Uncertainty and Probability—in Halacha, in Thought, and More Generally—Lesson 30 – Rabbi Michael Abraham
This transcription 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
- Human choice versus probabilistic considerations
- Engagement gifts, custom, and majority in Jewish law
- The question about psychology: why statistics do work on human beings
- Maimonides and the Raavad in the Laws of Repentance: the decree concerning Egypt and free choice
- Intentional and unintentional action, randomness, and agency
- Determinism, libertarianism, and influences that do not determine
- Group versus individual: when statistics are relevant
- Psychology as research versus therapy
- Statistical evidence in law: the prisoners case versus eyewitnesses
- Maimonides in the Laws of the Sanhedrin: monetary law versus capital law
- Presumptions, one witness in prohibitions, and punishment
- Reasoning, textual derivations, and scriptural decree
- The dispute over the source of Jewish law: reasoning versus verse
Summary
General Overview
The text argues that when an action is the result of a person’s free decision, there is no justification for assigning it a probability based on a general statistical distribution, as illustrated by choosing a single black ball out of a hundred when the choice is not random but depends on taste and intention. At the same time, it claims that in psychology and in describing human behavior we nevertheless do use statistics, and the solution is to distinguish between describing large groups and predicting a single individual. The discussion then moves to a similar tension in law: why purely statistical evidence is not enough for conviction, while witness testimony is, even though testimony would seemingly be less probabilistically “secure”; later Maimonides’ distinction is presented between monetary law and capital law with respect to the laws of evidence. Finally, the text argues that halakhic distinctions derived from textual interpretations are not irrational “scriptural decrees,” but rest on reasoning, and it also examines the possibility that the reasoning is probabilistic or legal in nature.
Human Choice versus Probabilistic Considerations
The text states that when a person chooses intentionally, one should not apply statistical considerations to his action as one would to random events. Therefore, even if there are ninety-nine red balls and one black ball, the principled chance of taking out the black one is fifty-fifty, because the choice depends on the chooser’s will. The claim is that inferring from the distribution in a group to the “probability” that a particular person will choose one way or another is mistaken, because the choice is not a lottery but a decision. The text emphasizes that even if an overwhelming majority behaves one way, a person can honestly claim that he belongs to the minority, and the very fact that this is a decision rules out using statistics against him, even if one does not necessarily trust his claim.
Engagement Gifts, Custom, and Majority in Jewish Law
The text cites Nachmanides on engagement gifts and argues that in his view one does not follow the majority there, because sending a gift or betrothal money depends on the man’s private decision even in a place where there is a majority custom. The text sharpens the point that when the man and woman make opposite claims about the intent of the gift, uncertainty is created, but one should not return to statistics in order to decide, because the case is built on decision rather than chance. The proposed solution is to remain in doubt and apply the laws of doubt, and in monetary matters one says, “the burden of proof rests on the claimant,” while noting that there are disputes among the medieval authorities (Rishonim) on the subject and presenting a dissenting approach that sees room to follow the majority even in acts of choice when the case is treated as a “random draw” from a group.
The Question About Psychology: Why Statistics Do Work on Human Beings
The text raises a difficulty: in psychology, human reactions are described through statistical laws even though the reaction is itself a decision, and it asks why choice-behaviors obey a stable distribution the way dice do. The claim is that the law of large numbers is clear in random and independent events, but in human choices each individual decides on the basis of considerations, so it is not self-evident that the same distribution will emerge in every large sample.
Maimonides and the Raavad in the Laws of Repentance: the Decree Concerning Egypt and Free Choice
The text presents Maimonides as distinguishing between a collective decree and individual choice, such that God decrees that the Egyptians as a whole will enslave Israel, but each Egyptian chooses for himself and is therefore punished for his choice. According to the text’s account, the Raavad finds this difficult, because if each individual has a choice, it is possible that all individuals will choose otherwise and the collective prophecy will be nullified; the difficulty reflects a refusal to apply the law of large numbers to human choices. The text explains that Maimonides assumes that in the aggregate a distribution emerges even where there is free choice, whereas the Raavad sees the collective as merely a collection of individuals that does not guarantee any distribution.
Intentional and Unintentional Action, Randomness, and Agency
The text distinguishes between an action in which a person acts as an agent and decides, and an action that is not a decision but rather “happens to him,” connecting this to the Talmudic text in Makkot and to the idea of circumstances being arranged. The claim is that statistics belongs to random or non-decisional processes, whereas in intentional wrongdoing a person decides, and therefore his act should not be described as a probabilistic rule.
Determinism, Libertarianism, and Influences That Do Not Determine
The text presents a deterministic attack claiming that genetics, the brain, and the environment determine behavior and therefore there is no choice; against this it argues that an “enlightened” libertarian admits influences but claims they influence rather than determine. The text maintains that even with deep familiarity with motives and psychological laws, there can be no certain prediction of what a single individual will do, and that free choice is not a tiny statistical fraction but a person’s ability to choose otherwise despite his tendencies.
Group versus Individual: When Statistics Are Relevant
The text argues that there is no contradiction between not applying statistics to the choice of an individual and using statistics in psychology, because psychology deals with distributions of large groups rather than prediction of a particular person. The claim is that about a group one can say “999 out of 1000” even if each person chooses, but one cannot infer from that that a particular person’s probability is 999/1000 without specific information about him. The text connects this back to the halakhic distinction between kavua and parish, where random separation allows one to follow the majority, while non-random choice rules that out.
Psychology as Research versus Therapy
The text argues that confidence in the discipline of psychology is limited, and that it is “certainly not a science” in the area of treating an individual person, because one cannot scientifically apply statistical tools to him. The claim is that basic psychological research on groups, which identifies correlations through statistics, can be considered a scientific field when done systematically.
Statistical Evidence in Law: the Prisoners Case versus Eyewitnesses
The text presents Pappino’s example: a hundred prisoners, ninety-nine murdered a guard and one did not, and with respect to one defendant there is no specific evidence against him but there is a 99% probability that he is guilty; nevertheless there is a “consensus among jurists” not to convict on that basis. Against this it presents a situation of two witnesses testifying to a murder when their estimated reliability is 97%, and in that case it is accepted that one does convict; the text sharpens that this gap does not fit probabilistic logic. Pappino is presented as proposing a distinction between “knowledge” in testimony and lack of “knowledge” in statistical evidence, but he too rejects that distinction on the grounds that the decisive question is the probability that the verdict is correct.
Maimonides in the Laws of the Sanhedrin: Monetary Law versus Capital Law
The text cites Maimonides (Laws of the Sanhedrin chapter 24), who says that in monetary law a judge rules according to what his mind inclines toward as true even without clear proof, illustrating this with a case of Rava who relied on his wife’s whisper to reverse an oath, and establishing that the main thing is the judge’s conviction rather than formal evidence rules. At the same time it cites Maimonides (beginning of chapter 20), who says that in capital law one does not punish based on estimation, and even in the case of a pursuer found with a blood-dripping sword next to the murdered person one does not execute without seeing the striking itself; the text attributes this to a formal requirement of witnesses rather than a probability assessment. The text notes a dispute with Tosafot, who understand the refusal to execute as a factual doubt, whereas Maimonides understands it as a structural rule in the law of evidence, and it discusses implications such as DNA and cameras regarding their status as evidence.
Presumptions, One Witness in Prohibitions, and Punishment
The text states that ed echad ne’eman be’isurim is not a formal category of testimony but a mechanism of being convinced regarding the status of permission or prohibition, and therefore one may rely even on a woman, slave, or minor if they are convincing. The text cites Maimonides’ view that one witness can establish a prohibited status, and if afterward two witnesses testify regarding the act of eating, lashes can be administered. It also explains that in “they flog, stone, and burn on the basis of presumptions,” the presumption establishes the background status (such as paternal identity), but the act of transgression itself still requires testimony and warning.
Reasoning, Textual Derivations, and Scriptural Decree
The text argues that the distinction between witnesses and circumstantial evidence cannot be a senseless “scriptural decree,” because the details are not explicitly written in the Torah; therefore if the sages derived it that way, there must be some reasoning behind it. It also argues that textual derivations are not arbitrary, because if there is no reasoning it is unclear why the derivation includes this particular thing rather than some other thing. The example given is “You shall fear the Lord your God,” which includes Torah scholars rather than something random. The text distinguishes between probabilistic reasoning and legal reasoning, and gives as an example “the burden of proof rests on the claimant” as a rule justified legally without claiming that the current possessor is more likely to be right.
The Dispute over the Source of Jewish Law: Reasoning versus Verse
The text presents opposition to the claim that Jewish law is determined by reasoning alone and that the verses are merely a “fig leaf,” emphasizing that in matters of prohibition and punishment one does not punish based on reasoning alone, because “one punishes only where there has first been a warning.” The example is meat and milk, where a textual derivation expands the prohibition to eating and thereby allows warning and lashes; on the other hand, a conceptual dispute is brought regarding “an eye for an eye” and the possibility that some hold it literally, or in certain contexts.
Full Transcript
[Rabbi Michael Abraham] First of all, I still need to complete what we did last time, so I’ll remind you of the background. We started with the claim that when we’re dealing with actions that are the result of human choice, it’s not correct to apply statistical, probabilistic considerations to them. The example we took from the explanations we saw about kavua was this: a person is presented with a container full of balls, ninety-nine of them red and one black, and now the person has to choose a ball. He looks, he chooses a ball in an obviously non-random way, and the question is: what’s the chance that he’ll take out a black ball? In principle the chance is half, even though there’s only one black ball out of the hundred. Why? Because he’s really taking out whichever color he chooses, whichever color he wants. Now, since I don’t know this person’s taste—whether he likes red balls or black balls—I assume it’s fifty-fifty: fifty percent that he likes red, fifty percent that he likes black. If he likes black, then he’ll search for that single black ball in the container and take it out. If he likes red balls, it’ll be easier for him and he’ll immediately look for a red ball and take it out. But for our purposes, the fact that there are ninety-nine red balls does not mean that the probability that the ball will be red is ninety-nine percent. The probability is half, because the action the person is doing is not a random extraction, a random draw of a ball, but the result of a decision. And we brought other examples of this in Jewish law as well: an ox for plowing versus slaughter, or engagement gifts, Nachmanides speaking about engagement gifts, where a person decides whether to send a gift to his fiancée, and we’re wondering—in other words, a person sent a gift to his fiancée and we’re wondering whether it was betrothal money or just an ordinary gift. And the Talmudic text makes it depend on the question of what the local custom was: did people first give betrothal and only afterward send gifts, engagement gifts, or did they first send the gifts and only afterward give betrothal money? And simply speaking, you follow the majority. If most people there customarily send engagement gifts first, then that’s the law; if they first do betrothal, then the law is different. Nachmanides says no, this does not go by the majority in the case of engagement gifts, because with engagement gifts it’s the person’s decision what to give first. Even in a place where the majority give the gifts first, there are people who give them afterward; there’s a minority like that. So the person says, look, I belong to the second type of people, I decided to give the engagement gifts afterward. Therefore I can’t tell him, look, but most people in this place first give the betrothal money and only afterward the engagement gifts. Fine, most people do that, but this is my decision; I chose; I belong to the people with the other taste. I’m one of those who like black balls, not red balls. You can’t handle this issue with probabilistic tools if we’re dealing here with a human decision. So that was the point.
