in the population is negligible. True, if they did the investigation properly and it passed through the prosecutor’s office, then they already filtered and say there are at most three suspects in this case; the others weren’t in the area, they have no motive, whatever. So it’s only three. Out of those three, if you have evidence of ninety-nine percent, that’s fine. But check very carefully whether the police and the prosecutor’s office did the filtering properly. Because if not, then all your evidence is worth nothing. If they bring you ninety-nine-percent evidence without doing an initial filtering, it’s worth nothing. Absolutely nothing. It’s like the doctor who sends you for a ninety-nine-percent test. Same thing. In fact there’s an article in Assia that I wrote about this once, where I brought this up in two topics in the Talmud as well, also questions from later authorities (Acharonim) that hover around this issue. Okay, so that’s just regarding the probabilistic calculation. Here I’m basically finishing the main line of thought, and now I’ll just make a few comments. The first comment concerns situations where I do follow the majority. So I think I mentioned this once. And the well-known story about Rabbi Yonatan Eybeschutz, where the priest came to him and said: why don’t you go after us? After all, we Christians are the majority; it says in the Torah, “follow the majority.” And Rabbi Yonatan Eybeschutz told him that he follows the majority where he has a doubt; if he has no doubt, he doesn’t follow the majority. And this is often told as a joke, but it’s completely serious. That’s the correct answer to the question. It’s not a joke; it really is the answer. Because when I find a piece of meat in the market, and there are nine non-kosher butcher shops in town and one kosher shop, but the piece of meat has a premium kosher seal on it, then should I follow the nine non-kosher shops and declare it non-kosher? Obviously not. Why not go after the majority? No—there is a Torah law to follow the majority, but not where I have no doubt. I know this is kosher. If I have a piece and I don’t know what its status is, I’m in doubt, so I follow the majority. But if I’m not in doubt, why should I follow the majority? Okay? So therefore this point of following the majority—and very often they do bring these indications—and we talked about Yitzhak Ben-Aharon, I started with that story, Yitzhak Ben-Aharon who said that the people should be replaced because the people made a mistake in the elections, back in ’77—that statement is, in my view, completely legitimate. So what if the majority said otherwise? If I have a position of my own, why am I supposed to assume the majority was right? I think the majority was mistaken. If I’m in doubt, you can tell me maybe the majority is right—and even there I’m not sure—but maybe you can make the argument that the majority is right. But where I’m not in doubt, why should I go with the majority? I have a clear position of my own. And this has many implications. For example, there’s Rabbi Kook’s interpretation in Ein Ayah on the Talmud in tractate Shabbat 30. The Talmud brings there two cases where someone comes to someone else and says, “Your wife is my wife and your son is my son,” and the reverse, one says this and that and says to him, “Your wife is my wife and you are my son”—I don’t remember exactly who says what to whom. So he says to that fellow: if you want to drink a cup of wine, drink and burst—he drank the cup of wine and killed him, I don’t know, or brushed him off. What do you mean? There is one witness here testifying that there is a problem. How can you ignore such a thing? There is, after all, a chance that he is right, no? So Rabbi Kook’s claim is that in order to be in doubt you need a reason. Meaning, if I have no reason to doubt, then I don’t doubt. If I am in doubt and then you bring me one witness, I’ll go with the one witness. But if I am not in doubt, then what do I care that there is one witness saying something? If it were two witnesses, that’s something else, because two witnesses are already evidence; that’s not just a reason to arouse doubt. But one witness is not that—it’s only a reason to arouse doubt. In order for doubt to be aroused, I need a reason. Meaning, if I have no reason to cast doubt, then I don’t relate to the witness’s words. That’s what the Talmud means when it says: “We stone and burn on the basis of presumptions.” Yes? When someone grows up in a home with two parents, then “one who curses his father and mother,” “one who strikes his father and mother shall surely be put to death”—but maybe it’s not his father? There is a presumption. After all, there are mamzerim; sometimes a child grows up under a father and he’s really not his son.
[Speaker C] And if that were actually the case and it really wasn’t his father, then there would be no death sentence, right? Why—what do you mean, not his father—
[Rabbi Michael Abraham] His, meaning maybe an adoptive father. There’s room to discuss that. And if it’s a mistake, then no, he isn’t his father. So the claim is that since he is presumed to be his son, and we have no reason to assume it isn’t so—even though, again, there are such cases, it’s not something that never happened; after all, there are such cases, maybe even quite a few such cases, of children who are mamzerim—but in order to start doubting that a particular child is a mamzer, you need a reason. That is a halakhic / of Jewish law claim. And since that is so, then even though there is some chance that he is right and that I am a mamzer, I don’t relate to it. And not only do I not relate to it, but even in capital cases they execute me: “one who strikes his father and mother shall surely be put to death,” even though there is some chance that I am not right. And in capital law, as you know, they look for every possibility—“and the congregation shall save”—they look for every possibility to save the person from the death penalty. But not here, because “we stone and burn on the basis of presumptions.”
