Doubt and Statistics – Lecture 13
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
- Statistical evidence versus eyewitness testimony and the prisoners example
- The first explanation: a majority present before us, randomness, and Eliav’s comment about probability
- The second explanation: majority in monetary law, majority of a court, and Rabbi Shimon Shkop on a majority that depends on choice
- The presumption of fitness, reasonable claims, and the need for positive evidence against the defendant
- The third explanation: a doubt where a prohibition is fixed, ignorance versus information, and the certain conviction of an innocent person
- Summary of the distinctions, legal considerations beyond statistics, and Pappino as an intuition without an explanation
Summary
General Overview
The text presents the difficulty of convicting on the basis of statistical evidence alone through Pappino’s example of one hundred prisoners, ninety-nine of whom attacked a prison guard. It contrasts this with the fact that eyewitness testimony is still sufficient for conviction even though its reliability is not necessarily higher than that of circumstantial evidence. Three complementary lines of explanation are proposed: an initial distinction between a majority present before us and a majority not present before us, together with the claim that in the prisoners case there is no “information about the person” but only a random datum; Eliav’s corrective comment that in the prisoners case there really is plain probability, like balls in a barrel; and two further explanations that are mainly legal-halakhic: the defendant’s right to make a reasonable claim and to be supported by a presumption of fitness, and a distinction between a doubt that has a positive basis in the facts and a doubt that stems from ignorance, together with a policy consideration of avoiding a mechanism that certainly leads to convicting an innocent person. In the end it is emphasized that even legal systems that do not rest on a scriptural decree sometimes operate on legal considerations that do not align directly with rankings of statistical certainty, and sometimes we are left with a strong intuition without full rationalization, as described in Pappino, before moving on to Enokh’s article.
Statistical Evidence versus Eyewitness Testimony and the Prisoners Example
Pappino’s example describes one hundred prisoners in a prison yard, ninety-nine of whom pounced on a guard, and asks whether every prisoner brought to trial can be convicted because the chance that he participated is ninety-nine percent. The text states that in the legal world it is accepted that one does not convict that way, and contrasts this with the accepted view that conviction can be based on eyewitness testimony even though the reliability of eyewitness testimony is not necessarily greater than circumstantial evidence and may sometimes even be lower. The text emphasizes that circumstantial evidence is not a synonym for weak evidence and seeks an explanation for the gap between not convicting on the basis of statistics and convicting on the basis of eyewitness testimony.
The First Explanation: A Majority Present Before Us, Randomness, and Eliav’s Comment about Probability
The first explanation proposes seeing the prisoners example as a case of a majority present before us, because there is no “law of nature” saying that when there are one hundred prisoners in a yard, ninety-nine percent of them attack a guard. So this is a random feature of the situation, not general information about the nature of the world. The claim is that a conviction based on this kind of statistical evidence has no positive evidence about the individual on trial, but only an assumption about a distribution, similar to the assumption in the case of a piece of meat separated from one of nine stores, where there is no knowledge of how the piece separated, so this is not knowledge but conjecture. Eliav comments, and the text accepts, that in the prisoners case, when it is known that ninety-nine are guilty and one is innocent, and one is chosen at random, a real probability is obtained, like calculating balls in a barrel. Therefore the claim that a majority present before us is “not really probability” is incorrect here. The text suspends the first explanation and suggests that the relevant distinction going forward will be legal rather than probabilistic.
The Second Explanation: Majority in Monetary Law, Majority of a Court, and Rabbi Shimon Shkop on a Majority That Depends on Choice
The difficulty raised by the medieval authorities (Rishonim) in Tosafot on Bava Kamma 27 and Sanhedrin 3 is brought here regarding the contradiction between the ruling that in monetary matters we do not follow the majority in the passage about an ox sold and later found to be goring, and the Jewish law that we do follow the majority of judges in monetary cases by virtue of “incline after the many.” Two directions are presented: either the majority in court is a special case against the general rule that in monetary matters we do not follow the majority, or the opposite: the case of the majority for plowing is the exception, while generally we do follow the majority in monetary matters. One explanation attributes the uniqueness of a court to the fact that the minority is halakhically nullified into the view of the majority, so that the ruling is treated as the unified ruling of all three judges. Therefore this resembles “nullification in a majority” rather than following the majority in the ordinary sense. Consequences are noted, such as all three judges signing the ruling, and the prohibition of “Do not go about as a talebearer among your people” regarding revealing the minority position to a litigant.
The text also focuses on the opposite direction through Rabbi Shimon Shkop, according to whom when the majority depends on human decision and this is a normative and reasonable action, one does not extract money from someone in possession merely because the claimant relies on a minority possibility. The difference is explained between a “reasonable” claim of belonging to the minority, such as “I sold it for slaughter,” and a suspicious claim such as “I repaid within the term,” which is seen as contrary to “human nature” and is therefore called a presumption rather than a majority, where the majority is a result of the presumption and not the other way around. From this an analogy is drawn to the prisoners example: the prisoner can reasonably claim that he chose not to participate, because the decision whether to participate in the crime is a human choice, and it is known that at least one person in fact did not participate. Therefore there is no legal justification to convict without direct positive evidence against him.
The Presumption of Fitness, Reasonable Claims, and the Need for Positive Evidence against the Defendant
The text is framed as a legal position according to which, in order to convict, some kind of positive evidence relating to the defendant himself is required, and ninety-nine percent statistical persuasion is not enough when the defendant enjoys a presumption of fitness and makes a reasonable claim in his defense. The text distinguishes between a claim about behavior that is the result of a decision and a description supposedly based on “human nature,” and suggests that even if a large majority of people react violently to provocation, this still does not completely rule out the reasonable possibility of self-restraint, though there the claim may be weaker than in the prisoners example. The text also addresses the question whether a prior criminal record changes the picture, and suggests that the presumption of fitness and the presumption of innocence are not canceled, even if such things may affect statistical intuitions.
The Third Explanation: A Doubt Where a Prohibition Is Fixed, Ignorance versus Information, and the Certain Conviction of an Innocent Person
A halakhic distinction is brought between a doubt involving “one piece out of two pieces,” where a prohibition is fixed, and a doubt involving “one piece,” where in the first case there is positive information that at least one of the pieces is forbidden, only it is not known which one, while in the second case the doubt arises from lack of information about the piece itself. The text illustrates this through a bet on a fair coin, where the fifty-fifty ratio stems from information about the distribution, as against a bet on an unknown coin, where the fifty-fifty ratio is a product of ignorance and symmetry considerations rather than knowledge of a distribution. This distinction is applied to the prisoners example by saying that here there is positive knowledge that there is one innocent person, and therefore a consistent policy of statistical conviction would certainly lead to the conviction of one innocent person, unlike eyewitness testimony that is ninety-five percent reliable, which does not guarantee that in the present case there is an innocent person who has been convicted. The text connects this to a formulation attributed to Maimonides, according to which it is preferable that many offenders go free than that one innocent person be convicted, and presents this as an indication that in the prisoners example there is no direct evidence about the person himself.
Summary of the Distinctions, Legal Considerations beyond Statistics, and Pappino as Intuition without Explanation
In the end the text accepts Eliav’s comment that in the prisoners example, although it is a majority present before us, it is still a real probability, and that the weight of the difference from eyewitness testimony is not on the statistical plane but on the legal plane. The text argues that even evidence that is statistically “better” will not necessarily be accepted, because the legal system uses legal considerations that are not exhausted by quantitative certainty. This parallels, to some extent, the way that in Jewish law people sometimes explain rules as a “scriptural decree,” but it turns out that general law also works that way. The text points to situations in which one can successfully provide a legal rationalization for the distinction, and also to situations in which a strong intuition remains without a full explanation. It cites Pappino as someone who, after rejecting the explanations, remains with “this requires further analysis” and yet is still certain that this is how one should proceed. The text concludes with a reference to column 227 on the author’s website, a reference to Pappino’s article and its translation on the Alaxon website, and an announcement that the discussion will move on to the article by Enokh and his colleagues.