[Speaker B] But the woman says the opposite?
[Rabbi Michael Abraham] I didn’t understand. The woman, in the engagement gifts case—yes.
[Speaker B] Right, so it’s not the point. Obviously he can say “I, I,” but we have to stand on the side here; there are two people saying with certainty what they saw and what they did, and we want to examine it statistically, and if this custom of giving engagement gifts is ultra-rare—one in a billion gives engagement gifts—then obviously we wouldn’t make a statistical consideration?
[Rabbi Michael Abraham] No, no, no.
[Speaker B] Why not?
[Rabbi Michael Abraham] Why wouldn’t we go by the majority? Because since for the person this is the result of his decision, then if he claims that he decided to give the betrothal money first, let’s say—in other words, to go like the minority—then he has the right to claim that he’s from the minority, even though the majority act differently. So what?
[Speaker B] But there’s also a woman here. Our question isn’t whether he has the right to claim it—of course he has the right to claim it—but she claims the same thing in the opposite direction. She was also there at the event and she knows; he told her, “I’m giving you this for the sake of betrothal.” Okay, so it’s a dispute. So because she claims against him, we’re in doubt. But that doesn’t mean we need to go back to statistics.
[Rabbi Michael Abraham] No, you’re saying that because she claims against him, we need to go back to statistics? No. She claims against him and therefore we’re in doubt, but that doesn’t mean we need to go back to statistics. Statistics speaks about a case that happens randomly, but here this is an action that is the result of a decision. So even if I don’t believe him that he belongs to the minority, I don’t need to believe him; it’s enough that this thing is the result of a decision for me not to apply statistical tools here. At most, you can say that if the woman claims against him, then I won’t believe him—fine. But on the other hand, obviously I also won’t apply statistical tools against him. So what if ninety-nine percent of the people here act differently? I’m claiming that I acted this way. So I’m saying your argument would be relevant if I were saying that we need to believe him. But if I’m claiming no, not that we need specifically to believe him—we’re in doubt because maybe he’s lying, maybe the woman is lying, we’re in doubt—but it’s an even doubt. It doesn’t matter that ninety-nine percent of people here behave unlike him. That’s my claim—not that we have to believe him specifically.
[Speaker B] What if he said, “I don’t remember”? He said, “I blacked out, I don’t remember why I gave the engagement gifts.” So here surely we would agree?
[Rabbi Michael Abraham] There’s room to say that even there, no. Because since this thing is the result of a decision, so what if he doesn’t remember? When he did it, he performed an act of decision. He just now doesn’t remember what he decided. But when he did it, he performed an act of decision, and that act can’t be judged using statistical tools. I can’t say there’s a ninety-nine percent chance that the decision was such-and-such.
[Speaker B] But why not?
[Rabbi Michael Abraham] On that kind of statistic I’d go to elections. So what if I don’t remember now?
[Speaker B] Yes, but when we need to weigh what we think this creature was thinking at that moment, then it’s ninety-nine percent, or even one in a billion, that he thought differently.
[Rabbi Michael Abraham] And we would say no—not one in a billion that he thought differently, exactly the opposite. Let’s say there’s only one in a billion—
[Speaker B] people who give engagement gifts.
[Rabbi Michael Abraham] One in a billion of people gives engagement gifts, that’s true. But from here you inferred that it’s one in a billion that he thought differently, and that’s already not true, because you’re assuming that the distribution among people projects onto the probability of each person individually, and that’s not true. Exactly like with the balls in the jar. In the jar there are ninety-nine red balls.
[Speaker B] If the Rabbi were in Las Vegas and could put down money, bet on what he really chose, and let’s say it would be possible to find out afterward by some prophecy, then the Rabbi wouldn’t put down a huge amount because it’s one in a billion that he thought otherwise? The Rabbi wouldn’t stake a billion, all his money, on that long shot?
[Rabbi Michael Abraham] Maybe I would put down some amount, but that doesn’t mean that what’s operating here is the statistical majority. Would I now say the probability of this is 999 out of 1000? No. Maybe I’d say yes, it’s more likely than not anyway. But when I want to apply the statistical majority here, I’m basically claiming that the statistical majority is relevant to this. I’m saying no, it isn’t relevant to this. It may still tilt me a bit more this way than that, but I still can’t claim that the probability is 999 out of 1000.
[Speaker B] I conduct an in-depth psychological investigation of him over hours and days, and a team of experts sees that he matches exactly, and all his opinions and habits are similar to the general public. The general public, in one out of a billion cases, does give engagement gifts. Let’s say he suddenly turns out to be some bizarre creature that popped up.
[Rabbi Michael Abraham] If you actually did that statistical study, then our situation is different; maybe there we really would go after the majority. Fine. But you didn’t do that. You don’t know whether he’s built like everyone else. He claims he’s not built like everyone else. You can’t say he isn’t. You don’t have to accept what he said, but you can’t say that the statistics stands against him.
[Speaker B] So leave us already, let’s get to the next stage.
[Rabbi Michael Abraham] Wait, wait, wait, you haven’t gotten to the next stage. That’s exactly where I’m heading, in just a second. So on the one hand, I’m basically saying that you can’t apply probabilistic tools to human decisions. That’s one side. The other side of the coin—and this is what I ended with last time—is that we do in fact apply probabilistic tools to human behavior. All psychological statistics are basically built on that. When I establish some psychological law—yes, the only psychological law is that there’s a connection between frustration and aggression, as everyone knows—so let’s say I say there’s a ninety-nine percent chance that someone who is frustrated will react aggressively. Okay, so there’s some statistical claim here that psychology stands behind, and therefore the assumption there is that we do apply probabilistic tools to human behavior, even though the decision whether to behave aggressively or not is a decision a person makes. It’s not only instinct. There are people who respond instinctively, but in principle it’s a decision, the result of a choice, and nevertheless we use probabilistic tools. And I explained that this is really not a trivial matter. In other words, when we use probabilistic tools on random events, that sounds altogether natural to us. But when we use probabilistic tools on human behavior, why assume that human behavior obeys statistical laws? After all, no individual human being is conducting a lottery—yes, we talked for example about rolling a die. With a die roll, precisely because each roll by itself is completely random and because there’s no connection between one roll and the next—the rolls are independent—precisely because of that we have the law of large numbers. I know that the distribution will be one-sixth for each outcome, for each number, each face of the die. But with human behavior, there may be independence between people, but the behavior of the individual person is not probabilistic behavior. The person doesn’t conduct a lottery in order to decide; he deliberates, uses judgment, and arrives at a conclusion about what he wants to do. So in that kind of situation, why assume—
[Speaker C] that statistical tools are relevant?
[Rabbi Michael Abraham] —that an individual case, out of many other cases that may be independent, but the individual case itself is not a case that proceeds randomly or by chance. And apparently we are at least used to thinking that probability is something that describes events that happen randomly. And the law of large numbers says that if many, many such events happen and they are independent, then I claim they’ll be distributed the way I would expect—yes, like I predict. If it’s a fair die, then there’s a one-sixth chance of getting each face. Okay? But with human behavior, on the face of it, I wouldn’t expect that to happen. In human behavior, each person individually makes a decision based on all kinds of considerations of his own. Why assume that every million people we take will always yield a distribution where eighty percent of the frustrated react violently and only twenty percent of those not frustrated react violently? Why assume that in every million, every group of a million people, it will behave the same way? This is human decision, yes? Remember Maimonides and the Raavad in chapter 6 of the Laws of Repentance—that was basically exactly their dispute. Maimonides claims that the Holy One, blessed be He, decreed concerning the Egyptians that they would enslave Israel, but on the other hand every individual Egyptian had free choice to decide whether he would enslave Israel or not, and therefore God’s prior decree does not dictate each Egyptian’s individual choice, and so that Egyptian deserves punishment. If he chose to enslave Israel, then he deserves to be punished. Why? But it was already said in advance that the Egyptians would enslave Israel. So Maimonides says that the statement was a collective one: the Egyptians as a whole will enslave Israel. Each Egyptian individually—that’s his decision whether he’ll do it or not. So in effect, according to Maimonides, human behavior is indeed treated—or should be treated—by probabilistic means. In other words, if God determined, say, that there was an eighty percent chance, He tilted the balance in such a way that each Egyptian had an eighty percent likelihood of enslaving Israel, then the assumption is that among all the Egyptians each one has free choice, but eighty percent will enslave Israel. Let’s say if the balance were fifty-fifty, then I’d assume that each one again had free choice, but the distribution would be fifty-fifty: fifty percent of the Egyptians would enslave, fifty percent would not. If God tilted the balance—yes, hardened Pharaoh’s heart, tilted the balance—now it’s eighty-twenty for each Egyptian, then still each one has free choice, but when we look at all the Egyptians, eighty percent will enslave Israel and not twenty percent. So it turns out that the law of large numbers works even on actions that are products of human decision. In other words, if you go to large numbers and the decisions are independent, then we still talk about using probabilistic tools, even though the events we’re treating are human choices, human decisions, not random events. And in fact that’s what the Raavad says against Maimonides; that’s really what stands behind the Raavad’s words. The Raavad argues: if the Holy One, blessed be He, did not decree upon any specific Egyptian, only upon the Egyptians as a whole, then it could happen that every Egyptian individually would choose not to enslave Israel, and then in practice it would come out that no Egyptian would enslave Israel, and God’s prophecy—in the Covenant Between the Pieces to Abraham our father—would turn out false. So how can God know in advance? After all, each Egyptian has free choice. What is the Raavad really saying? He is basically saying: I’m not willing to accept the application of the law of large numbers to human behavior. It’s their choice. If this Egyptian chooses the good and doesn’t enslave Israel, and the second Egyptian, and the third, and all the Egyptians choose not to enslave Israel, then in practice no one will enslave Israel. You can’t decree something about the collective that doesn’t find expression in the particulars. The collective is nothing more than the sum of the particulars, and if each individual makes decisions independently, then it may also happen that they all decide not to enslave Israel. So there is some dispute here over whether we can. All of us, in our simple intuition, it seems to me, think this way too. This isn’t some crazy notion of—
[Speaker B] But that’s because modern psychology understands that there is no free choice at all. There’s determinism exactly as with a die. Even with a die we know it’s not random; it’s deterministic.