[Speaker C] Why—I’m going back to the earlier story. I didn’t understand why one witness is not a reason. It’s not a reason for doubt? What do you mean? A doubt means there is a reason; you have one witness, no?
[Rabbi Michael Abraham] No, you can argue about what exactly the threshold is that is sufficient to arouse doubt in a place where there is a clear presumption and the whole world regards you as the child of these two parents. One witness is not a sufficient reason to doubt—that’s the claim. I’m not saying it has to be that way. I’m only saying that assuming that’s what is being said there, the explanation could be this. And the point behind it is certainly a correct point. Where exactly the line turns—you can argue. But it is certainly a correct point that in order to be in doubt you need a reason. It’s like Russell’s celestial teapot, right? Russell was a famous atheist, as you know, in the twentieth century, in the first half of the twentieth century, and he said that his attitude toward traditional testimonies about divine revelation or various things like that is like someone who comes and tells him that around the planet Jupiter there is orbiting a small transparent teapot. And if you ask him why I don’t see it, then obviously it’s because it’s small and transparent; that’s why you don’t see it. Now he asks: am I supposed to treat such a statement as a doubtful proposition? Fifty-fifty—I have no information, the two possibilities are equally weighted a priori, and therefore it’s fifty-fifty, so it’s a doubt. He says no: there is no doubt, because in order to doubt you need a reason. Meaning, and therefore he claims that there is no reason here at all, and claims like that are utterly absurd. Very similar to that statement of the Talmud there that I brought earlier. I think he is mistaken in that context, and that’s another discussion, but the logic of the matter is exactly that logic. In order to be in doubt you need a reason.
[Speaker F] According to that, why does one witness obligate an oath? What? According to that, why does one witness obligate an oath?
[Speaker E] Because it’s a doubt.
[Rabbi Michael Abraham] One witness, for purposes of obligating an oath, is enough to arouse doubt. For purposes of being concerned that I’m a mamzer, apparently it’s not enough. I said, this is what Ari asked earlier: one could have said the opposite—one witness is not a negligible thing. The question is where you place the threshold. As far as Jewish law is concerned, the threshold regarding concern for mamzer status is above one witness. Okay?
[Speaker C] Seemingly one could understand that there, in that story—so is it clear that that’s really the Jewish law? Meaning, that story can seemingly be understood differently, I think. Because the statement there relates somehow to what impression he got from that person or from the way it came. He got an impression? What?
[Rabbi Michael Abraham] There is one witness—what does “he got an impression” mean? Sit in a religious court and let it discuss it. You can’t judge yourself at all; you go to a religious court.
[Speaker C] He didn’t come to a religious court.
[Rabbi Michael Abraham] Exactly, he suddenly came to him. And it’s someone who came in that kind of way with someone. But in the end, you are supposed to bring the matter to a religious court. Why don’t you bring the matter to a religious court? You should be concerned that you’re a mamzer, shouldn’t you? Maybe you’re mistaken; you can’t judge yourself. No, in any case, that’s how he explains it. The principle is certainly correct. I’m saying one can argue about the interpretation of the Talmud, one can argue about the Rabbi, but the principle is a correct principle: in order to be in doubt you need a reason. And that’s that. Therefore one of the implications, for example, is that it is accepted—accepted in the world, I mean in the halakhic world—to think that if I find, say, some dispute among medieval authorities (Rishonim) or a dispute among later authorities (Acharonim) on a certain issue, then the laws of doubt apply here. And I also claim that this is absurd. Because if I myself am in doubt, then the laws of doubt apply. But if there is a dispute among medieval authorities (Rishonim) or later authorities (Acharonim), the question is: what state am I in? And if I have a position of my own, then that is what I should do. Then I am not in doubt, so why should I follow the majority or the rules of halakhic ruling? The rules of halakhic ruling address someone who is in doubt; then he needs the rules, because he has no way to formulate a position on his own. So he has to determine according to rules what is nevertheless to be done. But where I have a position of my own, I do not need rules. So therefore in places where I have a clear halakhic position of my own, then what do I care that later authorities (Acharonim) disagreed about it, or that medieval authorities (Rishonim) disagreed about it? I have a position of my own. If I don’t have a position of my own, I’ll follow the majority, I’ll follow a Torah-level doubt leniently or stringently or whatever it may be. If I’m not in doubt, then why should I treat a Torah-level doubt stringently? I’m not in doubt. The fact that Rashba and Maimonides disagree—what does that have to do with me if I’m not in doubt?