Full Transcript
[Rabbi Michael Abraham] Good. Last time we discussed statistical evidence. The example I brought there, not mine, from an article by Pappino, is an example of a hundred prisoners attacking a guard in the prison yard. Ninety-nine—in other words, there are a hundred prisoners in the yard, ninety-nine of them attacked the guard. The question is whether you can convict every prisoner who comes before us because the chance that he participated in that attack is ninety-nine percent. So in the legal world, the accepted answer is no. On the other hand, eyewitness testimony is still enough to convict a person, even though the reliability of eyewitness testimony does not necessarily exceed that of circumstantial evidence. On the contrary, sometimes it is even weaker. And I talked about the fact that circumstantial evidence is not a synonym for weak evidence. So the question is: why? Why, in the first case, do we not convict, while in the second case we do? Now I said that I would offer three explanations for this, which maybe in a certain sense complement one another. Maybe they are not identical, but it seems to me that they give some kind of shared idea. The first explanation that I proposed is that the prisoners in prison are really a case of a majority present before us, because there is no law of nature that says that people of this kind attack guards, right? It is not something connected to the nature of the world. I have no indication that this person actually has that sort of character or that he participated in this act. Rather, this is statistical evidence built from random data. In this case there happened to be a hundred prisoners, ninety-nine of whom attacked. In a different yard there might be a hundred prisoners of whom only twenty attacked, or none at all. In other words, there is no law of nature that when there are a hundred prisoners in a yard, ninety-nine percent of them always attack guards. So this is not general information about the nature of the world; it is a random fact true of this situation. In that sense, this is similar to a majority present before us and not to a majority not present before us. And the claim—and we saw that the difference between a majority present before us and a majority not present before us is that the latter is a generalization from a sample, whereas the former is not a majority built the way a law of nature is built, but is really some kind of a priori assumption about reality. It is not a kind of knowledge about the world or the result of a sample-based generalization, though that is still how scientific knowledge accumulates. Therefore I wanted to argue that when people say you cannot convict on the basis of a decision based on statistical evidence, what they mean is that you cannot decide on the basis of a majority present before us. Because a majority present before us is not something you really know about the nature of the world. As for the person standing before you, you really have no positive evidence that he took part in the crime. Rather, from the accumulation of circumstances you estimate that the distribution among all the prisoners is the same, so you assume that the probability here is ninety-nine percent. But that is not really the result of observation, not the result of knowledge about the world, but rather some conjecture of yours.
Now here I want to note—I think Eliav commented on this in the previous lecture—and in the end I do accept your comment. Meaning, I also wanted to argue, after saying all this, that what I’ve said so far has a measure of truth in it, but we’ll see it in the next explanations. But what I wanted to argue was that the explanation I proposed here is really an explanation connected to reality itself. It’s not a legal explanation, it’s a statistical or factual explanation. Because my claim was that I genuinely have no information about this person that he is a criminal, just as I have no information about the piece of meat I find in the street that it is kosher. I only know—or rather, sorry, assume—that it is most likely kosher, because I assume the separation from all the stores is equally weighted. But that’s an assumption, not knowledge, and therefore the claim was—not only that there is some legal logic in not relying on such evidence but on other evidence, say, evidence that brings specific information about the defendant, not complete information, not certain knowledge, and yet I still have knowledge about the defendant that he murdered or participated in the crime. Here I don’t. I merely have a conjecture as to how the guilt is distributed, and according to that conjecture it comes out that this piece is ninety percent kosher or that this person probably participated in the attack.
Now at this point I think that what I said is not correct. In other words, I accept Eliav’s comment, because basically—let’s formulate it this way—suppose I place these hundred prisoners who were in the yard outside the courtroom, arrange them all in threes, close my eyes, mix them up, and take one out. What is the probability that this one is a criminal? Answer: ninety-nine percent. And that is probability, that is a probability. Because once it is clear—after all—we know there are ninety-nine guilty people here and one innocent one. That I know. Now I also mixed them all up and chose one at random. So here this is not similar to the separation of the piece of meat from the stores. In the case of the piece of meat separating from the stores, I truly don’t know how pieces of meat fall or separate from stores, and what the chances are that a piece gets lost from store A, B, or C—whether that chance is the same or not. I have no way to test my thesis that this piece is kosher with ninety percent probability or non-kosher with ninety percent probability. All those things are true regarding pieces of meat, in the case of the nine stores. But in this case this is plain statistical calculation, right? It’s obvious. Since I have a hundred people and among them one—like balls, right, the classic examples in combinatorics—you have a hundred balls in a barrel, one of them green, ninety-nine red. You mix them well, put in your hand, close your eyes, and pull one out. What is the chance that it’s green? The chance is basically one in a hundred, right? In other words, that really is probability, not conjecture. It’s probability, because in this case I really do know how the balls are drawn from this barrel. In the case of the stores I have no information about how the piece of meat separated from the store and got lost, and what the chance was that it separated from store A, B, or C. But here I have a hundred balls and I choose randomly; obviously the chance of each one being drawn is equal. So it is true that this is not a generalization from a sample, and in that sense it is a majority present before us, but the logic that says a majority present before us is not really probability—that is not true here. In this case, although it is a majority present before us, it really is probability. That is basically Eliav’s point, and I accept it. I think this really is a probabilistic statement, and it could still be that this distinction can explain things—but then it would be a legal explanation, not a probabilistic one. But for that we need to get into the next two explanations. I started that last time.
[Speaker B] Rabbi, Rabbi, just a second? Yes, yes. Maybe one could say—I want to retract the claim I made—and maybe one could say that even in the case of the prisoners it’s not entirely probability, but there is an element of assumption, because maybe one could say that the chance that all the prisoners would kill someone is not the same chance. One is a very wicked prisoner, one less so, one less so.
[Rabbi Michael Abraham] That doesn’t really matter. Let’s assume all of them—the ninety-nine prisoners—were holding clubs of the same weight and struck the man with the same force. All right? We’re using a hypothetical example. So in that example, can I convict the person? The answer is no. Legal systems do not convict a person in such a situation. So that’s not important to me right now. Maybe that example wasn’t the best one; let’s take another example. I’m talking right now on the hypothetical level.
[Speaker B] No, maybe I didn’t say it well, but that’s not what I meant. What I meant was that the element of assumption in the case of the nine stores is that we don’t really know how meat falls and how a piece falls and so on and so on. So in the case of the prisoner too there is an element of assumption, because we don’t really know whether all the prisoners—
[Rabbi Michael Abraham] That doesn’t matter. As I described it before, in the end there were ninety-nine who had this potential, this desire, and one who didn’t.
[Speaker B] Who says they all had that potential? That’s an assumption.
[Rabbi Michael Abraham] They all participated. What? I didn’t understand. All of them—the fact is, they did it. Yes, yes, they killed him all together, and then someone chooses one at random. So the probability is one percent, there’s no—
[Speaker B] Okay, okay, I forgot that point. Okay.
[Rabbi Michael Abraham] Okay. In any case, fine, so that’s the first explanation. I’m freezing it, or suspending it for now, until we see the next ones, and then maybe we’ll come back to it. The second explanation—I started it, and I brought there the contradiction raised by the medieval authorities (Rishonim), Tosafot in Bava Kamma 27 and in Sanhedrin 3, regarding the question of whether in monetary matters we follow the majority. The original passage is a dispute between Rav and Shmuel about a person who sold an ox to someone else, and the ox turned out to be a habitual gorer. The buyer wants to return this ox to the seller—this is not a good ox, I didn’t intend to buy such an ox. The seller argues that he sold him the ox for slaughter. If it’s an ox for slaughter, what difference does it make whether it gores or not? The buyer argues that he bought it for plowing; he wants an ox that will stay with him for years and help him plow the field. And a goring ox can’t be used. Now the Talmud says—and this is apparently agreed to by both Rav and Shmuel—that most oxen are sold for plowing and not for slaughter. The question now is whether the one who claims he bought it for plowing is right, or whether Jewish law follows him, because in monetary matters we follow the majority, or not. Rav argues that in monetary matters we do follow the majority, and Shmuel argues that in monetary matters we do not follow the majority. Therefore the one who claims he bought it for plowing cannot be helped by the fact—which is true in itself—that usually most oxen are bought for plowing. And the Jewish law follows Shmuel, since in disputes between Rav and Shmuel, in monetary law the Jewish law follows Shmuel, and in prohibitions it follows Rav. So in monetary matters we do not follow the majority.