[Rabbi Michael Abraham] I don’t know what modern psychology thinks; I know what I think. And I think that a person has free choice, and nevertheless I think about these events using statistical tools. Among psychologists too, some think this way and some think that way. But I don’t think it’s connected to that. It could be connected to that, but I’m claiming there are people who don’t think that and still would apply probabilistic tools there. Like me, for example.
[Speaker B] But with a die, the Rabbi agrees that in truth it isn’t random, it’s deterministic.
[Rabbi Michael Abraham] Yes, but that’s the previous example. Still, it isn’t an action of decision. In that sense it is random. It’s random like the Talmudic text in Makkot 10—I think I brought this—where someone who kills another unintentionally, the assumption of the Talmudic text there is that if he killed unintentionally, apparently the other one deserved to die and the Holy One, blessed be He, arranged it. But if I kill intentionally—
[Speaker D] That’s at the assumption stage, not the conclusion. What? That whole passage in Makkot is only at the level of assumptions.
[Rabbi Michael Abraham] I didn’t understand. The Talmudic text in Makkot? Why? That is the conclusion. Why not? Of course it is. That’s what the Talmudic text says. The Talmudic text says: what is the case of unintentional killing, what is intentional killing, what is unintentional killing? The Holy One, blessed be He, brings them to one inn, and that’s the whole thing. And if you kill intentionally, then the murdered person may be someone who did not deserve to die. That’s the Talmudic text in Chagigah: “there is one swept away without justice.” And why? Both actions are actions of human beings, but the unintentional killer—that’s just a person acting in some ordinary way, it’s not the result of a decision. He didn’t decide to kill that particular person down below. By contrast, the intentional killer decided to kill the person he murdered. When a person decides, that’s a decision in which he acts as an agent, yes, a doer. That is, he acts as someone who makes decisions. Sometimes a person does not act as an agent; he acts like a physical entity. That is, he doesn’t decide what to do—what happens, happens. So the Talmudic text says: in the second mode it’s a decision of the Holy One, blessed be He; in the first mode it’s a decision of human beings and not of the Holy One, blessed be He. And if by “the Holy One, blessed be He” I mean statistics—yes, nature, natural process—then I’m saying: with actions that are human decisions, there are no statistics. And with actions that are not human decisions, there are statistics, because they’re actions that happen to them, actions that happen to them. But here, when we talk about probability in psychology, we’re applying probabilistic tools to things that are indeed human decisions. Decisions whether to react violently or not react violently, decisions—it doesn’t matter, any psychological phenomenon whatsoever. We apply probabilistic tools.
[Speaker B] If we assume that God had said that in an aggregate, statistical, probabilistic sense I created a situation where 99.9% of the Egyptians will enslave Israel and a tiny fraction of a thousandth of a percent won’t, and there is free choice, which the Rabbi maintains—okay, then the Rabbi would still say fine, there’s still free choice; Maimonides at least would still say that each one has free choice. Now all the Egyptians come as a nation—right, the Rabbi has spoken elsewhere about essence—
[Rabbi Michael Abraham] No, no. If you’re talking about a decision of a nation, then the discussion doesn’t even begin. If you’re talking about a decision being made as a single decision, then there is no law of large numbers because there is dependence between the events. The law of large numbers exists only when there’s no dependence.
[Speaker B] Meaning, God has no claim against the Egyptian nation, the Egyptian people, only against each individual.
[Rabbi Michael Abraham] That’s Maimonides’ argument. Leave theology aside for the moment. That’s what Maimonides writes, yes; he talks about a discussion concerning each Egyptian individually. Obviously, if the decision is a collective decision, then that’s not the law of large numbers. That’s simple.
[Speaker B] But it’s interesting that God’s punishment, for everything they did, wasn’t directed at individuals; it was directed at the nation. He drowned the whole army; He didn’t check each one and ask what he really would have chosen.
[Rabbi Michael Abraham] Here there’s room to discuss His goodness, His justice, when He administers collective punishment. Not important—that’s maybe questions on Maimonides. Right now I’m talking about what Maimonides says. So the claim, in the final analysis, is that contrary to what may seem at first glance, even actions that are actions of human decision, we still use probabilistic tools to describe them, to deal with them. And that basically means—and I said there’s a model that might perhaps describe this matter—that there’s the deterministic attack on libertarians, where they say to them: look, genetics determines our traits and all sorts of things about us. The brain, genetics, it doesn’t matter, the environment, the circumstances determine what we will do. Therefore obviously a person has no free choice. The world is deterministic. Now that attack basically assumes some very childish, naive libertarian conception. You think the libertarian imagines that a person acts in a vacuum. There are no influences on him. But here we see there are influences, so obviously you’re wrong. But an intellectually serious libertarian doesn’t think that. An intellectually serious libertarian understands that there are influences from genetics, from the environment, from the brain, from all kinds of such things. That’s obvious. The whole difference between him and the determinist is that in his eyes the influences do not determine the result but influence the result. But the person can still decide not to go with those influences, to go against them, to veto them, as I mentioned with Libet’s experiment and the imposition of a veto. What?
[Speaker B] Meaning, the influences will show up in the statistics, but in the end they won’t determine, they won’t negate the person’s final choice. Right. So I’m saying: if we had tools to examine and make deeper and deeper resolutions of those same influences, then does the Rabbi agree that little by little the statistics would become more and more precise, and little by little choice would keep shrinking?
[Rabbi Michael Abraham] No, no. Absolutely not. I claim that in contrast to the deterministic view, which says that the complexity of psychology, our inability to predict psychological behavior, simply stems from complexity—but if we knew all the psychological laws and all the structure of the person before us, we could say exactly what he would do at every moment—I, by contrast, say no. I say that even if we knew all the motives and phenomena and his entire psychic structure and all the laws of psychology and everything, we still could not predict with certainty what he would do.
[Speaker B] But does the Rabbi agree that the asymptote would keep getting closer and closer to the infinity of absolute precision, only that there would still remain some tiny fraction—
[Rabbi Michael Abraham] I’m claiming that there is the law of large numbers. The law of large numbers says that if you do this with a million people, then the distribution will match the statistical distribution. But a specific individual has free choice.
[Speaker B] So I’m saying: if the general distribution says there’s a 99.999 percent chance that you’ll choose this way—does the Rabbi agree that theoretically it could arrive at such a level of precision? So what is the meaning of this person standing there with this detached free choice of one in a billion?
[Rabbi Michael Abraham] Shmuel, Shmuel—about the group, not about an individual person.
[Speaker B] Yes, but they tell him: listen, there’s a chance—but the chance is one in a billion—please choose.
[Rabbi Michael Abraham] No, you don’t have a one-in-a-billion chance. You have a chance of whatever you choose. That’s exactly the difference. No—you’re once again jumping from a distribution over a group of people to a probability about an individual person. That jump is invalid.
[Speaker B] But he asks us before he chooses—we’re talking with him an hour before. He says to us: tell me, can you make a statistical estimate of what I’ll choose? I myself don’t know what I’ll choose. They tell him: listen, statistically—does the Rabbi agree probabilistically?
[Rabbi Michael Abraham] No. No, I don’t agree. I agree that if there were a million people here, I’d claim that 999,999 of them would behave this way. But when a single person stands before me—I don’t know. It depends what you like, red or black, like with the balls in the jar. A person comes before me and says, look, I don’t know what my taste is, all right? I have some taste but I don’t know, I forgot it; only when I see the balls will I remember what I actually like, red or black. Now I need to predict whether he’ll choose red or choose black.
[Speaker B] We know his taste—we did the precise investigation on him and reached amazing resolutions in the genome, in the—
[Rabbi Michael Abraham] No, again, I’m not talking about an investigation. Again, if you’re talking about an investigation then that’s something else; maybe I could give a statistical estimate even for an individual person, not with certainty but statistically. But if I’m looking at the collective and I say there are 999 out of 1000 who behave this way, that does not mean that the probability for the individual person is 999 out of 1000.
[Speaker B] But when I do statistics on Israelis and say, are you Israeli or Chinese? He says, I’m Chinese—then the statistics isn’t relevant to you. Obviously I look at the specific person, examine his characteristics. That’s not right, Shmuel, simply not right.
[Rabbi Michael Abraham] I’m looking at, let’s say, in my world there are—let’s go back to the balls in the jar, all right? Let’s say I reached the conclusion that there are 999 people who like the color red and only one who likes black. In our city there are a thousand people: 999 like red and one likes black. Okay? Now a person comes before me and says, look, I don’t know what my taste is, all right? Can you predict what I’ll choose? I don’t know. You ask him where he was born?
[Speaker B] He says, I was born where all the 999—
[Rabbi Michael Abraham] Not all—the majority.
[Speaker B] The majority, where the majority were born, that’s where I was born; I’m from the same people.
[Rabbi Michael Abraham] I don’t know. But if your taste is among those who like black, then you’re among those who like black.
[Speaker B] Only if all human beings came before—
[Rabbi Michael Abraham] before me, then I’d say that 999 of them would choose red.
[Speaker B] But the Rabbi would certainly put a lot of money on the fact that he belongs to the group of—
[Rabbi Michael Abraham] Belongs to the group—that’s already a subtler question. What do you mean, belongs to the group? If I now take a random person, what exactly is the specific question? And I ask the Rabbi: if I now take a random person from the group of a thousand and ask whether I drew a person who likes red or a person who likes black—here the probability is 999 out of 1000.
[Speaker B] Right, so you know—
[Rabbi Michael Abraham] But a specific person comes before me—I didn’t draw him randomly. A specific person comes before me and I have no information about him. I don’t know. He may belong to the black, he may belong to the red.
[Speaker B] So it’s all just a problem of information. If we gather all the information, narrow the information space—
[Rabbi Michael Abraham] But there is no information. Again, if we gather the information we’ll have information, but right now we don’t have information. We only have collective information.
[Speaker B] That’s what we have. But the Knower of secrets knows all the information. The Knower of secrets knows what God chose. He knows exactly what he chose. No, but we’re talking about divine choice—not in our eyes what he’ll choose, but how God should judge this Egyptian.