[Speaker C] Yes, but the statement that this is the situation—is that maybe not considered something?
[Rabbi Michael Abraham] Yes, but I’m saying it has no basis. I claim it has no basis. And it’s not… Therefore the point is that this rule of following the majority, like any other rule concerning halakhic ruling in the last example I gave, is a rule that applies only after I’ve decided that I need rules at all. Only where I am in doubt, and then I no longer go by my own decision but by the set of rules that tell me what to do in the absence of a decision. Then I can begin discussing rules, majority, all kinds of things like that. Where I know, then I don’t need rules. By the way, this is, in my opinion, one of the reasons for the many exceptions to rules in the Talmud. Right? “The Jewish law follows Rava against Abaye except for Y-A-L K-G-M,” right? The Jewish law follows Rava in all his disputes with Abaye except for six cases, whose mnemonic is Y-A-L K-G-M. There are other cases where they rule like Abaye. “If he did it, it is ineffective,” “do not form factions,” Maimonides rules like Abaye. And that is not one of those six cases. How can one understand such a thing? So there are all sorts of rule-specialists who split hairs and invent sub-rules: if Abaye and Rava disagreed within a tannaitic dispute, not in a dispute of their own, then maybe the Jewish law can indeed follow Abaye. Maybe. I don’t think that’s so. I think one sees this in general: all these rules always have exceptions. And I think the point is that these rules are there… to tell you what to do when you have no position. Like any rule. If you have no position, then know that you should rule like Rava, except for six cases that were already decided like Abaye. If you have a position, do what you think. So if someone thinks the Jewish law is like Abaye, then he doesn’t need this rule. There are places where they rule like Beit Shammai. In the Talmud they rule like Beit Shammai. And already in the time of the tannaim a heavenly voice came out saying that “the words of Beit Shammai in place of Beit Hillel are of no consequence.” The Jewish law is like Beit Hillel everywhere. And then amoraim rule like Beit Shammai in a few things. How is that possible? With all due respect to rules, rules are for when I don’t know what to do. So the heavenly voice tells me: if you’re looking for a tree to lean on, lean on Beit Hillel. But if I’m not looking for a tree to lean on, I have a position of my own, then what do I care that the Jewish law follows Beit Hillel in their disputes with Beit Shammai? Another example: if I hold like Beit Shammai, personally, then I certainly would do that even if Beit Shammai and Beit Hillel had disagreed about it, right? Because that is my position. Now Beit Shammai also support what I am saying. So now I’m supposed not to do what I think just because Beit Shammai support what I’m saying? Because Beit Hillel oppose what I’m saying? Beit Hillel oppose it—so what?
[Speaker C] “The words of Beit Shammai in place of Beit—”
[Rabbi Michael Abraham] “Hillel are of no consequence.”
[Speaker C] But my own words in place of Beit Hillel are not of no consequence. Meaning, the fact that Beit Shammai are with me doesn’t worsen my situation. So if I have a position of my own, then I’ll do what I think. If I don’t have a position of my own, then they establish a rule of what to do, and the rule is that we go with Beit Hillel.
[Rabbi Michael Abraham] What about tradition? What is tradition? Where does that tradition come from? From the fact that they thought that way; they reached a conclusion. Tradition is perhaps the mode of thinking, but the laws themselves are what crystallized in Beit Shammai and what crystallized in Beit Hillel. I see no reason to assume anything else. In any case, at some point the tradition split. There were some disagreements that split the tradition, right? It only pushes the doubt backward. Where did it split for the first time?
[Speaker C] Yes, but still, each of them claimed—each really tried to reconstruct it, and each one’s claim is tradition. No, it’s not just a symbol, but practically—
[Rabbi Michael Abraham] In the end, the split is real. But at the root there is an argument in reasoning.
[Speaker C] Or one of them is tradition and the other is not tradition, or both are tradition. So what do I care if it’s because both are tradition?
[Rabbi Michael Abraham] Of course I can. So those two who are the fathers of the tradition, they are the disputants, right? I don’t know who they are, but never mind. Now I agree with that one, so I do like that one.
[Speaker C] But maybe one of them isn’t tradition.
[Rabbi Michael Abraham] What do I care if one of them isn’t tradition?
[Speaker C] It’s not interesting.