On the other hand, there are places where we see that in fact we do follow the majority in monetary matters. The example Tosafot brings is the Talmud in Sanhedrin regarding the majority in a court in monetary cases. A court sits on a monetary case—three judges. Two obligate Reuven and one clears him. So the ruling follows the two judges against the one; we follow the majority. And the source for this is the verse “incline after the many,” which is basically the verse that serves as the source in general for the rule of following the majority. So we see that even in monetary law we follow the majority of judges. Tosafot asks: then we see that even in monetary matters we follow the majority, and that contradicts what Shmuel says—and that is how we rule in Jewish law in the case of the majority for plowing—that in monetary matters we do not follow the majority. So various other explanations can be offered. I want to focus on one explanation that appears in Tosafot in Sanhedrin. And Tosafot argues that the majority—I said there are two principal directions among the medieval authorities (Rishonim). One direction says that the majority in court is special, and therefore in a court we do follow the majority in monetary matters, but generally the rule really is that in monetary matters we do not follow the majority. And the second approach is the opposite: generally we do follow the majority in monetary matters. The majority for plowing in the oxen case is the special one; there specifically we do not follow the majority in monetary matters. In short, the difference is whether the majority in court is the exception and everywhere else we do not follow the majority, or the other way around: the plowing case is the exception, and there we do not follow the majority, but everywhere else we do. That’s the discussion. Why would the court be an exception? Because in court this is a majority present before us, isn’t it? What? In court it is a majority present before us, and for plowing it is a majority not present before us. Why is the majority for plowing not present before us?
[Speaker B] Because it’s the nature of the world that most buyers—
[Rabbi Michael Abraham] That’s not so clear. Not clear at all. Suppose I live in a certain place and I know that here people buy for plowing, even though maybe elsewhere in the world they might buy for slaughter. At least such a situation is possible. That probably is not the case, but I don’t think the distinction between present before us and not present before us is essential here. In any case, by the way, the medieval authorities (Rishonim) mention explicitly that the distinction between present and not present is not relevant. They reject the possibility that with a majority present before us we follow the majority, while with a majority not present before us we do not. They write this explicitly—Tosafot explicitly.
Okay. In any case, for our purposes, the one who says that the majority in court is the special case—he simply says, this is Tosafot in Bava Kamma, I think—he argues that in court the rule is that one needs—Rabbi Chaim explains it this way. Rabbi Shimon gives a similar explanation; there are some differences, but several later authorities (Acharonim) go in this direction in explaining Tosafot. Because in monetary law the Torah says that three judges are needed to rule. “God, God” is written three times in the passage. So you need a panel of three judges. Now if, when it’s two against one, we go with the two, then the Torah’s requirement of three judges is not fulfilled. In effect we ruled on the basis of two judges, not three. Three sat on the panel, but the ruling was issued by two and not by three. So if the Torah says that three judges suffice and that we follow the majority in court—“incline after the many”—the Torah is telling us that the minority in court is as though it were not there. Or, in less mystical language, in the end all three judges sign the ruling. It is true that only two held Reuven liable and one wanted to acquit him, but when the court finally rules, all three judges sign below that this is the law. Therefore the baraita says that the clear-minded people of Jerusalem would check who sat with them at a meal—that is, as a parable for the court. Meaning, judges who sit on a particular panel should carefully check whether they are prepared to join the other two judges sitting with them, even on the assumption that they may be in the minority and will have to sign a ruling that the others reached by majority opinion. If you belittle them or cannot accept the possibility that you are mistaken, if you are sure you are wiser and that if they disagree with you then they are simply fools—do not sit with them in judgment. Because you will have to sign the ruling even if you are in the minority. And there is even a prohibition—“Do not go about as a talebearer among your people”—against revealing, going to the litigant and telling him, “I really meant to clear you, they held you liable, what could I do, I was forced to sign.” You are forbidden to say that to him. That is “Do not go about as a talebearer among your people.” The document that comes from the court is a unified document: the court rules that Reuven is liable, and all three judges sign below. Their internal deliberations are their own internal business.
So what does that basically mean? That when two judges prevail over one judge by majority opinion, as far as we are concerned, all three judges ruled that Reuven is liable. And therefore in this case we really do follow the majority. Why? Because here there is no minority. The majority is basically three judges—halakhically, of course, not factually, but halakhically. And therefore here, even though it concerns money, we follow the majority. In ordinary contexts like the majority for plowing and the like, there we do not follow the majority in monetary matters.
[Speaker D] In monetary matters against someone in possession.
[Rabbi Michael Abraham] And therefore that is the special case.
[Speaker C] Rabbi, I didn’t understand the distinction you made between halakhically—
[Rabbi Michael Abraham] One second, someone started speaking, so just—
[Speaker C] Me. I didn’t understand the distinction you made here between a halakhic majority and a factual majority. Meaning what? That Jewish law treats it as though now they all agree?
[Rabbi Michael Abraham] Factually, there were two judges who held Reuven liable and one judge who acquitted him. But Jewish law views that situation as though all three judges held him liable. This is basically a kind of nullification in a majority. The minority judge is nullified into the view of the majority, and now we have, as it were, a mixture all of which obligates Reuven. This is of course a novelty of the Torah; it is not the factual truth. Factually, only two judges held him liable, not three. But the Torah views the ruling as a ruling that came out from all three judges at once. Because even the minority judge ultimately agreed that if the majority overruled him, he would join them. He was not persuaded, but he joins them. If he is not in such a situation, if he did not accept that arrangement, he is forbidden to sit with them in judgment. And therefore this really is a special case, because here there is no majority and minority; rather, the whole court determined that Reuven is liable. So this is not called following the majority. This is an exceptional case. I would even say it’s like “the majority is as the whole,” or nullification in a majority. It’s not really a matter of following the majority. By contrast, in all the other cases of following the majority, there in monetary matters we really do not follow the majority. Okay? It’s not—
[Speaker C] Can I ask one more question? Yes. Please. You said that if he doesn’t agree afterward to sign, then he is forbidden to sit with them in judgment. But then what happens in the Great Sanhedrin? Because then he has nowhere to run, because if he doesn’t sit with them in judgment, in any case he is bound by the Sanhedrin, no?
[Rabbi Michael Abraham] No, no, he doesn’t join the Great Sanhedrin. What do you mean? If he doesn’t trust the people there—what he is obligated to observe, fine, but he doesn’t sign the ruling. He will comply under protest. He’ll say, “I don’t agree, but what can I do, this is the great court and I have to follow it.” But when he signs the ruling, he is saying, “I think this is the law,” and that he is not willing to do.
[Speaker C] Okay, thank you.
[Rabbi Michael Abraham] So that is one track, the track that says that basically the case of the court is the exceptional case. What is interesting for us is the opposite track, the one that says the majority for plowing is the exceptional majority. I brought Rabbi Shimon Shkop, who argues that when the majority depends on human decision—a majority regarding gifts sent before marriage and so on, we saw that Rabbi Shimon Shkop there—then we do not follow the majority. What does that mean? True, most people buy or sell an ox for plowing; a minority buy it for slaughter. But there is nothing problematic or flimsy about the claim that I bought or sold an ox for slaughter. There are people who sell an ox for slaughter. It’s not some statement like, “I’m a strange person,” or “I behave irrationally or non-normatively.” I’m behaving in a completely rational, completely normative way. It’s true that in the minority of cases people buy an ox for slaughter. In fact, all people buy an ox for slaughter—except maybe they buy one ox in their lifetime for slaughter and many more oxen for plowing, say, just for the sake of discussion. There is nothing in that claim that makes it unreasonable or irrational behavior. It simply happens in the minority of cases. Fine. If, say, the oxen were enough for all my meals for life, then for my meals I would only need to buy one ox in my whole life. But still, I do buy one ox. There is nothing problematic in the claim that I sold or bought this ox for slaughter; it is not a problematic claim.