[Rabbi Michael Abraham] If I knew his psychic structure all the way through, then I would apply the laws of probability to him as well—I agree. But I’m saying that I don’t know his structure. All I do know is what happens among a thousand people.
[Speaker B] Right, but God knows his structure.
[Rabbi Michael Abraham] Fine. So if He knows, let Him do what He wants. But even there it’s not certain prediction, only statistical. But fine, the Holy One, blessed be He, can give a statistical prediction; I can’t.
[Speaker B] Fine, but regarding the Egyptians, we’re not the ones supposed to judge and punish them—God is. He knows everything, to the infinite limit of knowledge. So what is left for him to choose?
[Rabbi Michael Abraham] No, what do you mean he still had the ability to choose? He still had the ability to choose that despite his character being to enslave the Jewish people, he could have chosen not to enslave the Jewish people. Even though statistically his character is nine hundred and ninety
[Speaker B] and nine point nine nine nine percent?
[Rabbi Michael Abraham] Right, but there was still a chance.
[Speaker B] But he could have chosen otherwise. So that makes the choice very easy; it makes the punishment very light, as it were.
[Rabbi Michael Abraham] So he has arguments for leniency in punishment. Right. But he still could have chosen, that’s all.
[Speaker B] Absolutely. There remains a fraction of a percent of…
[Rabbi Michael Abraham] Shmuel, we’re just arguing for no reason. These are arguments for leniency in punishment. The question whether he has free choice or not is a different question.
[Speaker B] Yes, but I’m saying it’s turned into some kind of fiction of…
[Rabbi Michael Abraham] Fine, let’s move on. We really are repeating ourselves.
[Speaker E] Rabbi, can I ask one more question? Yes. After we really understood the difference between the two things, and that you really have to decide between them—if we go back to that Talmudic passage about the gifts and betrothal, then I understand that apparently you can’t rely on what most of the city does or something like that, but rather on his choice. And because we don’t know what his choice is, then it’s… But I’m asking myself: if that’s not how we decide, then how do we decide? What else is left to us?
[Rabbi Michael Abraham] Nothing is left. Regarding human actions, you can’t decide statistically—that’s exactly the claim.
[Speaker E] So then how do you decide?
[Rabbi Michael Abraham] It’s a doubt, so you remain in doubt.
[Speaker E] And then how do you decide? There are laws of doubt.
[Rabbi Michael Abraham] So she is of doubtful betrothal.
[Speaker E] She is doubtfully betrothed? Yes. Meaning, we actually don’t decide.
[Rabbi Michael Abraham] Correct. The laws of doubt are part of Jewish law.
[Speaker E] Because here you can’t rely on the majority, so we remain in doubt. Right. And that’s how the Talmud rules as well?
[Rabbi Michael Abraham] Yes, that’s Nachmanides—according to Nachmanides, yes, that’s the claim. And also in monetary law, we do not follow the majority in monetary matters. What does it mean that we do not follow the majority in monetary matters? And according to Rabbi Shimon Shkop, who explained that the reason we do not follow the majority in monetary matters is because in the case of plowing-oxen, it’s a majority based on human decision. So what remains? The laws of doubt remain. And what are the laws of doubt? The burden of proof is on the one who seeks to extract money from another.
[Speaker E] Meaning, that’s the practical difference, basically: every time there’s a decision that is human, you can’t rely on the majority of the population, and therefore we…
[Rabbi Michael Abraham] Yes. There are some disputes about this, but according to the approaches I presented here, then yes. I said there are disputes about this among Tosafot; there are disputes among the medieval authorities (Rishonim). But yes, there is such an approach among the medieval authorities, and that’s how they explain the Talmud—that human decisions are not subject to probabilities. I said that the second opinion can be understood in a slightly different way, and that will also be important later. I can say that if I were now, say, choosing a random person in this city and asking myself whether he sells oxen for plowing or for slaughter, then I would say he probably sells for plowing, because on the collective level there are nine hundred people out of the thousand who sell for plowing. So if I pull a person at random, it’s likely I drew him from the group of nine hundred and not from the group of one hundred. Therefore those who argue that even in actions that are matters of human decision there is room to follow the majority—that is probably the position of the medieval authorities who disagree. But the medieval authorities who say no claim: I’m not pulling a random person here; I’m talking about a specific person, and that person says, I belong to the minority—what do you want from me? You didn’t pull me at random. Exactly.
[Speaker E] The difference is really in his claim, basically. Just—
[Speaker F] So that I understand: when I pull a person at random, why then do I still not ask who says—meaning, maybe he’s a person whose nature is always to choose the red ball?
[Rabbi Michael Abraham] Because if I pull someone—I know that in the group there are nine hundred people who pick red and one hundred people who pick black, that I know. Okay? Now I drew a person at random from the group. What is the chance that he belongs to those who pick red? Ninety percent. Yes, ninety percent.
[Speaker F] Right. But then what’s the difference? I mean, basically, even if the person now arrived and I didn’t draw him from a group…
[Rabbi Michael Abraham] Then I don’t know. He was not drawn from the group at random; I have a lack of knowledge about him. It’s not the same thing. You assume—and that’s the simple assumption—that even in such a case I can assume it’s like a random draw of the person. No. It could be that the person who arrived now is a person who belongs to the minority.
[Speaker F] Okay, just so I understand: the person standing before me—if I now bring another nine hundred and ninety-nine people and then choose him, would that then count as if he was drawn at random?
[Rabbi Michael Abraham] Then you haven’t defined the question well. What does it mean, I’ll bring another nine hundred and ninety-nine? There’s a person before me. How does it help if I now bring another nine hundred and ninety-nine?
[Speaker F] I’m asking what the difference is. As I understand it, I seem to have two possibilities here, and I’d be happy to understand what the rabbi means. Either the problem with the person I brought now not at random is that the non-randomness can affect what he will choose, that maybe the fact that he is here shows—
[Rabbi Michael Abraham] indicates the course of what he’ll choose…
[Speaker F] It doesn’t affect it; he’ll choose what he chooses.
[Rabbi Michael Abraham] The non-randomness means that I cannot apply statistical tools to him.
[Speaker F] What’s the problem—that I don’t have a reference group? Is the problem that I don’t have a reference group?
[Rabbi Michael Abraham] No. The problem is that we are dealing with an individual person, and on an individual person you can’t apply statistical tools. Only if I draw him at random from a group. In other words, the difference between fixed and separated—that was exactly the explanation I gave a few lessons ago, the difference between fixed and separated. A piece of meat that separated from shops—then something random happened here. If something random happened, then my assumption is that it separated from the majority group, from most of the shops. But if it is a case of fixed status, then I go to the shop and take out a piece of meat. Here it depends on the question of which shops I like to enter, and therefore this is not a random separation. Once there is no random separation, then I do not follow the majority.
[Speaker F] Okay, I’m asking about that too. Assuming I know that out of every thousand people I choose in this city, nine hundred people will choose the red ball—so I said, if before me… then what is the problem if I bring here nine hundred theoretical straw men…
[Rabbi Michael Abraham] You can’t posit another nine hundred and ninety-nine theoretical straw men. The question is not defined. I’m explaining again: there is no such thing as theoretical straw men. Show me the experiment you are performing. You’re saying words, but those words have no meaning. What are theoretical straw men? If you chose him at random from a thousand people, you are right. But if he appeared before you not as the result of a random choice, then you cannot assume that it’s ten percent. So what helps me if you now dress up another nine hundred silhouettes here? It’s not interesting. Show me the experiment you are performing.
[Speaker F] To describe the experiment? I’m simply taking one person. The first experiment is with the original thousand people I took five years ago. That is, one time I took a thousand people and saw that nine hundred of them choose the red ball. From that I make a deduction for every thousand people under roughly the same conditions, that again nine hundred of them will choose the red ball. And then when one person comes to me, I say that whatever group he may be from, it’s as if—if there were another nine hundred and ninety-nine people here—he would…
[Rabbi Michael Abraham] No, no, that’s wrong, that’s wrong. Why? Because about the person standing before you, you do not say, whatever group he may be from. He is not from any group. Since from every group he comes from, you are assuming that his separation was random. But no! He did not separate randomly, so it is not relevant which group he belongs to. Only someone who separates randomly from the group—you need to ask yourself from which group he separated. But if he did not separate randomly, then the group is not relevant.
[Speaker F] But is he still considered as having separated from a group? Because he separated from some group…
[Rabbi Michael Abraham] No, no, no—he didn’t separate.
[Speaker F] He separated not randomly, but he arrived
[Rabbi Michael Abraham] now from some city where there are a thousand people… That is not called separating. If it isn’t random, it isn’t called separating. Separation is…
[Speaker F] What is the logic behind this rule, that if he
[Rabbi Michael Abraham] doesn’t arrive randomly, it isn’t called separating? Very simple logic. Once a person does not separate from a group randomly, then statistical tools cannot be applied to it, so what do I care that you can call it non-random separation? What do I care? It’s like… a die roll that is deterministic always lands on six, okay? I throw the die. Why in the world should the chance be one that it lands on six? After all, it’s one in six. No, because it’s not a random event. It always lands on six. So what do I care that it has five other shadow-faces? That’s not interesting. It always lands on six. It is not a random event. Only a random event is discussed as separation from a group, and then you can apply statistical tools. The person standing before you did not separate randomly from anywhere. And therefore could it be that his reference group would all choose the black ball, say? There is no reference group; there is no reference group. The concept of a reference group is not relevant to him. There isn’t one, there isn’t. A reference group is when you have separated randomly from some group. Then you can say: you belong to that group, you separated randomly, so your distribution is uniform across the whole group, and then I apply probabilistic tools. If you did not separate randomly, there is no such thing as a reference group. So to whom do I… but yes,
[Speaker F] we agreed to disagree.