[Rabbi Michael Abraham] I’m talking now about what the truth is, not what tradition says. One of two disputants can also have been mistaken about the tradition.
[Speaker C] Tradition is what the truth is. Do you mean tradition from Sinai?
[Rabbi Michael Abraham] No, but where would you invent such a thing from? Is it a law given to Moses at Sinai? Maimonides says that there is never a dispute about a law given to Moses at Sinai. So it can’t even be that.
[Speaker C] So what is the dispute?
[Rabbi Michael Abraham] Things that were created after Mount Sinai. What do you mean? There was no tradition. Most things were not received from Sinai.
[Speaker C] What? There’s no such thing.
[Rabbi Michael Abraham] The overwhelming majority of the Torah we have in hand—the Torah-level law, never mind rabbinic law; Torah-level law—they didn’t dream of it at Sinai. These are things that developed throughout the generations. That’s obvious.
[Speaker E] Wait, I want to understand for a second what you’re saying. Suppose the Talmud says the Y-A-L K-G-M rule; the Talmud brings it.
[Rabbi Michael Abraham] Of course, there are many rules in the Talmud.
[Speaker E] Wait, so what you’re saying is that Y-A-L K-G-M—let’s say, how shall I put it—the Talmud wasn’t precise and there are other cases in the Talmud like that; that’s one possibility. Or the second possibility is that the Talmud really said Y-A-L K-G-M, but that’s because the Talmud was in doubt. But Maimonides, eight hundred years later…
[Rabbi Michael Abraham] The Talmud wasn’t in doubt; the Talmud decided. Am I bound by the Talmud’s decisions? All halakhic decisors are bound by the Talmud’s decisions; that is binding. Now, Y-A-L K-G-M—I claim that this is not a decision, it is guidance. That’s my claim. It is guidance that says: listen, if you have a position, do what you understand. If you don’t have a position and you want something to rely on, then rely on Rava except for six things that were ruled like Abaye. Where there is an explicit ruling like Rava, then they’ll also go with Rava regardless of my own position because the Talmud—
[Speaker E] decided.
[Rabbi Michael Abraham] In those six things they will go like Abaye regardless of my own position because the Talmud decided. I’m talking about situations where it’s not clear in the Talmud what the Talmud decided, but ostensibly this rule itself is a decision. No—the rule is guidance, not a decision; it gives direction. And Rabbi Elchanan writes somewhere—this is a question of how one views these rules: whether one views them as summary rules or as directive rules. If they are summaries, then there cannot be an exception, because they summarized. A rule that I stated just now—even there there can be exceptions. I don’t know whether all rules are like that, but all rules are presumed to be like that unless it becomes clear otherwise. If you show me that they aren’t—so for example, Rabbi Elchanan writes somewhere, in Bava Batra in Kovetz Shiurim, that there is in Eruvin, page 29-something or 34, a whole series of rules of halakhic ruling between tannaim: Rabbi Yosei and Rabbi Shimon, the law is like this one. So he brings there that in one place the Talmud asks—I don’t remember exactly which tannaim are involved—Rabbi Yosei and Rabbi Shimon, the Jewish law follows Rabbi Yosei; Rabbi Shimon and Rabbi Yehuda, the Jewish law follows Rabbi Shimon; therefore Rabbi Yosei and Rabbi Yehuda, the Jewish law follows Rabbi Yosei.
[Speaker D] A transitive relation. Yes.
[Rabbi Michael Abraham] So Rabbi Elchanan says: if that’s so, then you see that these rules are substantive rules. Because if these rules were summary rules, then there is no necessity that it should be so. That’s just how it came out. It’s not because we decided a priori that Rabbi Yosei is wiser than Rabbi Shimon or more correct—never mind wiser—or that Rabbi Shimon is more correct than Rabbi Yehuda, but simply that’s how it came out for us: in all the disputes it turned out that for some reason the Jewish law follows Rabbi Yosei. But then there is no necessity that between Rabbi Yosei and Rabbi Yehuda it should also be so. If you make such an inference from transitivity, you are essentially assuming that this is a substantive relation, not some accidental collection or summary of what happened to come out. That’s his claim. One could say that it is a summary that serves as an indication that apparently it isn’t accidental. One could say that.
[Speaker C] According to this, all the exceptions need to have a substantive explanation.
[Rabbi Michael Abraham] What? According to this, all the exceptions need to have a substantive explanation. Right. Either there is another Talmudic passage that goes that way, or I think that way—yes, indeed.
[Speaker G] Or that it’s an exception. What? An exception obviously needs a substantive explanation. It has to.