There is a difference between saying this is a minority of people who buy for slaughter, and saying that if you claim you bought for slaughter, that is itself a problematic claim, one that looks dubious. No, it does not look dubious. A woman, for example, who says, “I gave birth at seven months.” Okay? Even that—although we know there are women who give birth at seven months—that is still a problematic claim. Why? Because few women give birth at seven months; most give birth at nine. And after all, it is not her choice; it was not your choice; the birth happened to you. The woman does not decide that. So there one can still say that the claim “I gave birth at seven months” is a problematic claim—at least when there is a dispute. We talked about the possibility that maybe there has to be a dispute for that, because really, there are women who give birth at seven months, so what’s the problem? Never mind. But even there one might say it. In a case where the majority depends on a person’s decision or choice, and the person says, “I wanted this ox for slaughter—what do you want from me?” The need to buy an ox for slaughter is a completely reasonable, normative, logical need. You cannot say that this person is an unreasonable person acting unreasonably, abnormally, against the nature of the world. There is no nature of the world that says that a majority buy for plowing. By the way, this also connects to what was raised earlier—whether the majority for plowing is a majority not present before us or a majority present before us. Again this challenges the accepted definitions. On the one hand, it is seemingly a majority not present before us, with the qualification I mentioned earlier. On the other hand, it is obviously not something in the nature of the world that people tend to buy mostly for plowing. There is no special motivational force to buy an ox for plowing; it just turns out that usually, in most cases, people buy for that because it is the greater need. Maybe one can call that the nature of the world in some sense, but it is not some positive statement about the person—“we have clear indication that you in fact bought it for plowing, so don’t sell us stories that you bought it for slaughter, because that claim is dubious.” No, it does not look dubious.
So Rabbi Shimon Shkop says that the case of an ox sold for plowing or for slaughter is the exceptional case. Why there do we not follow the majority in monetary matters? Not because in principle we never follow the majority in monetary matters—on principle, we do follow the majority in monetary matters. But with a majority of this sort we do not. Why? Because a person has the right to claim that he belongs to the minority. He chose to act the way a minority of people or a minority of cases act, as long as that choice is a rational, normative choice—everything is fine. By the way, in this context, for example, when a person says, “I repaid the debt within the term.” Usually people do not repay within the term. “I repaid within the term.” Ostensibly, that too is just the person’s decision; he decided to repay within the term. So what? It can happen. A few people repay within the term. I decided to repay within the term. It’s not the same thing. It’s not the same because every person’s interest is to repay after the term. So when you claim you repaid within the term, that is a problematic claim. It raises suspicion. Don’t tell me you repaid within the term; reasonable people don’t repay within the term—even though some do. I’m not saying it never happens. But anyone who comes and tells me that, the claim sounds suspicious. By contrast, a person who says, “I bought an ox for slaughter”—there’s nothing suspicious about that claim. What’s the problem? People buy for slaughter. It’s a perfectly normative purchase by perfectly rational people. It just happens that usually oxen are needed for plowing and not for slaughter. But there is nothing in the nature of the world saying that one who buys for plowing is doing something unreasonable or problematic. Yes, you wanted to comment?
[Speaker B] Yes. “I repaid within the term” is like giving birth at seven months, right? Like what? Like giving birth at seven months, where there is an interest.
[Rabbi Michael Abraham] Yes, it’s like giving birth at seven months. I’m saying again—it’s stronger than giving birth at seven months. Because someone who says, “I repaid within the term,” that is a problematic claim. It goes against human nature. Human nature is not to repay within the term. Every person, including the person standing before me. Why did he repay within the term? Maybe he had explanations. But human nature is to repay after the term. With giving birth at seven months, there are women who give birth at seven months and women who give birth at nine months. There is absolutely nothing wrong with a woman giving birth at seven months. Therefore there I said that it is very likely only because there is a dispute and perhaps there are consequences to the claim that she gave birth at seven rather than nine—for example, whether the child is a mamzer or something like that—that there I still would not accept that claim as a reasonable one. But in itself, if nothing depended on it and a woman came and said, “I gave birth at seven months,” I would accept it. Why not? There are women who give birth at seven months. There is nothing there. So you can rank this as a three-level hierarchy. Repaying within the term is an unreasonable claim, even though it also happens; it is unreasonable. Anyone who makes that claim—it smells suspicious.
[Speaker B] Even in a dispute.
[Rabbi Michael Abraham] “I gave birth at seven months,” in itself, is a fine claim. There are women who give birth at seven months. The woman says, “Tell me, I belong to the minority, what can I do?” True. But when there is a dispute and we know that usually women give birth at nine months, and that this is the physiological nature of a normal woman—to give birth at nine and not at seven—and now suddenly a woman comes and says, “I gave birth at seven months,” and she probably also has some interest, so that this won’t be a mamzer child but rather a child—whatever, for various reasons—then there I won’t accept it. It sounds suspicious to me. In the case of the majority for plowing and the minority for slaughter, there the person says, “I chose to buy an ox for slaughter—what do you want from me?” The fact that a minority of people do that doesn’t contradict anything, not even human nature. In other words, a majority for seven months contradicts the physiological rules. The physiological rules say that usually people give birth at nine. True, some give birth at seven. But it is still true that the physiology of the ordinary person, of the ordinary woman, is childbirth at nine. Here, with the majority for slaughter, there is nothing unreasonable about the fact that I bought for slaughter. There is nothing there, not in nature and not in anything else. There is a reality in the world that people usually buy for plowing and not for slaughter. Therefore this is the lowest level of majority. And after such a majority, we do not go in monetary matters, as Rabbi Shimon explains the view of Tosafot.
[Speaker E] Excuse me, there’s something I don’t understand in the facts here. The one claiming that he bought it for plowing is the person who bought it, not the person who sold it. Right.
[Rabbi Michael Abraham] Who—
[Speaker E] The one defending himself and saying, “I sold it for slaughter” and relying on the minority, yes, that’s the one who basically didn’t disclose that this ox had gored three times. Right. So the plaintiff here is the one who bought the ox. He is suing and wants his money back. And in the case of “a person does not repay money within the term,” that is a claim made by the defendant. They come and tell him, “Sir, give me back the money.” He says, “No, I already paid.” There is a big difference.
[Rabbi Michael Abraham] No, fine, I didn’t mean to compare the cases. What is clear—
[Speaker E] No, but you keep saying this is a claim one can understand, that someone bought it for slaughter. But that is not the claim here. The seller says, “I sold it for slaughter.” Yes. But a person comes and says, “I bought it for plowing”—so why don’t you let him benefit from the majority?
[Rabbi Michael Abraham] Who says that? After all, he wants to return the ox to me. It could be that he is lying, that he changed his mind or something like that.
[Speaker E] Right, because he was deceived, because they didn’t tell him, because they didn’t tell him—
[Rabbi Michael Abraham] They didn’t tell him that the ox was dangerous. No, because he told me he bought it for slaughter, and therefore I didn’t tell him.
[Speaker E] Meaning the minority—and you are trying to defend the person who sold it for slaughter because you say that a person can also sell for slaughter, can want to buy for slaughter. But on the other side there is a claim supported by the majority: the person comes and says, “I bought it for plowing, and indeed most people buy for plowing,” so why am I giving an advantage to the minority?
[Rabbi Michael Abraham] So I said, because in the end, first of all, the one making the minority claim is the defendant, and therefore he has an advantage.
[Speaker E] He has an inherent advantage: the burden of proof is on the one seeking to extract from his fellow.
[Rabbi Michael Abraham] You’re suing me.