[Rabbi Michael Abraham] Let’s continue. So in the end, what I actually want to argue is that there is no contradiction between these two claims. On the one hand, I said that events of choice are not subject to statistical tools; on the other hand, we see that in psychology it does work. The answer is: in psychology we are talking about what a group of people will do. We are not talking about what one individual will do. We are talking about what a group of people will do. And since I am saying what a group of people will do, I can speak in statistical terms, even if the actions are actions of choice. But if I am talking about what one individual does—then one individual, if the actions are actions of choice, there is really no room to apply statistical tools. That is basically the claim. And therefore, in another context, I also wrote—and I think I also spoke about this in one of the series—that in general I have a problem with psychology. Yes, I’m in good company; Popper also had one, at least with psychoanalysis. But my trust in that discipline is very limited. It’s certainly not science, let’s say—that’s clear. But when I say it’s not science, I’m talking about treatment of an individual person. When someone treats an individual person, that is not a scientific action. That is, you cannot do it with scientific tools. It is much more an art than a science. But basic psychological research—not the action on a single person, but the research that does statistics on groups of people and finds correlations between phenomena, or action and reaction, or things of that sort—that is a scientific field. Why? Because there I really am applying statistical tools. About groups I can speak statistically. But a specific individual who stands before me—I cannot say there is a ninety percent chance that he will respond this way. It depends how he is built, whether he is this sort of person or another sort of person. And therefore I cannot apply statistical tools, and so the therapeutic field is not a scientific field. The field of psychological research—assuming it is done properly and so on, which many times it is not—but assuming it is done properly and everything, that is a scientific field. Because there you work with statistical tools, and if you do it systematically and in an orderly way, then fine, there is no problem. And that is exactly the same distinction. In other words, the difference between applying statistical tools to a large group of people, even when the actions are actions of decision—that, yes. If you are talking about one individual, then in an action of decision he decides what he decides; there is no point in treating him with statistical tools. Okay, so that is regarding the…
[Speaker F] And also when it comes to a collapse range, statistics applies even to a single particle.
[Rabbi Michael Abraham] Again, I didn’t understand.
[Speaker F] When it comes to a collapse distribution—I don’t remember the name of it—of particles.
[Rabbi Michael Abraham] Exactly, that is precisely the proof of my point. Even to a single particle statistics applies. That is exactly the proof of what I’m saying. Why? Because a particle does not make decisions. For a particle, this is a random process. In a random process, that’s exactly the difference. In a random process, the statistics over many particles, or the distribution among many particles, basically means what the statistics is for the single particle. That is exactly what I’m arguing: when we are dealing with actions that are actions of decision, not random actions, then you can do statistics on the group. But that does not mean that the probability for the individual is ninety percent, say, if in the group it is distributed ninety-ten. That is exactly the claim. Okay, so let’s—what I want now is really to move to the next stage. What I want to do now is talk about—and I’ll make use now of the distinctions I made here—but actually this is already the next chapter in the series. I want to talk about statistical evidence in law. Again, there are two columns on my website that deal with this, 226 and 228. In any event, this topic of statistical evidence in law—I dealt with it through a case that appears in an article by someone named Papino from the Alaxon website; the article appears there in my column. And Papino basically suggests the following example. Imagine there are one hundred prisoners exercising in the prison yard. Suddenly ninety-nine of them attack the guard supervising them and kill him. One does not. That is, ninety-nine out of the hundred attacked this guard. Now they want to try the prisoners for murder. Now one of the prisoners is sitting there before the judges, and the judges have no specific evidence—specific evidence against him. But still, the claim is that the probability he is guilty is ninety-nine percent, because ninety-nine out of the hundred were partners in this murder. Only one was not. So basically the probability that this prisoner is guilty is ninety-nine out of a hundred. Can he be convicted on that basis of reasoning? On the face of it, ninety-nine percent is a good probability—I mean, for a criminal trial, ninety-nine percent is a good probability.
[Speaker E] But why here do we apply probability?
[Rabbi Michael Abraham] I don’t know; I’m asking whether he can be convicted.
[Speaker E] No, it seems to me the same thing.
[Rabbi Michael Abraham] The same as what?
[Speaker E] Like the gifts?
[Rabbi Michael Abraham] Ah, wait—you’re already applying what we saw. I’m opening a new page now. We’ll come back to those arguments, but first let’s think for a moment about this case on its own.
[Speaker B] It depends—if he lives in Gaza and there everyone murders everyone, then maybe it’s not unusual.
[Rabbi Michael Abraham] Okay, your protest has been noted, Shmuel. In short, in such a case there is consensus among legal scholars that he is not convicted. He is not convicted. That is a given, okay? Not a given that it is correct—a given that this is what legal scholars think. Okay? Now there is another case, okay? There is a situation in which two witnesses come to court and testify that Reuven murdered Shimon. Okay? Now we know that people’s vision is sometimes disappointing, yes? It is not always accurate. Let us assume for the sake of discussion that under the circumstances prevailing there, there was—let’s say—ninety-seven percent chance that the witnesses saw correctly, but a three percent chance that they missed it. I think anyone familiar with this—I’m speaking with lawyers, my wife is also a lawyer—it’s unbelievable, you never hear the same testimony from two people. That is, testimony is something far less reliable than laypeople think. In the courts, surprising things become clear in this regard. In any case, eyewitness testimony came before us, and let’s say for the sake of discussion that there is a three percent chance these witnesses got it wrong, didn’t see well. Okay? Now, do we convict the murderer on the basis of their testimony or not? It is commonly thought that yes. Yes, we convict. Now notice: here it is ninety-seven percent that he murdered, and we convict. In the previous case it was ninety-nine percent that he murdered, and there we do not convict. Why the difference?
[Speaker D] Because there is no proof at all. Why the difference?
[Rabbi Michael Abraham] There is no proof anywhere. The question is what percentage I am right.
[Speaker D] No, when two witnesses come, there is room to say yes, because it is supported, whereas there there is one where for sure one person was outside and there is no proof whatsoever against the person—
[Rabbi Michael Abraham] that he belonged.
[Speaker D] Now you pulled another joker.
[Rabbi Michael Abraham] That is, first you started by saying there is no proof. There is no proof in either case. In both cases there is some probability that he murdered. Now you’re adding something else. You’re saying: in the case of the prisoners, if we go systematically and try all one hundred prisoners, then we have certainly convicted one who is not guilty. Fine, we’ll get to that later. But let’s leave that for a moment. So basically the question Papino asks there in this article is: what is the difference? From a probabilistic point of view, apparently the prisoners’ case is a safer conviction than the case of the eyewitnesses. So why, in the case of the eyewitnesses, does everyone agree that we convict, while in the case of the prisoners everyone agrees that we do not convict? There is wall-to-wall consensus on both sides of the equation, and the two sides are opposite—opposite to probabilistic logic. So why don’t we… In short, why don’t we punish on the basis of statistical evidence? That is basically the question. Now he himself—and it is common in that field to say this—his claim is that in the case of the witnesses, we have knowledge that Reuven murdered Shimon. Not certain knowledge, but knowledge. In the case of the prisoners, we do not have knowledge that this man murdered. There is some probability that ninety-nine out of a hundred murdered, but we have no knowledge at all about this man that he murdered. Now he himself, after saying this—because many legal scholars and philosophers of law, or whatever you want to call them, say this—he himself is unwilling to accept this distinction. And why? Because what do I care whether we call this thing knowledge or don’t call it knowledge, when in the bottom line I ask what the probability is that I ruled correctly. What difference does it make whether you call it knowledge or not? Now what is the probability that I ruled correctly? In the case of eyewitness testimony, the probability is ninety-seven percent. In the case of the prisoners, the probability is ninety-nine percent. So why do I convict here and not convict there? What difference does it make whether one uses the word knowledge or the word probability or whatever else one wants to use? The question is: what is the probability that the verdict is correct? On the face of it, that is what ought to determine it. Okay? Therefore he says he is not comfortable with this distinction between yes-knowledge and no-knowledge. Now there is—I’ll bring a similar halakhic / of Jewish law example. Maimonides in the Laws of the Sanhedrin makes a distinction between monetary law and capital law. In monetary… I’ll share the screen for a moment. So, chapter 24 of the Laws of the Sanhedrin: “A judge may judge monetary cases according to matters toward which his mind inclines that they are true, and the matter is strong in his heart that it is so, even though there is no clear proof there.” Yes, in monetary cases the judge should rule according to what his heart tells him, according to what appears to him to be true, regardless of the formal laws of evidence. “And needless to say, if he knows with certainty that the matter is so, that he judges according to what he knows. How so? If a person became obligated to take an oath in religious court, and a person whom the judge trusts and whose words he relies upon told the judge that this man is suspect regarding oaths, the judge may transfer the oath to the opposing litigant, and he will swear and collect, since the judge relied on this person’s words. Even if it were a woman or a slave whom he trusts, since the matter was found strong and correct in his heart, he relies on them and judges.” Okay? What is Maimonides saying? The source of Maimonides’ words is actually a Talmudic passage. The Talmud speaks there about Rava, who judged a certain case, and a woman came before him and he wanted to make her swear. Then Rava’s wife entered the court—back then the court was in the judge’s house—and she came in and whispered to him: listen, this woman is a pathological liar; you shouldn’t make her swear. What did he do? He transferred the oath, meaning he imposed the oath on the other litigant, and he swore and took the money. Yes, the woman could have sworn and been exempt, but he says to him: you swear and receive the money. Why? The woman is not a valid witness, and certainly when she is his wife, a relative of the judge—the witness is related to the judge—so she is not valid for testimony. How can it be that he extracts money without having proof? So from here Maimonides learns—and this already begins with the Rif and others—that in monetary law, the formal laws of evidence are not binding. The judge must be convinced that this is the truth. Once his heart is convinced that this is the truth, that is what he should rule, and it does not matter whether it fits the formal laws of evidence or not. That is not what matters in monetary law. By the way, this is exactly what I spoke about this morning; I don’t know how I reached the same topic in two lessons on the same day. I just remembered I said something this morning about one witness being believed in matters of prohibition. People often think that one witness is believed in matters of prohibition is a rule of testimony, just as there are two witnesses in monetary matters or capital matters or matters of sexual prohibition. So in matters of prohibition, one witness is believed. But it is quite clear from the Talmud, and even more from the medieval authorities (Rishonim), that this is not really a category of testimony. What is really required in matters of prohibition is that I become convinced that this is the case. If that happens through one witness—and it doesn’t matter if he is a woman, a slave, a minor, whoever it may be—if it happens through him and he convinced me, then even though in terms of formal categories of testimony one witness who is a woman is not valid, still he convinced me, so that is perfectly fine. In matters of prohibition I can eat on the basis of his testimony. Yes, people often ask a halakhic decisor: tell me, can I trust a shop owner who testifies that his merchandise is kosher? Yes, very common questions. There is no kosher certificate, but the shop owner says, listen, it’s kosher here. Assess the situation yourself: decide whether it is reasonable to trust this person; if yes, then eat, and if not, then don’t. It’s not a halakhic question. What?
[Speaker D] In kashrut that’s not how people rule.
[Rabbi Michael Abraham] What? In kashrut—
[Speaker D] they don’t rule that way.