[Rabbi Michael Abraham] After all—
[Speaker C] If it’s an accidental summary, you don’t need a solution.
[Rabbi Michael Abraham] No, I do need one, because you still have to explain to me why you rule that way.
[Speaker C] Because from the summary I tell you that in these cases it didn’t come out that way.
[Rabbi Michael Abraham] It didn’t come out that way, and therefore what? But explain to me why you nevertheless decided to rule that way. Fine, you don’t have to follow Abaye only in Y-A-L K-G-M, but you still need to explain to me why you decided to go with Abaye. Here I do need some explanation.
[Speaker C] But the explanation needs to be substantive.
[Rabbi Michael Abraham] What’s the difference between substantive and non-substantive?
[Speaker C] Meaning, if you say the rule is that at the end of the day Abaye is more correct than Rava, then in all the exceptional cases you have to explain: despite the fact that Abaye is more correct than Rava, why there it’s like Rava.
[Rabbi Michael Abraham] I said even more than that: if the rule were that Abaye is more correct, or Rava is more correct than Abaye, it could be that the medieval authorities (Rishonim) would not have allowed themselves to deviate at all. Because the Talmud itself says that Rava is more correct. Now true, that isn’t an unequivocal decision. Here it may be that even with an explanation I would not have allowed myself to deviate. If you understood it as a directive rule only because Rava is wiser, because that’s how it came out, then fine—if I have a reason, I’ll deviate.
[Speaker C] So then it’s not a directive rule, it only summarizes. If that’s how it came out, that’s how it came out.
[Rabbi Michael Abraham] No, there is the possibility of saying that it is an a priori rule stating that Rava is always right, and there are six cases where Abaye is right, but in principle Rava is wiser—let’s put it that way. Okay? If that were so, I’m not sure the medieval authorities (Rishonim) would allow themselves to deviate. Because at the end of the day, when there is a dispute between Abaye and Rava, apparently Rava is right, and if I think Abaye is right, apparently I too am mistaken, because Rava was wiser than I am as well. Okay? But if I say no, there is no difference in wisdom between Abaye and Rava; that’s not the point. The point is that we know there are six cases where the Jewish law follows Abaye, and generally it turned out that the Jewish law follows Rava. It is a summary of what happened—not a one-for-one summary, but a guiding rule: generally speaking, to the extent I can tell you something sweeping, the Jewish law follows Rava except for six cases where I can tell you it was ruled like Abaye. Then I say: I can deviate.
[Speaker C] And the question is what about the cases that aren’t brought in those six cases.
[Rabbi Michael Abraham] There the Talmud said nothing. If the Talmud didn’t say that the Jewish law follows Rava and didn’t say that the Jewish law follows Abaye, there is only the rule that except for Y-A-L K-G-M the Jewish law is like Rava. Fine, but in a case where it seems to me like Abaye, then that’s okay, because I don’t need the rule; I have a position of my own. Okay, so that’s the first comment. The second comment, since I mentioned it, is a majority of opinions that are dependent on each other. I talked about this when we discussed, I think, the more educated and less educated regarding faith / belief, with the antics of the atheist websites. And I said there that when there is a majority of opinions that are dependent on each other, there is no point counting them as separate opinions. The source for this is a Talmud in Sanhedrin 36, I think—someone asked me about it, which is why it reminded me of this. The Talmud there: a father and son—well, father and son are disqualified altogether; a father and son cannot sit in judgment. But a rabbi and his student are not counted as two, but as one, in a religious court. If they sat in a religious court, and the rabbi and his students sat there, seated in rows like that, in the end they would ask all of them their opinions, but a rabbi and his student would be counted as one and not as two. But in matters of ritual impurity and purity they are counted as two. What’s the difference? The explanation here is that in a religious court a rabbi and his student cannot be counted as two independent opinions, because the student is basically just saying what his rabbi says. So it’s not that there are two opinions here in favor of a ruling; rather, he is carried along by what the rabbi says. In impurity and purity—that’s the novelty—but in impurity and purity the point is that this is something that doesn’t require a religious court; it’s matters of prohibition and permission. In prohibition and permission, even one person is enough to decide. But even in prohibition and permission, when there is a dispute, we still need to count according to the majority. For example, when there is a dispute among doctors regarding desecrating the Sabbath on Yom Kippur, when there is a sick person and the doctors disagree whether he needs to eat on Yom Kippur or not, then we go after the majority. Even though this is prohibition and permission, it’s not a religious court; there is no need here for a court ruling. So in prohibition and permission one doctor is enough, or a panel of one, to decide—or a rabbi in the case of ordinary prohibition and permission. But if in fact several were asked and there are differences of opinion, then even there one follows the majority. But there a rabbi and his student are counted as two. Very interesting. In a religious court, no, but there yes. Why indeed? That needs thought. One thing: this is also ruled as Jewish law in both Maimonides and