[Speaker E] No, but he doesn’t—no, no, the problem is that his inherent advantage is flawed because—you know—he deceived him, he didn’t disclose.
[Rabbi Michael Abraham] No, he didn’t deceive him, not at all. You claim I deceived you; I say I didn’t deceive you. You asked for an ox for slaughter. If you ask for an ox for slaughter, I don’t have to disclose that it gores. What difference does it make? Why is that relevant?
[Speaker F] No, he didn’t say he wanted it for slaughter. He didn’t say anything at all.
[Rabbi Michael Abraham] נכון, we struggled with that in the passage, but right now I’m talking about a situation where I claim that he bought it for slaughter, not that he said nothing. Because if he said nothing, that is more problematic. Right now I’m setting up the passage this way just for the sake of the explanation.
[Speaker E] Fine, but I think there is still something illogical here, again: the person who says, “I bought it for plowing; you didn’t tell me this ox was dangerous; most people buy for plowing,” and suddenly I’m giving a right of defense and relying on some minority, and I justify it by saying, “But it’s not terrible, because there are also people who sell for slaughter, who want to buy for slaughter.” But on the other side the majority stands in his favor, his claim stands in his favor.
[Rabbi Michael Abraham] Right, but in order to extract, in order to take from a person, you need to show that his claim is an unreasonable claim. The claim “I sold it for slaughter” is a completely reasonable claim. And you are coming to extract from me; I am the defendant. Now if I make a reasonable claim, then on what basis do you want to extract from me? That’s the whole idea that in monetary matters we do not follow the majority. Even if in principle we do, not with a majority like this, when my claim is a reasonable claim. In monetary matters we follow the majority when that majority turns the defender’s claim, the defendant’s claim, into an unreasonable claim. But here, when his claim is reasonable, how can you take money from him? By the way, “a person does not repay within the term” is the same thing in this context. Now there too it is the defendant’s claim. Who is the defendant there? The defendant is the one who says, “I repaid.” Right, but repaying within the term is done only by a minority. So there too the majority works against the defendant.
[Speaker E] So there they do follow it? What? And there they go according to the majority. They tell him, “A person does not repay within the term, and therefore you must return the money.”
[Rabbi Michael Abraham] Exactly. So I asked why. In both cases, in both cases, the defendant wants to rely on the minority. In that sense it really is similar. Why is it that there we follow the majority—or a presumption, that’s what they call it there, it doesn’t matter—and in the case of the majority for plowing we don’t follow it? Because in the case of the majority for plowing, the claim itself is not problematic at all. But “I paid within the allotted time” is a problematic claim in itself; it’s not just a minority case. That’s why they call it a presumption there and not a majority. That’s the standard explanation. What’s the difference between a presumption and a majority? The presumption that a person does not repay before the due date—isn’t that the same as saying most people do not repay before the due date? Why is this called a presumption and that called a majority? After all, in monetary law we follow a presumption; it’s only about a majority that there is a dispute between Rav and Shmuel whether we follow it or not. Isn’t it true that every presumption is really a majority? The accepted explanation is what I said here. When you say that you repaid before the due date, it’s not only that most people don’t repay before the due date, but that it’s not logical to repay before the due date. A person who repays before the due date is making a problematic claim. I’m not saying it can’t happen; it can happen. But on its face it’s a problematic claim, because it contradicts human nature. So there is a presumption that it’s not true. The majority is the result of that presumption. Because logic says that people do not repay before the due date, if you examine all repayments that happened, you’ll discover that most of them indeed happened after the due date. But it doesn’t begin with the majority; it begins with the presumption. It begins with the fact that there is a good reason not to repay before the due date. Consequently, if you check, you’ll see that indeed most people don’t repay before the due date. In the case of the majority for slaughter versus plowing, it works the other way around. You start with the majority. There is nothing in the motivation of the individual person saying that buying for plowing makes sense but buying for slaughter doesn’t make sense. There’s no problem at all with buying for slaughter. The statistical fact is that most people buy for plowing. Therefore there it’s called a majority and not a presumption. Clear? In both cases there is a majority; the whole question is whether the majority is the result of the presumption, or the majority is… what creates the presumption. Okay? That’s basically the… no, I’ll say it as a final line.
[Speaker E] What seems a little outrageous to me is this: there’s a factual dispute. Before us there is a factual dispute. Why, when there’s a rule of majority—I’m not getting into the question right now of whose Jewish law follows, whether in monetary cases we follow the majority or don’t follow the majority. I’m not getting into that for the moment. I’m just presenting a fact, a factual dispute. One says he bought it for plowing, the other says I sold it to you for slaughter. The majority supports the one who says he bought it for plowing; that is, the majority grounds the factual claim of the one who says, “I bought it for plowing.” One fact exists on the ground here: there was non-disclosure of something that happened.
[Rabbi Michael Abraham] That’s not a fact.
[Speaker E] No, they didn’t tell him, he didn’t know. Non-disclosure.
[Rabbi Michael Abraham] Non-disclosure already isn’t a fact.
[Speaker E] No, non-disclosure of a fact that he didn’t know.
[Rabbi Michael Abraham] Fine, I didn’t need to disclose it.
[Speaker E] No, okay, but factually we know there was a sale here without telling someone that this ox was a habitual gorer.
[Rabbi Michael Abraham] No, he—I claim that he told me he bought it for slaughter.
[Speaker E] Again, I’m distinguishing. They didn’t say it was a habitual gorer, and one says, “I didn’t say it because it was for slaughter, so I didn’t say it.” But still, you didn’t say some fact that exists, regardless of whether it’s important or not.
[Rabbi Michael Abraham] Let’s say this ox is two years old—so what?
[Speaker E] What I don’t understand is why, from the standpoint of—let’s say—justice, if a person has the majority on his side, and he claims that factual claim, and the other person claims something that is the minority, still I’m going to prefer the minority. That’s what I don’t understand.
[Rabbi Michael Abraham] So I’m saying, I think the explanation here is very simple. If I’m the defendant, the seller, yes? He’s the defendant, and the defendant makes a reasonable claim, then it’s not correct to say that the majority supports the plaintiff’s claim. The majority does not support his claim. True, there is a majority, but you also claimed that the majority supports the plaintiff’s claim. How does it support the plaintiff’s claim? Only if the defendant’s claim, in light of that majority, is perceived as unreasonable. But if the defendant’s claim—“I sold it for slaughter”—is reasonable, then it’s not correct to say that the majority supports your claim. Good. Now—
[Speaker G] Let’s go back to the matter of—
[Rabbi Michael Abraham] Payment before the due date.
