Doubt and Statistics – Lecture 11
This transcript was produced automatically using artificial intelligence. There may be inaccuracies in the transcribed content and in speaker identification.
🔗 Link to the original lecture
🔗 Link to the transcript on Sofer.AI
Table of Contents
- Democratic majority, social power, and the legitimacy of disproportionate influence
- Statistical evidence in law and the gap between it and eyewitness testimony
- Circumstantial evidence in Jewish law: Rabbi Shimon ben Shetach and Maimonides versus Tosafot
- Monetary law, estimation of intent, trust in judges, and a fraudulent case
- Legal reasoning versus probability, and preparation for a future explanation
Summary
General Overview
The text continues the distinctions between types of doubt, majority, and statistics, and sets up a contrast between a majority whose purpose is to get as close as possible to the truth and a democratic majority whose purpose is to represent the will of the public. It argues that when democracy is viewed as representing public will, there is room to understand politics as a broader clash of forces than the ballot box alone, including economic, military, media, and cultural power, and even to justify disproportionate influence by those who contribute disproportionately. From there it moves to the question of using statistical evidence in law, presents the intuition and practice that one does not convict on the basis of statistics alone, and parallels this in Jewish law through the prohibition against conviction in capital cases on the basis of circumstantial evidence even when it is extremely strong.
Democratic majority, social power, and the legitimacy of disproportionate influence
The text states that a majority in a religious court, according to Sefer HaChinukh, is meant to get as close as possible to the truth and to the correct ruling, whereas a democratic majority is meant to represent what the public wants. It argues that when the public is divided, adopting majority rule is a simple tool for representing the public will, but there is room for further weighting of desires according to the contribution and price borne by different groups. It presents the view that someone who contributes disproportionately to the functioning of the state, for example through economic power or military service, is entitled to disproportionate influence, and compares this to the U.S. veto right in the Security Council as opposed to states that do not bear similar costs.
The text recounts a panel in the tent of hunger strikers in Jerusalem with Inbal Hermoni, and describes an argument in which he claimed that the influence of people with economic power on policy—even if it delays the will of the elected political majority—is not necessarily unfair but may be justified. He describes a shocked reaction from the participants to the claim that someone who earns little “is worth less” than a wealthy tycoon, and ties the dispute to the clash between capitalism and socialism, while arguing that in the ballot box everyone has equal power but in the market economic power is legitimate to use. He points to an internal contradiction in the position of those opposing the reform, in that the very same economic forces and serving elites who, in his view, stopped the reform are precisely the forces whose actions the opponents benefit from in practice even while denying their legitimacy.
The text ties this to the distinction between two roles of majority and argues that in the conception of majority as a measure of truth there is no justification for giving extra weight to those with power, because there is no reason to assume they are wiser than the average citizen. It argues that in the conception of democratic majority as a measure of public will, democracy is a complex game beyond the ballot box in which power struggles, demonstrations, and economic threats are part of the mechanism that generates a “collective will,” and therefore the influence of different forces can be legitimate within certain limits.
Statistical evidence in law and the gap between it and eyewitness testimony
The text turns to the use of statistics in legal questions and refers to columns 226 and 228 on statistical evidence, arguing that the philosophical and legal literature has not given a convincing explanation of the problem. It quotes an article on the Alaxon website by David Papineau about a case of one hundred prisoners, in which ninety-nine attacked a guard and the hundredth did not, and a random individual stands trial with no additional evidence, so the chance that he is guilty is 99%. It presents the accepted intuition and the claim that legal systems do not convict on the basis of statistical evidence alone, and also brings up the British woman who was convicted on the basis of a statistical claim about crib death, noting that in the end it turned out that the statistical reasoning was flawed.
The text presents the puzzle that courts are willing to convict on the basis of eyewitness testimony even if it is viewed as having something like 95% reliability within the standard of “beyond a reasonable doubt,” but are unwilling to convict on the basis of a 99% statistical datum. It rejects the standard conceptual distinction in the literature between “information” and “belief” as something that does not explain the gap, and argues that labels do not change the central question of the probability of error. It announces that he will offer three explanations of his own, and later present and critique the explanations in the literature.
Circumstantial evidence in Jewish law: Rabbi Shimon ben Shetach and Maimonides versus Tosafot
The text presents a “halakhic perspective” through the Talmudic text in Shevuot 34 about Rabbi Shimon ben Shetach, who ran after a man into a ruin and found him with a sword in his hand, blood dripping from it, and a murdered man still twitching, and said to him, “Wicked man, who killed this one? Either I or you,” but concluded, “What can I do, for your blood is not delivered into my hands,” because “by the testimony of two witnesses or three witnesses shall the dead man be put to death.” It defines this as foundational for Jewish law’s attitude toward circumstantial evidence and emphasizes that circumstantial evidence is not necessarily weak, and may even look almost absolutely certain.