[Rabbi Michael Abraham] I know. That’s why I’m saying: many people are mistaken in this matter, in my opinion, and I don’t think that is correct. In the laws of permitted and forbidden food, what determines things is whether you became convinced that it is correct. If you became convinced that it is correct, then everything is fine. There are no laws of evidence in the laws of permitted and forbidden. Maybe, maybe there is room to discuss laws of evidence if you yourself were not convinced. Say one witness came and said this meat is kosher. Now you are not convinced; something about him doesn’t seem right to you, I don’t know exactly what. It may still be that you are permitted to eat because one witness is believed in matters of prohibition. But if you are convinced that yes, then it doesn’t matter on what basis you were convinced—eat, because all you need is to know that it is permitted. That’s all. At most you are taking a risk: if later it turns out that you ate something non-kosher, they will tell you, fine, you sinned unintentionally or under compulsion or something like that. But you are permitted to take the risk, because if you were convinced, then everything is fine. This is not the same as in what is called criminal law, when we are dealing with a death penalty or lashes. There, what determines things is the formal procedure, the formal laws of evidence. More than that: in an obligation of death, for example, or also an obligation of lashes, you are not obligated at all until the religious court determines that you are obligated. That is Rabbi Akiva Eiger in tractate Makkot page 5: you are not obligated at all; the court constitutes the obligation. Okay? When you look at this, let’s look for example at a loan. The court adjudicates a loan. I claim that you borrowed from me and you claim that you did not, and we come to court and the court says that you are liable because you borrowed, and you must repay. Before the court—assuming you are a liar and really did borrow—and the court did not rule, or ruled incorrectly, or there is no court at all here—are you obligated to repay me? Certainly yes, right? If you borrowed, you have to repay me. That means that in monetary law, in this sense, it is similar to matters of prohibition; it is not similar to criminal law. In monetary law, what matters is whether you borrowed. If you borrowed, then you pay; if you caused damage, you have to pay. It does not matter whether there is a court ruling or not. We go to court when there is disagreement between us and I want the court to decide in my favor and force you to pay me because I think I deserve it. If I succeed in persuading the court, then the court will compel you and you will pay me. But the fact that you owe me is not because the court decided; you owe me because you borrowed. Reality determines it, not the court ruling and not the laws of evidence. Even if there are no witnesses that you borrowed, if you borrowed, you have to pay me. So there are no witnesses—so what? Maybe if there are no witnesses you can lie and I won’t be able to prove in court that you are lying, but that does not mean you are not obligated. By contrast, if there are no witnesses that you murdered or ate pork, then you are not liable to death and not liable to lashes. You are not liable. It’s not that we don’t know you are liable. Yes, that was the difference. What—
[Speaker B] What isn’t clear to me is that these are different questions altogether. It’s like what the rabbi said to that Danish judge: if there is doubt, then you go with statistics and with statistical probability. But in principle these are two different questions. When witnesses come before the judge, when he examines their reliability, at most he is saying whether he believes the witnesses. But he is not really saying what the correct case is; he is saying whether I agree that they are not lying. But that still does not tell him—if, let’s say, suppose I ask the rabbi this: if all the Jewish people saw the murder, every Jew in the world saw the murder, physically saw it, and now judges have to be appointed to judge him. Twenty-three judges sit and have to judge the person for murder; they themselves saw the murder with virtually zero distance. What do they do now? They know with certainty from themselves that he murdered.
[Rabbi Michael Abraham] That is a discussion in the Talmud. It is a discussion in the Talmud regarding whether a witness can become a judge, and therefore in such a situation what they do is designate two judges who will serve as witnesses; the rest of the judges will receive the testimony from those two witnesses, and then—the question is whether they can afterward rejoin as judges again or not—but that is it. But you need testimony.
[Speaker D] Some kind of ritual like that?
[Speaker B] Some mystical ritual like that?
[Rabbi Michael Abraham] What do you mean a mystical ritual? Yes, you need a legal process. Without testimony he is not liable to death. You can put him in a cell, you can do whatever you want, but a death sentence does not exist until the court establishes it.
[Speaker D] It can’t establish it in this case.
[Rabbi Michael Abraham] Fine, these are halakhic facts. That is not the issue right now; I’m not…
[Speaker D] No, I’m referring to his case: everyone saw it, they can’t judge it. A cell, right, but not this—not murder, not punishment.
[Rabbi Michael Abraham] Okay. In any case, what I basically want to say is that in monetary law, in that sense, it is like matters of prohibition. That is, the judge needs to be convinced that it is true. In other words, the procedural order or the formal laws of evidence are not relevant in monetary law. By the way, the Rif, who is the one who learns this from that Talmudic passage about Rava—he is the one who innovates this, and Maimonides basically took it from him—the Rif also writes that nowadays this is no longer true. That is, after the two academies in Babylonia, after the Babylonian exile was dispersed, we now judge according to the formal laws of evidence in monetary law, because minds have diminished and the people have dispersed and whatever, all kinds of explanations of that sort. So that was the basic law, but in practice today we do not do that. Although, again, there is the law of a fraudulent case. That is, in a place where you become convinced that the witnesses are not telling the truth, and you have two witnesses in the formal sense, and you caught them in a lie, but in your heart you are not convinced—then there is an exit through a fraudulent case. You can recuse yourself from the case. You can say: I cannot rule on the basis of these witnesses; my heart troubles me. Then another judge may come—if he is convinced, he will rule. So you can’t rule however you want, but you can withdraw. Okay. That’s one side. The other side of the coin is what happens in the same situation in capital law or punitive law in general, not only capital law. That is another Maimonides. Again, I’ll share it: this is Laws of the Sanhedrin, the beginning of chapter 20. The other was chapter 24. Here it is the beginning of chapter 20: “A court does not punish based on estimation of the mind, but only according to witnesses with clear sight. Even if the witnesses saw someone pursuing his fellow, warned him, and then lost sight of them, or they entered after him into a ruin and found him slain and writhing, and the sword dripping blood in the killer’s hand—since they did not see him at the moment he struck him, the court does not execute on the basis of this testimony. And regarding this and the like it is said: ‘Do not kill the innocent and the righteous.’ Likewise, if two testified that he worshiped idolatry—this one saw him worship the sun and warned him, and that one saw him worship the moon and warned him—they do not join together. As it is said: ‘Do not kill the innocent and the righteous’—since there is some side by which to acquit him and regard him as righteous, do not kill him.” So here Maimonides is basically saying that unlike what we saw at the beginning of chapter 24, there in monetary law, in punitive law—in what is called criminal law, in capital cases or lashes—it is not so. There there are formal rules: you need two valid witnesses with prior warning and all the rules. Without that, you do not punish the person. Now this is not arbitrary. Rather, even in a case where—you entered after him into a ruin and saw the victim writhing and the sword in his hands dripping blood—that is, it is obvious he killed him, yes, but there are not two witnesses to the murder. In such a case he is not executed. The source in the Talmud is in tractate Shevuot 34 with Shimon the Righteous, I think, who saw a person chasing his fellow, entered after him into a ruin—yes, with a knife he chased him, the other fled and entered the ruin. The second one with the knife entered after him into the ruin. Then half a minute later Shimon the Righteous went in after them and saw the dead man writhing, basically dead, dripping blood, and the person who had chased him with the knife—the knife was dripping blood. Obviously he murdered him, but there were not two witnesses to the murder. In such a situation he is not executed. Now there is a dispute between Tosafot and Maimonides. Tosafot claims he is not executed because there is doubt. Who knows—maybe someone else entered there; it is not certain; it is not clear. But Maimonides, from what I read now, learns it differently. He claims there is a formal requirement of two witnesses. Circumstantial evidence cannot establish a death penalty or a punishment at all, not even lashes. Only witnesses. Now one must understand that when we speak of circumstantial evidence, that does not mean that this evidence is weaker than two witnesses. The character of the evidence is that it proceeds according to circumstances and not on the basis of direct observation by witnesses. But the level of certainty I have about the event can be very high even with circumstantial evidence, like in the case we discussed earlier. Someone entered a ruin, and let’s say there are no other openings to the ruin except that one, and no one else is inside the ruin. I entered; I didn’t see anyone come out, no one else was there, and I see a knife in the killer’s hand, and I see the other person writhing, and he entered healthy and now he is writhing and bleeding. That is, there is no doubt he murdered him. What?
[Speaker D] Shimon ben Shetach, I think.
[Rabbi Michael Abraham] Shimon ben Shetach, right. Right, right, it was Shimon ben Shetach.
[Speaker B] Rabbi, maybe this teaches that the laws of evidence in punitive law are really aimed, in their overarching purpose, at finding a way not to punish—like Rabbi Akiva, under whom no one would ever have been executed.
[Rabbi Michael Abraham] If you want not to punish, then go with Rabbi Akiva; even with witnesses he did not punish.
[Speaker B] No, he found the way to find the—you didn’t see the…
[Rabbi Michael Abraham] Even with witnesses he did not punish. But in Jewish law we do not rule like Rabbi Akiva, rather like Rabban Gamliel who disagrees with him there. After all, that is the Talmud at the end of the first chapter of tractate Makkot, where Rabbi Tarfon and Rabbi Akiva say: if we had been in the Sanhedrin, no person would ever have been put to death. Why? Because if you want to judge a murderer, then we will ask the witnesses: did you perhaps see whether there was a hole at the place of the sword? Maybe where he inserted the sword into the victim, perhaps there had already been a hole there before, and the victim would have died even without that? Then you can’t execute them. Rabban Gamliel says to them: you would thereby increase murderers in Israel.
[Speaker B] Meaning… fine, but Rabbi Akiva testifies that if I had been there, that’s what would have happened. And the rabbi ruled not like him.
[Rabbi Michael Abraham] The law was ruled not like him.
[Speaker B] So he lied? If he had been in the Sanhedrin then he… He didn’t say what he would say; he said what the Sanhedrin would say.
[Rabbi Michael Abraham] If he had been in the Sanhedrin—and he wasn’t.
[Speaker B] Is there some halakhic ruling on this matter? Clearly, it’s an aggadic statement.
[Rabbi Michael Abraham] No, not at all. It is a halakhic dispute, completely halakhic. Rabbi Akiva and Rabbi Tarfon claim that halakhically one cannot execute a murderer. They are speaking about policy, about policy… and he says that this increases murderers in Israel.
[Speaker B] Fine, maybe that’s good, no?
[Rabbi Michael Abraham] But he says that this increases murderers in Israel. Meaning—after all, Rabbi Akiva was descended from converts; he did not sit on the Sanhedrin.