[Speaker D] That’s true only in the topic of payment before the due date. That’s true—the presumption is true only in the world of Jewish law where there are no witnesses. Since, for example, it happened in—
[Rabbi Michael Abraham] No, no, let’s not get into details now; obviously. I’ve already brought that example several times as something where reality can overturn the presumption. That’s obvious; there’s no point going into it. Right now I’m talking about the statistical logic or the logic of following the majority. So the claim is actually—so this is basically Rabbi Shimon Shkop’s explanation of this rule: that we do not follow the majority for plowing, we do not follow the majority in monetary law, in the case of the majority for plowing. Okay? What does that mean in our case? Let’s return to the prisoners in the jail. When this prisoner stands trial, they tell him: look, ninety-nine prisoners participated in the assault on the guard and one did not participate. There is a ninety-nine percent chance that you are guilty. So he says to them: but I’m telling you that I’m a righteous person; I chose not to take part in that offense. After all, this is the result of human decision, like the majority for slaughter and the majority for plowing. There’s no—what are you telling me, that it’s unreasonable that I didn’t participate? After all, there was one who didn’t participate; I’m claiming that was me. Now of course every one of the hundred will claim that he is that one, obviously, but on the principled level you can’t say to a person that the claim you’re making is unreasonable. There is no majority in the nature of the world that causes people to attack guards. It depends on their choice. Ninety-nine prisoners chose one way; I chose differently. Now not only is there a minority who buys for slaughter, but buying for slaughter, as we saw earlier, is a reasonable claim. Here it’s even more reasonable: certainly there was such a minority—after all, one person did not attack; we know as a matter of fact that one did not attack. Beyond that, he even acted properly. That means he really wasn’t a criminal. When a person comes and says, look, I acted lawfully—can you convict him because most people didn’t act lawfully? That’s absurd. Even worse than the majority for slaughter, the majority for plowing as against the claim “I sold it for slaughter.” There it’s a neutral act; you can sell for slaughter, you can sell for plowing. Here it’s not even neutral. All I’m claiming is that I acted as a reasonable person should have acted, should have behaved. I wasn’t a criminal. And the decision whether to be a criminal or not is my decision. So I’m telling you: I’m a righteous person, I don’t commit offenses—except for the offense that got me into this jail, of course; in the meantime I repented. So I don’t commit offenses. Therefore my decision was not to participate in this act. The fact that ninety-nine others decided yes to participate in this act—on that basis can you compel me or sentence me as guilty? On what basis? After all, my claim is a very reasonable claim, and there is one such person who didn’t participate, so why should I care that it isn’t ninety-nine percent? And therefore, just as in the majority for slaughter—even more so than in the majority for slaughter, majority for plowing—here the minority is actually an even more reasonable claim than the claim of the majority. Not merely no less reasonable, but even more reasonable—or at least, I don’t know if more reasonable, more proper, at least on the criminal level, yes? I wasn’t a criminal—what do you want? Usually, if you want to convict me as a criminal, you have to prove that I’m a criminal; there is a presumption of propriety. The simple assumption is that unless it has been proven, I am innocent. So all I’m claiming is that I’m innocent. Prove that I was not innocent. The fact that ninety-nine are guilty—true, but by the same token there is also one who was not guilty. I claim I’m that one. Do you have proof that I’m not? Now of course, notice: all one hundred made that claim. The claim is that you cannot convict any of them, because each and every one of them individually has the right to say: I was the one who was not a criminal. So obviously ninety-nine percent of them are lying—it doesn’t matter—but still, you cannot convict any one of them, because each one is making a reasonable claim. And therefore this analogy to “we do not follow the majority in monetary law” seems to me to offer another explanation for this rule in the prison assault case. Now, the claim here—I want to compare this explanation to the previous explanation. In this explanation, again, probabilistically we already agreed there’s a ninety-nine percent chance I’m guilty, on the probabilistic plane. Here it seems like a legal claim, not a probabilistic claim. A legal claim says that in order to convict a person you need positive evidence against him. It’s not enough to be statistically convinced at ninety-nine percent that he is guilty. I need to point to evidence that indicates him, even if it isn’t absolute, like eyewitness testimony. And once the person makes a reasonable claim, he has a presumption of propriety. You can’t convict a person who makes a reasonable claim unless you have direct evidence of his guilt. Again—not one hundred percent evidence, but direct evidence that speaks about him and his guilt. That, that is basically the claim. Notice that what I’ve said here sounds a bit similar to the previous explanation, because in the previous explanation I also said there is nothing that indicates that the prisoner before me participated in the assault. It’s only statistics; it’s a majority that is before us. So the logic is similar logic, but here I’m not hanging it on the fact that it’s a majority that is before us. The distinction is not because it’s a majority that is before us. It happens to be a majority that is before us, maybe, but that’s not why—not for that reason am I acquitting him. Rather, I acquit him because his claim is a reasonable claim. Therefore, in a claim that concerns human conduct, a person can say: I chose to act in such-and-such a way—at least if that choice is a reasonable choice. So regardless of whether this is a majority that is before us or a majority that is not before us, as long as the matter is not—let’s formulate it like this. Suppose, for example, that if someone provokes him in a certain way, usually ninety-nine percent of people will react violently—to a very blunt provocation, a very severe one. Fine? Now, can I convict a person—he claims, “I restrained myself and did not react violently,” and the other person who provoked him says, “Not at all, he reacted violently”—can I convict him of violent behavior because ninety-nine percent of people behave violently here? Notice that here there is something in human nature that causes this, it’s not just that ninety-nine percent of people behave violently; and still I can claim that I overcame it, I did not behave violently. It’s a decision; it’s an act that results from my own choice. I chose to overcome it. So it’s somewhat similar to the prisoners’ case, although the prisoners’ case is even stronger than that. Because in the prisoners’ case there is nothing in human nature that causes me to think that this person also did the same thing. If there were a nature of prisoners such that prisoners in jail always attack guards, then when the person said, “I’m the one who didn’t attack,” that would be a weaker claim, although we still know that one person didn’t attack. But here there is something in human nature that indicates that you too attacked, because every reasonable person would have attacked; so there is room to say that the burden of proof is on you. But there is no law of nature like that, that prisoners attack a guard. It’s not the same as provocation and violent response. There it really is in human nature: when a person is provoked in that way, he reacts violently. And if someone overcame it, that really is an unreasonable claim. It happens, but it’s an unreasonable claim. I can understand why the burden of proof would be on him. Okay? I’m just trying to illustrate the boundaries of this explanation, or of this claim. But this claim really is legal in character, not probabilistic in character. That is the second explanation—the second, or the first if we accept the first one at all.
[Speaker E] What if one of those ninety prisoners had already beaten someone, some guard, two days earlier?
[Rabbi Michael Abraham] Meaning, I know that he himself has some sort of nature of attacking.
[Speaker E] He already did something similar, so he shouldn’t be able to benefit from this.
[Rabbi Michael Abraham] So it may indeed worsen his situation; I can’t tell you for sure. It may worsen his situation. That reasoning is definitely running around now; I’m not sure it collapses. Because still, every person has a presumption of propriety; he has the right at least to claim that he is proper as long as it hasn’t been proven that he is not proper. The fact that he fell in the past—think, for example, in an ordinary court, someone was convicted of theft ten years ago, and now he is accused again of theft.
[Speaker G] That won’t help convict him.
[Rabbi Michael Abraham] Exactly, so that’s why I’m saying that. Why? Because there is a presumption of propriety.
[Speaker G] No, but here you’re relying on statistics.
[Rabbi Michael Abraham] Yes, you’re relying on statistics.
[Speaker G] So when you rely on statistics, you have support for the statistics because two days ago he confirmed it.