the Shulchan Arukh. I also brought the example of manuscripts: when there are differences in the wording of manuscripts, then one follows the majority. And if there is a manuscript copied from another manuscript, then that is not another textual witness. Meaning, it’s not two textual witnesses against one. Obviously—if you copied from that one, then you’re the same as it, unless there was an error in this copying too. So it certainly isn’t counted as two manuscripts. And there too this is some kind of majority, and although it’s interesting because there too it’s not a religious court, and ostensibly one could have compared it דווקא to impurity and purity here, something that doesn’t require a religious court, where there would indeed be room to count them—but there it’s apparently too blatant. Because the issue of rabbi and student is not that the rabbi must go like the student, only a concern that the student will go like the rabbi, just a reasonable concern that he will go like the rabbi. And indeed they make distinctions here, they draw distinctions. The question is whether it is a student, and this appears as Jewish law also in the halakhic decisors, in the Shulchan Arukh and in Maimonides: did the student receive this specific ruling from his rabbi, or did he receive only the modes of thought from his rabbi, but then he applies them to this ruling, in which case there is also a chance he will deviate? So there are different situations here. But where it is clear that the student is copying after his rabbi, there is no point counting him as an additional opinion. And since that is the case with textual witnesses, because there one simply copied from the other, then clearly there is no point counting that. And that would be true even in impurity and purity. So if there is someone who automatically says what his rabbi says, then he will not be counted even in impurity and purity as an additional opinion. I once brought the famous and dear-to-my-heart Magen Avraham, that it is permissible to say things in the name of a great person so that people will accept them from you. Yes, this is from the Talmud. That you can say—lie—and say that Rabbi Moshe Feinstein said this, so that they will accept from me the Jewish law that I am saying. Which on the face of it is a very strange thing. How can that be? Am I causing someone to desecrate the Sabbath? I tell him Rabbi Moshe Feinstein permits this, and in truth that’s not so; I think it is permitted—maybe I’m mistaken, I’m not Rabbi Moshe Feinstein. Maybe I’m mistaken. How am I allowed to cause someone to desecrate the Sabbath in such a situation? And what I said is that the only way to understand the Magen Avraham, in my opinion, is only because the assumption is that even if Rabbi Moshe Feinstein says something, I don’t automatically accept it. Only because of that am I allowed to lie to him and say that Rabbi Moshe Feinstein said it. If so, then why do I need to say it? What was the point of the whole story if he won’t accept it? The answer is: so that he will seriously consider what I’m saying. My feeling is that he doesn’t give serious weight to what I’m saying; he isn’t even willing to consider it. If I say it in the name of Rabbi Moshe Feinstein, he will consider it. He won’t accept it automatically, but he will consider it. That’s exactly the intermediate state. Because if someone automatically accepts what Rabbi Moshe Feinstein says, then he isn’t another opinion. He’s just saying whatever—he’s bound to him. Whatever Rabbi Moshe Feinstein says, he says too. And the Beit Yosef writes that the Rosh and the Tur, in a dispute among halakhic decisors, are counted as one.
[Speaker B] Meaning, if you know that if you say Rabbi Moshe Feinstein he’ll accept it because he accepts him, then it’s forbidden for you, right.
[Rabbi Michael Abraham] I explained that way the famous forgery of the Jerusalem Talmud. At the beginning of the twentieth century there was someone who forged a Jerusalem Talmud on Kodashim. He said he found a manuscript of the Jerusalem Talmud on Kodashim, and he inserted his Torah novellae there. Fine? So they came out strongly against him; there were major controversies; many later authorities (Acharonim) agreed and many later authorities (Acharonim) disagreed. There is an entire book by Klei Chemdah on this forgery to prove that it is a forgery. In the end, the legend says that the Rogatchover proved that it was a forgery. He had some kind of rule that in every chapter of the Jerusalem Talmud there appears a sage who did not appear in the previous chapters. And that is how he checked, and he saw that it didn’t hold there, and he decided it was a forgery. I don’t know—that’s a legend, and I haven’t checked even that—but I think that even if it held in the other chapters, I still wouldn’t rely on that in order to reject the Jerusalem Talmud on Kodashim, and in Kodashim it wasn’t that way. In any case, I’m saying the question was: why were you so angry at the forger? After all, he was simply following the Magen Avraham; he said things in the name of a great person so that people would accept them from him. What’s the problem? He did something permitted. What’s the problem? He said his own ideas in the name of amoraim so that people would accept them from him. The answer is no: because if you say it in the name of the Talmud, the Talmud we are obligated to accept. The Talmud is not just something we’ll seriously consider; the Talmud has authority. That you cannot do. Meaning, you can’t forge and say things in the name of someone who has authority, in the name of the Sanhedrin. You cannot lie and say that the Sanhedrin ruled this. No. You can say that a great person said it so that I will seriously consider his position too. Okay?