[Speaker B] Maimonides explicitly ruled in this dispute.
[Rabbi Michael Abraham] Maimonides, whom we just read, ruled it.
[Speaker B] No, did he explicitly address Rabban Gamliel and Rabbi Akiva?
[Rabbi Michael Abraham] Maimonides rules that if there are two witnesses, they execute the murderer on the basis of clear sight.
[Speaker B] No, but Rabbi Akiva doesn’t say not to.
[Rabbi Michael Abraham] Rabbi Akiva simply says not to.
[Speaker B] No, but he says it in language…
[Rabbi Michael Abraham] Not in any language—Rabbi Akiva says not to.
[Speaker B] That’s a very simplistic interpretation of the aggadah.
[Rabbi Michael Abraham] It’s not interpretation; that’s what he says. It’s not interpretation.
[Speaker B] No, I read it completely differently. He is saying that punishment policy is not a successful policy, and I would do everything not to punish but rather to educate.
[Rabbi Michael Abraham] The Talmud says that we ask them whether they saw that there was a hole at the place of the sword. That’s all. He says it explicitly. It’s not interpretation. So he does not execute a murderer, period.
[Speaker E] Rabbi, when the rabbi said that you cannot rely on circumstances but only on witnesses, what happens in a case where there was a camera at the scene of the murder?
[Rabbi Michael Abraham] Good question. Contemporary halakhic decisors already discuss this issue—whether a camera has the status of witnesses. There is room to discuss it. I think yes, but I don’t know; there is room to discuss it.
[Speaker D] There are only decisors who don’t want to rely on it.
[Rabbi Michael Abraham] DNA and cameras and all kinds of evidence of that sort, which we consider certain evidence—but DNA is pretty clearly circumstantial. A camera can really be seen as like witnesses, because just as I see through glasses I can see through a camera, fine, but bottom line I see. So there is room to discuss it. In any case, for our purposes, what I want to say—wait, in the age of AI… the distinction Maimonides makes between monetary law and capital law is that in capital law, circumstantial evidence is not accepted. Now notice that this is a fairly similar division to what we saw. Circumstantial evidence, which can be one hundred percent or ninety-nine percent or whatever, is not accepted; but eyewitness testimony, which may be only ninety-seven percent or ninety-five percent, is accepted. In other words, the point is that unlike what Tosafot says, according to Maimonides this is not a distinction based on the quality of the evidence or on the probability that I am right, but on the character of the evidence. Circumstantial evidence is not accepted; witness testimony is accepted.
[Speaker C] How does the rabbi explain the rule that we stone and burn on the basis of presumptive status? I didn’t understand. The rule that we administer lashes, stone, and burn on the basis of presumptive status. There we don’t have witnesses—meaning, there the witnesses come afterward regarding the fact that he did an act, but about the relationship between them, let’s say, we know it only because we saw them together.
[Rabbi Michael Abraham] That is Maimonides—Maimonides himself brings the famous example of what happens if there is one witness about a piece of meat that it is pork, and then I eat it in front of two witnesses. Do they give me lashes? After all, on the basis of one witness you cannot administer lashes. Maimonides says no: once the one witness established that it is pork, and one witness is believed in matters of prohibition, if there are two witnesses that I ate it, I receive lashes. Now also in the rule that we administer lashes, stoning, and burning on the basis of presumptive status, presumptive status surely means, according to Maimonides, the presumptions that establish the underlying framework. But the act of the transgression itself requires two witnesses in order to punish. The framework—for without it, that is, the framework that establishes that this is the father—say I cursed my father. Fine? One who curses his father or mother, or strikes his father or mother. So if there is a presumption saying that he is established as my father, then he is my father. But clearly, in order to execute me, there must be two witnesses that I did it. The presumption is not enough to execute me.
[Speaker D] And also prior warning.
[Rabbi Michael Abraham] Yes, of course, all the rules. Yes. So the claim in the end is that here in Maimonides we see a principle that is pretty similar to what we actually saw in Papineau’s article. There’s a difference between circumstantial evidence and witness testimony, and that difference is not necessarily because circumstantial evidence has a lower probability. No. Sometimes the circumstantial evidence has a higher probability and still it isn’t accepted. Now, usually when we see a distinction like that, the first tendency is to say, fine, it’s a Scriptural decree. A Scriptural decree. Which in other words means something without logic. But that’s a very problematic statement, because most of these laws, and maybe all of them, are not actually laws explicitly written in the Torah. The verse “By the testimony of two witnesses a matter shall be established” is written about monetary law. And in matters of forbidden sexual relations they learn “matter-matter” from monetary law, and there’s also a verse about murder, about capital law. But who says it means only two witnesses? And that circumstantial evidence on the level of two witnesses is not accepted? The fact is that in monetary law it is accepted. And there too it says, “By the testimony of two witnesses a matter shall be established.” Meaning, if the sages say that it means specifically witnesses and not circumstantial evidence, that means they probably had some reasoning or some logic behind this distinction. I’ve said this more than once: many times when the sages expound a verse or interpret a verse, we tend to think it’s a Scriptural decree. But that’s usually not true. Something written explicitly in a verse, you can say is a Scriptural decree. But something learned from an exposition or interpretation is never a Scriptural decree in that simple sense. Since, for example, “You shall fear the Lord your God” is expanded to include Torah scholars. So the word-marker comes to include, right? But why does that word-marker include Torah scholars and not telephones? “You shall fear the Lord your God” to include telephones. Why not?
[Speaker D] Because then there wasn’t anything else.
[Rabbi Michael Abraham] Because logic says that if I’m already including something, it makes more sense to include Torah scholars than telephones, right? Meaning, there’s a rationale behind this derivation even though the law does not emerge from the rationale by itself. The law emerges from the exposition. But obviously the exposition cannot give me the law unless the rationale says there is logic in this law. Okay? And here too it’s the same thing: when the sages say—and actually Tosafot does not accept this distinction, but Maimonides does—that circumstantial evidence is not accepted but two witnesses are accepted, they have some reasoning behind it. It is not a Scriptural decree in the sense of a law without a reason. Because it’s not written in the Torah. If we didn’t have some reasoning behind it, then we would interpret two witnesses as meaning two witnesses or any evidence on the level of two witnesses. Two witnesses would just be the example chosen for good evidence, but any other evidence equivalent to two witnesses would also be good. So it is clear that if the sages said specifically two witnesses and not circumstantial evidence, that means they had some logic for distinguishing between two witnesses and circumstantial evidence. By the way, I think I said exactly this—now I remember—when we talked about the law of fixed status. “And he rose against him and lay in wait for him” is the verse from which they derive the law of fixed status, in the lecture a few sessions ago when I spoke about the law of fixed status. And I said that it cannot be that the law of fixed status is a law without a reason, a Scriptural decree, as it is usually understood. That cannot be right. If it were so, they would not have learned it from the verse “and he rose against him and lay in wait for him.” Why single out fixed status specifically? Why not single out specifically someone whose name begins with the letter B? Because there is some logic to distinguishing between fixed status and separation; there has to be some logic. Otherwise I would not derive this thing from the verse. If it is written explicitly, then even if there is no logic, that is what the Torah says, so I accept it. But if it is something the sages interpreted or expounded from a verse, then there must be some kind of logic in it. Because without there being logic in it, we wouldn’t say it. And the same in our context: when we make a distinction between circumstantial evidence and witness testimony, since that distinction is not written explicitly in the verse, it is clear that the sages who made this distinction had some understanding behind it, some logic. The only question is that this understanding itself can be of several kinds. One kind is probabilistic understanding, which is what Tosafot understood: that circumstantial evidence is simply weaker. If there were hypothetically circumstantial evidence that was one hundred percent, then Tosafot says you would convict on that basis too. That is Tosafot’s position. Meaning, from his point of view, the difference between circumstantial evidence and witness testimony is in the level of probability or likelihood. Okay, this is ninety-seven percent versus ninety-nine percent. But according to Tosafot there is no principled difference between circumstantial evidence and witnesses. So he makes this distinction on the probabilistic plane. That is one type of reasoning, evidentiary reasoning or probabilistic reasoning. There is another kind of reasoning, and that is legal reasoning. There are all sorts of principles in Jewish law and in legal systems generally that are not based on factual logic. I act in a certain way because there is a legal logic to act that way, even though on the factual level there is no justification for it. But on the legal level there is justification. For example: one who seeks to take from another bears the burden of proof. A rule that exists not only in Jewish law but also in the legal world. One who seeks to take from another bears the burden of proof. Does that mean that most likely the person holding the item is the true owner? In my opinion, no. Why assume the liar is the claimant and not the possessor? But there is legal logic in not taking the money away from the possessor unless I have a reason to perform an action. Taking it away is an action. The religious court performs an active step. If it has no reason to perform an active step, it won’t do it. So this is a very sensible rationale, but it is a legal rationale; it is not a rationale saying that the possessor is more likely to be right than the claimant. Therefore I call this a different kind of rationale: it is legal reasoning and not probabilistic reasoning. So when I say that behind a certain principle there is a rationale, that still does not mean that this rationale is probability. That rationale can also be legal reasoning. And later we’ll talk more about the fact that even within legal reasoning there is more than one type, at least two if not more. But that really will be next time. Okay.
[Speaker G] Excuse me, Rabbi, Tosafot also agrees that the distinction between fixed status and separation is not probabilistic reasoning at all. They have to accept that the distinction is not— In which Tosafot? No, what the Rabbi mentioned, that Tosafot speaks about circumstantial evidence.
[Rabbi Michael Abraham] Yes. So what does that have to do with fixed status and separation?
[Speaker G] No, no, they won’t always explain things because of probabilistic reasons when there is…
[Rabbi Michael Abraham] You’re asking whether Tosafot rejects in principle the possibility of legal reasoning that is not probabilistic? Not necessarily. Regarding circumstantial evidence he thinks there is a probabilistic difference.
[Speaker G] But elsewhere he
[Rabbi Michael Abraham] can accept legal reasoning.
[Speaker G] But to explain, in their view, the distinction between fixed status and separation, there’s no other option except, as you say, legal reasoning.
[Rabbi Michael Abraham] We’ll come back to that soon—not soon, in the next lecture—but we saw there too some probabilistic explanations.
[Speaker G] Didn’t we reject that? Didn’t we reject that from the probabilistic standpoint…
[Rabbi Michael Abraham] In the end I said there are certain explanations that you can view as probabilistic and explanations you can view as halakhic explanations. Even the last explanation I mentioned today—that when you enter the store that is your own decision—that is not a legal explanation, it is a probabilistic explanation. My claim is that on the probabilistic level it is not correct to discuss this as though it were ninety percent. That is not legal; it is a probabilistic explanation. Right. Okay.