The text argues that the plain meaning of the Talmudic passage draws a categorical distinction in admissibility between direct testimony of witnesses and circumstantial evidence, not a distinction based on the quality of the evidence. It notes that Tosafot is troubled by this and argues that the circumstantial evidence is not absolute, and attributes the lack of conviction to the quality of the evidence, so that if circumstantial evidence reached a sufficient level of certainty it would be possible to convict. It presents Maimonides as learning differently and drawing a sharp distinction between monetary law and capital law: in capital law, a religious court does not punish based on estimation and inference, only on witnesses with clear evidence, even in a case of someone chasing another into a ruin with a blood-dripping sword, and he brings the verse “Do not kill the innocent and the righteous” as meaning one must refrain from killing when there is some formal basis to clear and justify him.
Monetary law, estimation of intent, trust in judges, and a fraudulent case
The text brings Maimonides in the Laws of the Sanhedrin, chapter 24, according to whom in monetary law a judge should rule according to what his mind inclines him to see as true even without clear evidence, and certainly if he knows it with certainty. He illustrates this with the case of Rava, whose wife whispered to him that the woman before him was “suspect with regard to an oath,” and he transferred the oath to the opposing party, and he presents the reaction of modern jurists who were outraged at the idea that a judge’s wife could influence a ruling in this way. He argues that from Rava’s perspective, if he knows that the one taking the oath will swear falsely, allowing the oath would leave stolen property in the hands of the person not entitled to it, and he suggests that the gap from the modern view depends on the degree of trust the community places in the judge and on fear of abuse.
The text quotes Maimonides, who explains that although this is “the essence of the law,” in practice, “once courts that were not decent became numerous,” most Jewish courts agreed not to judge on the basis of the judge’s own impression and not to reverse an oath except on the basis of clear evidence, so that no ordinary person could say, “My heart believes this one.” He interprets this mainly as a change in the community’s attitude toward the judge and an absence of willingness to trust him, so that justice must “be seen” and not only “be done.” He adds that even today there exists the concept of a “fraudulent case,” in which a judge who suspects deceit, even though he has no formal basis to disqualify witnesses, increases the inquiry and investigation, and if his heart is not at rest he withdraws from the case, linking this to the principle that a ruling one is not fully at peace with may lead to distortion.
The text raises an additional example in the context of the “scriptural decree” disqualifying relatives, and challenges the conception that relatives speak the truth but their testimony is simply inadmissible, because then a situation could arise in which truthful refuting witnesses are not accepted and an innocent person would be executed. It suggests that the law of a fraudulent case and withdrawal from judgment set a limit: a judge does not issue a ruling he knows is not true, even if the formal rules do not allow direct acquittal.
Legal reasoning versus probability, and preparation for a future explanation
The text presents a concept of “legal reasoning” from a pamphlet on migo, and argues that in legal rules and the law of evidence there are considerations that do not arise from probability but from what is legally correct, such as “the burden of proof rests on the one who wishes to extract from another,” which is not necessarily based on any statistic that the current holder is more likely to be right. It raises the possibility that the distinction between statistical evidence and direct testimony likewise rests on a substantive legal intuition and not on probabilistic calculation, but says that he wants to give that real content. He states that he intends to argue later that there may in fact be a probabilistic reasoning here that is not accepted in the literature he has read, and concludes with the claim that a legal system does not “bring justice” but “brings law,” and therefore it operates through legal rules.
Full Transcript
[Rabbi Michael Abraham] Okay, we’re continuing with matters of doubt and statistics, and up to now I’ve tried to define several kinds of doubts, or contexts in which statistical tools can be used. I spoke about vagueness versus doubt; then I moved to the two types of majority. Really, majority is the expression of statistical tools in Jewish law, following the majority, and we saw some failures in the use of statistics and majority, and I distinguished between a majority that is before us and a majority that is not before us. And after that I moved to democratic majority, and the claim was that democratic majority has a different purpose than majority in a religious court. The majority in a religious court, at least as Sefer HaChinukh explained it, its purpose is to get as close as possible to the truth, to the correct ruling. By contrast, democratic majority is meant to represent what the public wants. When there are disagreements, then you have to adopt some criterion that tells us what the public wants, because in the end the policy or decision the public is supposed to adopt should reflect what the public wants, and when there are disagreements we need to define how we represent what the public wants. We saw differences between those conceptions; the whole rule of the philosophers collapses once they understand it that way, the supposed advantage of the wise, and so on.