[Speaker D] What? In the book Besamim Rosh, is it okay there? What? There’s the book called—some call it Kizvei HaRosh, actually Besamim Rosh, not Kizvei HaRosh, yes, some do call it Kizvei HaRosh—so there is still a dispute there whether it is…
[Rabbi Michael Abraham] No, no, today that is no longer in dispute. It is a forgery, with the commentary Kisei Deharsena, and the Kisei Deharsena also forged Besamim Rosh.
[Speaker G] In short, people still use that book.
[Rabbi Michael Abraham] People use it, those who don’t know it’s a forgery. But today it is clear that it’s a forgery. Fine, in any case, what do I care even to use it? Look at it and see whether it is correct, that’s all I said. For me it is totally irrelevant whether it is a forgery or not. Look at the considerations and see whether you agree with them. What difference does it make whether the Rosh said it or didn’t say it?
[Speaker C] But when you are summing up the opinions of later authorities (Acharonim), that’s a problem.
[Rabbi Michael Abraham] Yes, if you take them into account numerically, if you are counting, then it matters—it’s a problem for whoever counts. In any case, the point is that a majority—this is my second comment—when I speak about a majority, the majority needs to be independent in order for it to be a meaningful majority. That’s the second comment. A third comment: there are situations of majority—and this is really fascinating—situations of majority in which the majority is created from different reasons. In the Shulchan Arukh there is a reference to this issue in section 25, paragraph 2, in the Rema. He brings there that if there is a majority in a religious court for different reasons, it is still a majority. And the Shakh goes on at great length there to explain that there are cases where it isn’t, and he distinguishes among different cases; it’s not so simple. What are the arguments on either side? There is what is called the doctrinal paradox. Think about a situation where there is a dispute among three judges regarding a contract between Reuven and Shimon. Reuven claims that he had a contract with Shimon and that Shimon violated it. Okay? And Shimon claims that he did not violate it, or alternatively that the contract did not obligate him to do that at all—it is permitted to make such alternative claims. Okay? Now they come before three judges. The three judges make the following decision. Judge A says: the contract did require it, and Shimon fulfilled what the contract said, therefore Shimon is okay. Judge B says: the contract doesn’t require it at all; Shimon did do it, true, but the contract doesn’t require it at all, therefore Shimon is okay. Judge C says: the contract does require it, and Shimon didn’t do it, therefore Shimon is not okay. So there is a majority saying that Shimon is okay, right? Therefore Shimon comes out okay in the end. Now count the opinions regarding the reasons. The question whether the contract required it—the answer is yes; there are two who said that the contract required it. Did Shimon violate the contract? Yes, since two out of the three said that Shimon violated the contract. So in fact the final ruling should be that Shimon is not okay. Meaning, if you count the majority on each of the reasons separately—and that’s altogether what logic says, after all you want to know: first, there are really two questions before the court here: first, did the contract require this thing? and second, did Shimon violate it? Right? Now, if we check the truth on each of the questions, on each of the questions there is a given majority and minority, and then the result will be the opposite of what comes out if you go by the majority of the bottom lines. So what do you do? In the Shulchan Arukh it is ruled that even if it is for different reasons, one goes after the majority of the bottom lines. And that is a very interesting claim. That is also accepted in court, by the way. Not only that; just now I ran across such a case, really such a case, in the Supreme Court. There was a story—just as an anecdote—you know about the kiddushin that we annulled, the second kiddushin, of the two that we annulled. There was some woman there—never mind—whose husband refused to give a divorce document. He was abroad; somehow he found this, surfed the internet and suddenly found it. It turns out there had been legal proceedings between this woman and her husband in 2010, 2011. The claim was that she kidnapped his daughter, their daughter, because this was already around the time of the divorce. She kidnapped his daughter and he sued her, and an extradition order was issued against her through Interpol, to extradite her to the United States. Again, they didn’t determine that she had kidnapped; they determined that she had to be extradited in order to adjudicate whether there was kidnapping or not, right? And that was enough. She was not willing to come and litigate, so the extradition was so that she would come litigate. Okay? And she did not agree to litigate. Now the question was whether to extradite her, or to return the child or not return the child, or whether to go—I don’t know. What happened there was that it went through three judicial instances. Very interesting, and the prosecutor’s office was involved, and even after the final decision the prosecutor’s office argued that this was not the right decision, and in the end it somehow got stuck; I don’t know what became of it. Meaning, it didn’t continue, and I