[Speaker D] By the way, when statistics companies do all kinds of tests—like what the public chooses, for example, and things like that—they supposedly refer to, say, a thousand people. Right. They survey a thousand people, what each one says, and based on that they build statistics. Okay. But actually in our case, if we take doubt and probability, after all this is a human factor, and each person really can’t be pinned down in the same way as something else, so this statistic is kind of absurd because it can’t be…
[Rabbi Michael Abraham] No, I’m claiming that even in probabilistic decisions, if you do statistics on a sufficiently large number of people, it works.
[Speaker D] No, that’s what I’m saying. Every statistical survey—say Bezeq does a check of how many people are satisfied—they take a quantity of a thousand or fifteen hundred people, they call them, because you can’t take too many, they have a limited staff and time and money and so on. Right, so it’s a thousand people or fifteen hundred or two thousand, usually not more than that; that’s how statistics work. And from that they get some kind of agreement about most of the public, that’s how it is in this country.
[Rabbi Michael Abraham] A representative sample, okay.
[Speaker D] A representative sample, right, but it can’t be. Why? Because if you take each individual person, that means each one of them could really be something else entirely. Meaning, to take two thousand people—
[Rabbi Michael Abraham] And therefore they use a large number, and with large numbers statistics do work even on human decisions; that’s what I said before. That’s exactly, exactly the surprising thing here: with large numbers, statistics work even on acts of human decision. That’s precisely the point.
[Speaker D] But in fact, for example in the previous elections, everything the statisticians said and all that—it didn’t work at all.
[Rabbi Michael Abraham] I don’t remember, but usually statisticians are amazingly accurate.
[Speaker D] No, in elections and so on, where there is…
[Rabbi Michael Abraham] I don’t specifically remember the last elections at the moment; I very much doubt you’re right. Usually the accuracy in elections is astonishing. The only thing is that because the balance between the blocs sits around sixty, you can make a mistake and say this bloc won when in fact the other bloc won, because that’s within the margin of error. Meaning, if the true ratio is, say, fifty-nine to sixty-one and the margin of error is five percent each way—and five percent is an excellent estimate when you’re talking about five hundred people as a representative sample for the whole country—so think about it: a margin of error of four or five percent is an excellent estimate. But when the whole difference is two percent, then with five percent you can make a colossal mistake in the final result. But the statistical prediction is still an excellent prediction.
[Speaker B] The error for each party was only a few percentage points; Nadav Shnerb has an article about it in one of his posts.
[Speaker D] Yes, I remembered there was some kind of error.
[Rabbi Michael Abraham] These polls are excellent polls, it’s amazing. And it’s unbelievable, because most polls are based on five hundred people. That’s the sample for the whole country.
[Speaker D] Yes, I know that’s a small number; that’s why I didn’t understand how doubt and probability work here—there aren’t large numbers.
[Rabbi Michael Abraham] Amazing. Well, that’s the art of it; statisticians have accumulated experience, that’s the profession. Meaning, if a person who isn’t a professional did it, he’d need to take a million people. A professional knows how to choose five hundred. But the five hundred they choose are not random—
[Speaker B] The wisdom here is how to choose the
[Rabbi Michael Abraham] five hundred in a
[Speaker B] representative way; that’s based on prior information.
[Rabbi Michael Abraham] Exactly, that’s the art here. The statistician’s expertise is expressed not in asking and making a distribution of who votes for whom—that I can also do—but in choosing those five hundred so that they form a representative sample. Then you don’t need to get to a hundred thousand or a million; with five hundred you do it, which is amazing. Meaning, if I were doing such a thing, I assume my errors would be very large.
[Speaker D] And could it be that the sages of Israel relied on exactly this—that they knew whom to choose—and then majority, separation, fixed status and separation are based precisely on that?
[Rabbi Michael Abraham] They didn’t do statistics at all—what do you mean? When they determined that most women are not barren, or most women… all these things—they didn’t do statistics. They simply formed an impression from what was around them. There was no awareness there of really doing precise scientific statistical work. They formed an impression and established that there is such-and-such a majority. They also didn’t determine what the majority was, whether it was ninety percent or seventy percent.
[Speaker D] No, but a majority is one more.
[Rabbi Michael Abraham] Meaning, not… Exactly because they didn’t really do scientific work and check. Rather, an impression says that most women are not barren, and I assume that the people I know are a representative sample. And by the way, that’s fairly true. Yes.
[Speaker B] Rabbi, the Rabbi said at the end of the lecture that we don’t make expositions without logic. Usually the expositions are also based on logic, on the reasoning of the sages who expounded them. But surely the Rabbi would admit—and I think can admit—that if they had a strong rationale, they wouldn’t have a problem finding a way to derive an exposition. So in the end, Jewish law was really determined by reasoning; it is a creation of the sages of Israel, and not because one verse or another said this or that.
[Rabbi Michael Abraham] Rather, our sages knew how to overcome the verse.
[Speaker B] I don’t agree.
[Rabbi Michael Abraham] I don’t agree. Again, I don’t agree a priori—not from the Talmudic material itself, but I don’t agree a priori because it sounds implausible to me. If that were the case, then let them be straightforward and say they are relying on reasoning. Why are they playing games with me of verbal analogies and inclusions and crowns and letters?
[Speaker B] But let’s think about that question, let’s think—it’s a real and serious question, but it’s not necessarily correct. It’s not necessarily that they were consistent in that. When Rabbi Akiva came and said, “that your brother may live with you”—your life takes precedence over your fellow’s life—and with that he eliminated the… or saved, it doesn’t matter what he did there.
[Rabbi Michael Abraham] No, there it’s something else. “That your brother may live with you”—no, but “that your brother may live with you” can also be an idiomatic phrase; it’s not necessarily a derivation from a verse. That’s why I say—because why? Because I know… No, I’m explaining: because there we are not talking about a law where in the end you would also flog the person who violates it. We are talking about rules of precedence; it’s not a question of lashes. But the question is… if as a result of an exposition you derive a law and someone violates it and you flog him, then there has to be a source for it in the Torah, because on the basis of reasoning alone we do not administer lashes.
[Speaker B] I understand. That story there, with two people who were walking, has significance for human life. The significance of that derivation is that maybe someone will die, maybe someone will live.
[Rabbi Michael Abraham] Of course. And that comes from reasoning. We apply reasoning to human life.
[Speaker B] So why does he cite a verse? I don’t understand. When Rabbi Akiva comes and says it, after all he doesn’t say to Ben Petora: you’re wrong, your reasoning is illogical. Even though obviously he disagrees with him.
[Rabbi Michael Abraham] I understand, that’s what he told him.
[Speaker B] I agree too. So why does he bring a verse?
[Rabbi Michael Abraham] Because that verse is a linguistic motto. “That your brother may live with you” means: that is the way he expressed his reasoning.
[Speaker B] And that motto has significance for human life.
[Rabbi Michael Abraham] Why can’t we extend that motto to all the expositions of the sages? No, human life is decided by the reasoning, not by the motto. That’s exactly what I’m telling you.
[Speaker B] So why not extend that motto to all the expositions of the sages?
[Rabbi Michael Abraham] No, that’s exactly the point. Because when you say, for example, “You shall fear the Lord your God,” that’s not a good example because it’s a positive commandment. But take an exposition—I don’t know—eating poultry with milk, meat with milk, sorry. Okay? So it says, “You shall not boil a kid in its mother’s milk”; that’s about cooking. How do we know it is forbidden to eat it? It’s written twice. So that is really an exposition. From the fact that it is written twice we learn also about eating. Now someone who eats meat and milk receives lashes. Right? How can he receive lashes if it’s just reasoning? If it’s only reasoning and has no source in a verse, you can’t flog him; punishments are administered only where there is a prohibition. Therefore it is clear that there they really think the verse says it.
[Speaker B] And Maimonides rules that one who eats is also liable. And once it is a law, then you can warn and administer lashes.
[Rabbi Michael Abraham] Right, and on the basis of reasoning alone we do not administer lashes.
[Speaker B] The problem is that it is so hard to find any kind of motto.
[Rabbi Michael Abraham] Not a motto—you need a source, that’s what I’m saying. A motto won’t help here. You need a source. Because the source says this is a prohibition. Once the source says this is a prohibition, if someone violates it I can punish him. Mottos do not do that job.
[Speaker D] But the verse doesn’t say the prohibition is eating. The verse says the prohibition is cooking. But there is an exposition.
[Rabbi Michael Abraham] But the exposition explains that this verse forbids eating. But I claim that the exposition really explains the verse; it does not use the verse as a fig leaf while really everything comes from reasoning. Because if that were so, they would not administer lashes for it.
[Speaker D] That’s clear, yes. But it’s still an exposition. A rabbinic exposition. Yes, obviously.
[Rabbi Michael Abraham] No, not rabbinic. An exposition. Rabbinic means an enactment. A Torah-level exposition. The sages expounded a Torah law, yes.
[Speaker D] The sages expounded, yes. The sages expounded, yes.
[Speaker B] “An eye for an eye, a tooth for a tooth, a leg for a leg.” In plain Hebrew it’s clear that the Torah is emphasizing something specific, and the sages said that’s not logical and found a way to expound it differently.
[Rabbi Michael Abraham] In complete
[Speaker B] contradiction.
[Speaker D] No, by the way, that’s not true. There are many who disagree with that.
[Rabbi Michael Abraham] Fine, I don’t agree, but okay.
[Speaker D] No, in principle. I heard Rabbi Asher Weiss being asked a question—what would be the interpretation in the case of a person who is extremely rich, so that paying money for the eye he actually took out from someone wouldn’t diminish him at all? He said it could be that at the Torah level they would take out his eye. Because this punishment of money, yes, it has to function as a punishment. He has an entire lecture on this.
[Rabbi Michael Abraham] That sounds very strange to me.
[Speaker D] You can find it. It’s online. Okay. Yes, I heard it, but it can be found, no problem. Okay. Meaning, there are people who say that in its plain sense it should be that way, and the sages wanted to prevent some kind of disaster in the world. Could be.
[Rabbi Michael Abraham] It doesn’t seem that way to me, but I don’t know, we’d have to see what his proofs are. Okay. All right, friends.
[Speaker D] Good Sabbath
[Rabbi Michael Abraham] Peace to you.
[Speaker B] Sabbath peace, thank you very much.
[Rabbi Michael Abraham] Sabbath peace.