[Rabbi Michael Abraham] נכון, but I’m explaining: the second explanation is not statistical, it’s legal. A person has the right to claim that he is proper. Until you manage to prove otherwise. So a person has that right even if he was convicted in the past. He has the right to claim he is proper. You’re right that here his claim may be weaker in a statistical sense, or whatever one calls it, but still there is something—what today is called the presumption of innocence, or in Jewish law the presumption of propriety—which says that until you’ve proven otherwise, I have the right to claim that I’m a proper person. And that perhaps brings me to the third formulation. In Jewish law there are two kinds of doubt. There is a doubt where a prohibition is established, and a doubt where a prohibition is not established. What does that mean? One piece out of two pieces, and one single piece. Suppose there are two pieces of meat before me. One is carrion and one is slaughtered meat. I know that. I just don’t know which is which. Now I took one of them and ate it. That is called a doubt where a prohibition is established, or by another name, one piece out of two pieces. Okay? There is a fifty percent chance I ate non-kosher, and a fifty percent chance I ate kosher. A second kind of doubt is the doubt of one piece. There is one piece before me. I don’t know whether it is non-kosher or kosher. Again, apparently it’s fifty percent, and if I ate that piece then apparently there is a fifty percent chance I committed an offense. In Jewish law there is a difference between these two kinds of doubt. For a doubt of the first kind, for example, one brings a provisional guilt-offering in places where such an offering applies. Something whose certainty requires a sin-offering—its doubt requires a provisional guilt-offering. By contrast, in the second type of doubt, one piece—that is different from one piece out of two pieces. For the doubt of one piece, one does not bring a provisional guilt-offering. It is considered a lighter doubt. Again, probabilistically, in both cases apparently there is a fifty percent chance I committed an offense. And nevertheless, the first case is considered more severe than the second. Why? Because in the first case, in the first case, I have a positive reason to doubt. One of the two pieces I know is—one of the two pieces. Now the question is whether I took that one or the other. In the case of one piece, I have no indication in the piece itself that it is non-kosher or not non-kosher. I only know that in the world there are non-kosher pieces and there are kosher pieces, so I have doubt whether this piece is non-kosher or not. But I don’t have a positive reason to doubt. It’s a negative reason. I doubt because of ignorance, because of lack of information. In the first case I doubt because I have information. In the second case I doubt because I lack information. I brought an example of this in the last column, and there I talked, yes, about two kinds of left or two kinds of equality. I said that I brought this example of established prohibition, but think, for example, of someone flipping a coin. And I know the coin is fair. Now I flip the coin: there’s a fifty percent chance it lands heads, a fifty percent chance it lands tails. Or in Haifa they say tour or yaz. So now that doubt is fifty percent, right? It’s a doubt that stems from the information I have that this coin is a fair coin. Meaning, I know the chance it lands heads and tails is equal. What happens in another case? They bring me a coin; I have no idea whether it’s fair or not, I know nothing at all about the nature of the coin, and they tell me: now bet whether it will land heads or tails. Statisticians will tell you that this isn’t even a problem in statistics. The problem isn’t defined in the statistical sense. Because in the statistical sense, a defined problem is one for which there is a distribution. Or in other words, I have some prior information. Not what will happen, but I know the distribution regarding what will happen. So if the coin is fair, I know the distribution is fifty-fifty. That is positive information. I know the distribution. So now the question of what will happen is a statistically or probabilistically well-defined question. In the second case I have no distribution. I don’t know how coins in the world are distributed, how many are fair, how many are not fair, what kind of unfairness. I have no idea, I know nothing. I have complete ignorance; I have no information at all about the situation. Now they tell you: you must bet. Even so, although the statisticians tell you this is not a probabilistic problem, you are required to bet. Otherwise I’ll shoot you in the head if you don’t bet. Fine? Now the question is: how much do you put on heads? I assume we would still bet fifty-fifty. Right? For reasons of symmetry. I have no reason to prefer heads over tails or vice versa. So I bet fifty-fifty. What else am I going to do? I have nothing better than that. But you understand that the second bet is a bet grounded in ignorance, not in information. I have no information, and in the absence of information all possibilities are equal. If I have information that the coin is fair, the fifty-percent bet is a bet based on information I have. I have information about the situation. In both cases I bet that it’s fifty against fifty or one against one. But the basis for that bet is different. In the first case it’s a bet based on information; in the second case it’s a bet based on lack of information or ignorance. You see the comparison to the doubt where a prohibition is established and where it is not established. When there are two pieces before me, one non-kosher and one kosher, I have a positive reason to think there is a fifty percent chance I took the non-kosher one, because after all there was one non-kosher and one kosher here. So I have information that leads me to assume there’s a fifty percent chance I took the non-kosher one. By contrast, if there is one piece before me, I have no idea what it is. I know nothing. I only know that it could be non-kosher and it could be not non-kosher. And I also don’t know the distribution in the world and, for the sake of discussion, suppose I know nothing at all. Okay? Still I bet fifty-fifty. The law is the law of doubts. We must be in doubt under the laws of doubts here. But it’s a doubt grounded in ignorance. It’s not a doubt grounded in information. It will still be fifty percent that it is non-kosher—I’m using percentages although it’s not really a probabilistic problem—but yes, I would bet one against one that it is non-kosher, even though this is just ignorance. Now, that is the distinction between these two types of doubt. In the case of the prisoners in Papiño, there are ninety-nine prisoners before me who attacked, but there is one who did not. There is one who did not. So here I have a positive reason to acquit him. Because if it were—if, say, it were the nature of the world that prisoners attack guards, then true, factually only ninety-nine percent of prisoners attack guards, because that I know; I have information about human nature. There is a ninety-nine percent chance that a prisoner will attack a guard. Not that there were a hundred prisoners and a concrete distribution, but I know about a nature—suppose I knew such information about the nature of the world. In ninety-nine percent of cases, people tend to attack guards in that situation. If that were the case, then I would have information that also pertains to the prisoner who is standing before me now for trial. Because I know something about human nature, and after all he too is a human being. He too is a human being, and the regular laws of human nature apply to him as well. True, the laws say that only in ninety-nine percent of cases will he attack, and one percent not. But that law saying that a person’s tendency is to attack is a law that applies to this person too. Therefore his conviction in that case—or acquittal or conviction in that case—has a side for stringency and a side for leniency. On the one hand I have a positive reason—let’s say for stringency—I have a positive reason to convict him. Because after all, human nature is to attack. And I have no positive reason to acquit him. Only that doubt that says, okay, but it’s only ninety-nine percent, maybe he belongs to that one percent? Okay? By contrast, if I’m speaking about a situation where I have no such information about human nature, there were simply a hundred prisoners here, ninety-nine of them attacked and one did not—then it’s not that in human nature there is a ninety-nine percent chance that every person in such a situation will attack. Here we begin from the facts and then create the majority or the presumption, rather than the presumption creating the majority, as I said earlier. In such a case, after all, there is a prisoner here who did not attack. How can you convict this person? After all, there is one here who did not attack. That parallels a case of established prohibition, only this time toward leniency. Meaning, when I look at the side for acquittal, not the side for conviction, there is a positive side for acquittal. Why? Because I really know that there is one prisoner here who did not attack. I don’t know who he is; maybe it’s the one before me. There is a positive side for acquittal. By contrast, in a case where I have information about human nature that there is a ninety-nine percent chance he will attack a guard and one percent that he won’t, I do not have positive information that there is a prisoner who did not attack. I have statistical information saying that there is a one percent chance that a person may not attack, but there is no concrete person before me who did not attack. Do you see? It’s exactly like one piece out of two pieces as opposed to the doubt of one piece. In one piece out of two pieces there is a non-kosher piece before me, not that I have some law of nature saying that fifty percent of pieces in the world are non-kosher and fifty percent are kosher. That would be the doubt of one piece. In the doubt of two pieces I don’t care what happens in the world or what the nature of pieces is; there are two pieces before me, one non-kosher and one not. I have a doubt grounded in the facts that are before me. Therefore it is a stronger doubt than the doubt of one piece. In our case I’m going in the direction of acquittal. When I have information about human nature that it attacks, then I don’t have a positive doubt for acquittal. I only have doubt in the direction of conviction. The simple tendency is to convict, just as when two people see that you committed murder. True, sometimes eyewitness testimony is misleading, and I have positive testimony that you committed murder with some doubt attached to it. I note on the side that there is some doubt, but I have positive testimony that you committed murder—not statistical, not general. In the prisoners’ case I have nothing whatsoever that you committed murder; on the contrary, I have clear evidence that one of the prisoners did not participate. How can I know it wasn’t this one before me? You see that this is already very similar to the previous explanation. The previous explanation says that I have the right to claim that I am the proper person, that I am the one who did not participate in this act. I’ll formulate it differently, in a form taken from Maimonides and often quoted. Let me prove it by contradiction. Suppose I choose the opposite policy. I convict on the basis of statistical evidence. Then in the prisoners’ case I am basically convicting all one hundred. Right? Every one who comes before me—there is a ninety-nine percent chance that he participated. So all one hundred will be convicted. Okay? That means I am necessarily convicting one innocent person. If I follow this policy and apply it consistently, an innocent person will certainly be convicted here. Okay? That we do not do. We do not do that. You cannot adopt a policy that certainly convicts an innocent person. Now notice, the certainty here is a somewhat misleading concept. Suppose I have only—suppose I have statistics, say, about eyewitnesses. I have eyewitnesses who saw Reuven murder Shimon, and I know that eyewitnesses are reliable ninety-five percent of the time. Here too someone could come and say: if you rely consistently on eyewitness testimony, then in five percent of cases you will certainly convict innocent people. That’s not the same thing. Why not? Because I do not have before me a person, I do not have before me a case where an innocent person will certainly be convicted here. I merely assume that statistically in five percent of cases I will make a mistake, but it may also not happen. Maybe in this case all these witnesses who came really are witnesses who saw clearly? I cannot know with certainty that an innocent person will be convicted here, and therefore here I allow myself to rely on eyewitness testimony.