don’t know for what reason. The claim of the Supreme Court—and I’m now coming to the punchline—when one kidnaps a child or does not return a child, whether by omission or by action, those are two possibilities but it doesn’t matter, in principle the child has to be returned; this is the Hague Convention. But there are three reservations; for our purposes two are enough. One is consent and the second is acquiescence. If there are some indications that the husband—or never mind, the side from whom the child was taken—had consented beforehand, even though now he claims he didn’t consent, but there are indications that then he in fact did consent, then it is not kidnapping and the child does not have to be returned. The second reservation or exception is if there are some indications that he acquiesced after the fact, after it had already happened. Consent is before; acquiescence is after. That he already came to terms with it and says, okay, then let the child stay with you. Okay? These are two exceptions, this appears in the case law and it is clear—these are two exceptions. There is also the child’s welfare, a third exception, but that is less important for our purposes. Now what happened? In the first two instances it was determined that she had to return the child, return the girl, or go to the United States and litigate there. Okay? The Supreme Court reversed the decision of the first two instances, but the split of opinions there was as follows: Arbel, Meltzer, and Vogelman. Okay? Arbel argued that there had been consent, and therefore the girl need not be returned. Meltzer argued that there had been acquiescence, and therefore the girl need not be returned. And Vogelman argued that there was neither acquiescence nor consent. All agree that there was kidnapping; she was in the wrong. The whole question now is, fine, but even if you kidnapped, it’s not certain that you have to return the child. So in the bottom line, two said the child should not be returned and one said yes. Therefore the ruling was not to return the child. But if you check it in slices, regarding acquiescence there are two against one, and regarding consent there are also two against one. In principle they should have returned the child even according to the Supreme Court’s own ruling, if one goes after the reasons—if one breaks the ruling down into reasons and discusses reason by reason. But no: both in court and in Jewish law—again, with the qualifications of the Shakh, but at least according to the plain sense of the Shulchan Arukh in the Rema—what determines things is the bottom line. And that is a very interesting question: why indeed? Because if we are looking for the truth, then ostensibly we should discuss the answers to the factual questions, and in the answers to the factual questions the majority determines the truth, as the distinction taught us. So if the majority is an indication of truth, then the truth is that there was no acquiescence and no consent, or that there was a contract and there was no violation. Right? So there is something problematic here. So here this is apparently a matter of simplicity or efficiency or something like that, and not of truth, I assume.
[Speaker C] Because otherwise even the very notion of majority is really a matter of—
[Rabbi Michael Abraham] No, we talked about that in previous classes; I don’t think that’s correct. But this is the difference between a democratic majority and a majority in a religious court. And here I really close the circle: all the qualifications I’ve just made are not relevant to a democratic majority. In a democratic majority, whether it comes from different reasons or a rabbi and his student or things like that—it doesn’t matter.
[Speaker B] At all—what does he want?
[Rabbi Michael Abraham] Exactly, because the whole question is what people want, what most people want. Why should you care whether I want what my rabbi wants, or whether I’m influenced by him, or by other reasons, and that’s what I want? What difference does it make? That’s what I want. We need to do what most of the public wants. Therefore all these reservations apply only to a majority in a religious court, not a majority that… Now one last point connected to this—I’m already skipping something—but one last point connected to the majority issue in Zadorov’s case. There was a very interesting situation in the proceedings there, in the last hearing. There were several hearings there. There it was two against one on whether to accept the appeal. Meaning, one judge said to accept the appeal, and two were against it. Then the claim came up that if one judge accepts the appeal, that means there is reasonable doubt here. And if there is reasonable doubt, then basically he should have to be released in any case. There was that debate. Now my view—my unlearned view—is that it depends on why the one judge argued that the appeal should be accepted. If he says, “Zadorov did not commit the murder,” then that is reasonable doubt, and he should be released. If the Supreme Court says that he did not commit the murder, that is certainly enough to constitute reasonable doubt. But if he says, “I have reasonable doubt,” then no. Because there is a dispute over whether this is reasonable doubt or not, and in that dispute there are two against one. And the dispute is about the very question of whether there is reasonable doubt here or not, and on that the majority decides. Okay? I think that’s also the structure. Good.