And now I’m suddenly reminded of an interesting point, maybe, just something that suddenly popped into my head. The day before yesterday, I think it was Monday evening, I was on some panel or something in the tent of the hunger strikers in Jerusalem. They invited me to speak there, and the chairwoman of the social workers’ organization was also there, Inbal Hermoni I think, something like that. And anyway, during the discussion, she—this Inbal—had a very left-wing worldview, I’d say radical, and she talked about all the demonstrations where they’re always just a few dozen people, and now finally they feel that they have more respectable numbers around them. In any case, during that discussion we got to a claim that I also made in some of the columns—I’m just remembering all this now in the context of what I was talking about—that I said there that the fact that people use their economic power, for example what we see today, in order to influence political decisions, and that’s also why you get these protests of “we are second-class citizens,” where basically the economic, academic, media elites, whatever, manage to move the majority away from its policy or freeze it or delay it or change it, soften it, whatever, with respect to its policy. And my claim was that there is justice in that. In other words, it’s not true that it’s unjustified. The fact that we are second-class citizens may be true, but you are justifiably second-class citizens.
Because the fact is—I said there, and I think maybe we talked about this—that the U.S. has veto power in the Security Council, Israel doesn’t, Belgium doesn’t either, and there’s justice in that. Why? Because once there’s some problem somewhere in the world, the one who sends an army there and the one who invests the money to deal with that problem is not Israel and not Belgium—it’s the U.S. So someone who contributes disproportionately deserves disproportionate influence. And that was basically the claim, a somewhat undemocratic claim, which is why I’m remembering it now.
So I said to them there that in my opinion the fact that people with economic power use their economic power to influence policy—maybe against the opinion of the majority, it doesn’t matter right now, it’s not exactly against the opinion of the majority, but let’s assume it is, against the elected political majority, let’s call it that. I don’t know whether that’s the majority in the public today, I don’t think it is there, but the elected political majority in principle has the right to set policy and implement it, and still there are very strong economic and social forces that are, meanwhile, managing to stop it, moderate it, halt it, and so on. The question is whether that’s justified. So the claim is that in the end, the person who really sustains many of us to a large extent, or a large part of the possibilities of what the state does, is actually based on the income of people whose marginal contribution is incomparably greater than that of the ordinary citizen, including myself of course. I think it is perfectly just that they use their economic power; there’s nothing unfair about it. In other words, they also provide their economic input and contribute out of all proportion to the state’s product and its ability to function, so in my view it’s perfectly just that they use that same economic power to influence things.
So she was really shocked—how can that be? What, I’m a social worker making, I don’t know, eight thousand—well not me, never mind—someone making eight thousand shekels is worth less than Tshuva? I told her yes. So she was—everyone was shocked. You could hear gasps and hissing from the audience, not hissing exactly, more like people could barely breathe. How can you say something so politically incorrect? And I said yes, that was part of the argument between capitalism and socialism. She’s socialist to the point of communism, and I was also representing capitalism there in some aspect of the argument, though it wasn’t really about economic systems. So I told her yes, absolutely yes. She was really shocked. I said to her, why are you so shocked? In the ballot box everyone has the same power, but in the market—well, that’s part of capitalism. Part of capitalism is that your ability to use economic power to influence things, if you have such power, is legitimate to use. Maybe within certain limits, but in general it’s legitimate to use it. That was my claim.
Now afterward I drew her attention to the unfortunate fact, from her perspective, that she hadn’t quite noticed, namely that there’s an internal contradiction in what she was saying. She’s from the camp most strongly opposed to the reform, right? I mean she’s extremely radical against the reform. And I said to her: listen, the people playing your game, the people who managed to stop the reform on your behalf, are those same wealthy elites. It wasn’t the socialist masses, the proletariat, who stopped Bibi. Bibi doesn’t care a bit about two hundred thousand demonstrators. His grandmother cares. What interests him is economic power and refusal to serve in the army. Meaning, the serving elites, let’s call them that, in the army, who certainly in large part are not in the coalition camp, at least a large part of them, and the economic elites, who probably in large part also are not in the coalition camp, and they use their power to push their agenda. So you’re telling me they have no right? You stop them, then, make them not do it. When they do it she’s silent in a thousand languages, but when I say that it’s legitimate then of course she protests.