[Speaker H] Sorry, one moment—
[Rabbi Michael Abraham] One moment. In the case of the groups of ten, I have absolute certainty before me that among these hundred people there is one innocent person who has been convicted. Such a legal policy is not adopted. Now the analogy I make to this is one piece out of two pieces versus one piece. Where a prohibition is established, if I eat both pieces on the basis of this permission, then I certainly ate a non-kosher piece. Right? Because one of them is non-kosher and one is kosher. If I have two individual pieces, each time one piece comes before me and I eat two, did I certainly eat a non-kosher piece? No. Each such piece—I don’t know, maybe it is kosher, maybe it isn’t—I ate it. The second piece—maybe it is kosher, maybe it isn’t—I ate that too. There is no certainty that I ate something non-kosher here. With one piece out of two pieces, if I apply the permission to both pieces, then I certainly ate a non-kosher piece. Therefore, although statistically eating one piece is only a fifty-percent prohibition in both cases, you understand that there is a difference—at least legally, even if not statistically—there is a legal difference between applying the rule to one piece out of two pieces and applying the rule to one piece. And I think the same is true regarding the groups of ten. Yes, someone wanted to comment earlier?
[Speaker B] So why does the direction change?
[Rabbi Michael Abraham] What? Why does—why does the direction change?
[Speaker B] Because in the case of the two pieces, you say he is forbidden to eat them because there is definitely a prohibition there, whereas in the case of the groups of ten you acquit him.
[Rabbi Michael Abraham] Very good, because the context is different. The consideration is exactly the same consideration, only the context is reversed. In matters of prohibition, if I might violate a prohibition then in a case of doubt I must be stringent. The whole case I am concerned about is a case where I might violate a prohibition. In the groups of ten, the case I am concerned about is the case where I would convict an innocent person, and that I am not willing to do. I mentioned Maimonides but didn’t end up quoting him. Maimonides says: why do we rely only on two witnesses and refuse to accept other kinds of evidence? Because it is better that many offenders go free than that one innocent person be convicted. And that line of Maimonides is often quoted.
[Speaker B] So it really isn’t similar to the second explanation, or the first either.
[Rabbi Michael Abraham] I didn’t understand.
[Speaker B] In principle, according to this third explanation, basically we should have applied the laws of probability—
[Rabbi Michael Abraham] But—
[Speaker B] There is a side problem.
[Rabbi Michael Abraham] This explanation is not—the problem is an indication. The problem is not—that’s exactly what I want to claim. The problem that one innocent person will certainly be convicted here is an indication. It’s not the problem itself. It’s an indication that such a conviction is problematic. Because basically it means you have no information about this person on the basis of which you are convicting him. And the proof is that there will be a person here who is innocent and you will convict him in the same way. So in fact you have no direct evidence about the person himself. Therefore I argue that the three formulations I’ve presented here—or at least the last two—are really different sides of the same coin. Perhaps formulations that seem different. The first formulation says: I always have the right to claim that I am innocent, and that is a reasonable claim, and the fact that most are not like that is irrelevant. Here I’m basically saying something very similar: you have no direct or positive evidence against the defendant standing before you; you have some probability. But notice: the indication that you have no direct evidence against him is that within this particular case before you, there will be one innocent person who is convicted. One innocent person who will be convicted if you do all one hundred.
[Speaker H] Can the Rabbi sharpen for a moment exactly what the picture is with the hundred, where we know that ninety-nine—or ninety-nine of them, basically—are guilty? What exactly happened there?
[Rabbi Michael Abraham] I’m saying: I looked from afar with binoculars and I saw that ninety-nine guilty ones attacked, and one remained sitting on the side.
[Speaker H] Meaning, I am a witness that those people were there.
[Rabbi Michael Abraham] No, but—
[Speaker H] I am a witness that there are ninety-nine people there who did it.
[Rabbi Michael Abraham] Correct, and one who didn’t. But I can’t identify who that one is. So now there will certainly be one person here—
[Speaker H] Even though he’s innocent, he will be convicted. Because when two witnesses come to testify and we say there is a chance, say five percent, that they didn’t see clearly, here too I didn’t see it clearly, but I did see that ninety-nine did do it.
[Rabbi Michael Abraham] Ninety-nine is not evidence against any person. Two eyewitnesses have a ninety-five percent chance of being right and a five percent chance of being wrong, but they are speaking about a specific person. There is positive evidence directed against him. True, it isn’t certain. There, by contrast, there is no evidence at all against the person. On the contrary, I have certainty that there was one innocent person here, and if I follow my method consistently I will convict one innocent person here. Okay. So in the end I accept the comment, because in the column itself I didn’t write it that way. In the end I accept Eliyav’s comment that after all these explanations, in the case of a majority that is before us—like one piece from the nine stores—I really think that there it is different from a majority that is not before us; it is different statistically and factually, not only legally, because I don’t think it is correct to speak there in terms of probability. In the case of a majority that is before us, of the tenth man—which is also a majority that is before us—there I do think it is statistical. Meaning, the difference between that and eyewitness testimony, which is also reliable ninety-five percent of the time, is not a difference on the statistical plane, contrary to what I tried to argue in the column. But still, we see here that in the legal context considerations can enter that go beyond statistics. Meaning, even evidence that would be statistically better, we still won’t use for legal reasons, let’s call them, and not for statistical reasons—not because of the level of statistical certainty. And that’s a whole story. It’s an interesting perspective on the meaning of statistical thinking in the legal and halakhic world. I’m supposed to start Enoch’s article now, so let me come back for a moment—I’m actually not going to start Enoch’s article now, but I’ll just mention this point, that indeed the distinctions we are making here are probably not distinctions on the statistical plane. Still, that doesn’t mean there is no legal justification for them. Meaning, in the halakhic world, when we talk about a rule that is disconnected from statistics, we tend to say, okay, it’s a scriptural decree. The disqualification of testimony by relatives or something like that—it’s a scriptural decree. Scripture decrees that we act in a way that apparently contradicts reason. But it turns out that even in the legal world, which does not operate with scriptural decrees, people sometimes act not only on the basis of reliability or statistical certainty. Rather, there are various legal considerations that can lead me to act in a certain way even though on the statistical level maybe it is not justified, or no more justified than another case where I do act that way. There are things called legal considerations. But even when I speak about legal considerations, I want to make a distinction here, because it’s a very significant distinction in my view. Sometimes there are situations where I can find the legal explanation, like what I just proposed regarding the tenth man. But there are situations where we sometimes have some legal intuition that this is not the right way to act, and I try to think to myself why, and I can’t find it. What do I do in such a situation? I have a very strong intuition, and I can’t rationalize it. Now if you want to see an example of this, look at Papiño’s article. The column is column 227 on my site, and there at the beginning I refer to Papiño’s article, to the translation on the Alaxon site. And after Papiño rejects all the possible explanations, he remains with the fact that basically he has no explanation for the difference between statistical evidence and other evidence, or between eyewitness testimony and the prisoners, and nevertheless it is clear to him that this is what should be done. He is left simply with a question needing further analysis as to why; he doesn’t understand why. He does not draw the conclusion: well, if I have no explanation and I don’t understand why, then apparently I just have some kind of thinking error. So I feel that this is different from that—fine, so I feel it; so what? We have lots of illusions. We have various cognitive illusions, perceptual illusions, yes, all sorts of illusions. So why pay attention to it? And this is a wonderful example of the fact that we have a very strong trust in our intuition, even when discursive reason cannot conceptualize it or rationalize that intuition. But everyone feels that this is the right way to act. And we’ll see this later as well in connection with the article by Enoch and his colleagues. Okay, I’ll stop here. Any comments or questions? Okay, Sabbath peace, and goodbye.