Okay, that’s just parenthetically. What matters for our purposes is that now I’m suddenly remembering: let’s try to think about the justification for using these forms of power—economic, whatever, cultural, intellectual, legal, military service, military standing, and so on—using these things in terms of majority. After all, they seem to be acting against the majority position. Is that legitimate or not? If the goal really is to arrive at the true decision, then she’s right. Because the person with economic power, or the person with abilities, or I don’t know, the one serving in the army, is not necessarily more intelligent or more likely to reach more correct conclusions than the average voter like me or you. Therefore, if the goal really is to reach the truth, and the assumption is that majority is a measure of truth—an assumption I think is absurd—then she really is right that there’s no justification for it.
But if the purpose of majority is to express what the public wants, then that’s a different question. Because as I said, the public is divided in its views, and the criterion of majority is ultimately just the simplest criterion we adopted in order to express what the public wants. But here there is definitely room for weighting. In other words, when I ask myself what the public wants, there is room for the claim that someone who contributes much more, say, to the functioning of the public, what he wants should be factored in differently when we sum it all up in order to decide. Because in the end I think there is logic in saying that his will is more significant, because he will pay a much heavier price if his will is not fulfilled than I will. I pay—yes, taxes like every ordinary salaried employee, not really, whatever, such and such amounts like everyone pays. And this person pays billions every year. And not only that, he supports me—I work for him, I, whatever, many people like me work for him. In other words, his contribution to the functioning of the economy is immeasurably greater than that of any of us, leave aside contributions on other levels. But first of all, in the most basic and most agreed-upon sense of economic contribution, there is no question that his contribution is much greater. And therefore if he uses that economic power, at least legitimately, to move decisions in the direction he believes in, in my view that is legitimate. Even though ostensibly it means going against the decision of the voter at the ballot box, because the voter at the ballot box determined that this government is the one that would lead. But I think the democratic game is more complicated than the ballot box. There are also, during those four years, power struggles, demonstrations, economic threats from all sides; that is part of the democratic game. That’s how the public speaks.
And therefore in my view—this just came back to me while I was talking about democratic majority—here, for example, is another practical implication. Am I willing to accept that there is greater strength and weight for forces that have economic power, security power, or whatever it may be, than for the ordinary citizen? And my answer is that according to, in light of the conception of democracy I described earlier, actually that is not absurd. Yes, there is room for it at least. Again, you can argue about it, but certainly there is room for it. In the conception of striving for truth, no—I don’t think Tshuva is smarter than me or you, or this pilot or that one, there’s no reason to assume he is wiser than we are. So if we are striving for truth, there is no justification for that. But it’s a mistake to think that the point of democratic majority is to hit the truth. The point of democratic majority is to represent what the public wants. And the public is some kind of clash of forces that ultimately produces some collective will. And in that clash, it’s a clash between fingers at the ballot box and economic, media, cultural, and military power—yes, that whole game, again within certain bounds of legitimacy—that is ultimately what gives rise to what is called the will of the public. And therefore in my view it is completely legitimate.
So there, I just remembered that this could be another implication of the distinction I made between the two roles of majority: majority as a measure of truth, and majority as a measure of what the public wants. Okay, so that’s just a completion of last time. I now want to get back to our basic line of thought. This democratic timeout happened because it was, yes, it was a required comment on Sefer HaChinukh, on majority as a measure of truth, so I contrasted that with democratic majority. And of course it was also timely, I thought it would be interesting to connect it to issues that are on the table. But now I want to return to our regular course. And here I want to move to the use of statistics in legal questions. And this, as we’ll see, is tied at the core to the foundations we saw in previous meetings and also to what we’ll see later on.
I want to start—there are columns on the site about this too—I want to start with column 226. There I dealt with statistical evidence, 226 and 228, I dealt with the issue of statistical evidence in law. This is a very broad issue; an ocean of material has been written about it, philosophers and legal scholars have dealt with it a great deal. In my opinion no one hit the explanation—again, no one, from what I saw—no one really took the bull by the horns and offered an explanation, I don’t know, that in my view is convincing or really holds water. I want to offer you my own explanation based on the distinctions we made, and in my opinion it’s the only explanation I know of that holds water. You judge.
Let’s begin with a quotation from an article on the Alaxon website, an article by someone named David Papineau, at the end of the paragraph I think. Look here. Imagine one hundred prisoners exercising in the prison yard. Suddenly ninety-nine of them attack the guard in a premeditated assault in which the hundredth prisoner did not participate. Right, ninety-nine out of the hundred attacked the guard. Now one of the prisoners is sitting in the defendant’s seat. There is no additional evidence, no concrete evidence at all. The probability that he is guilty is ninety-nine percent. Ninety-nine attacked and one did not attack; the question is who that one is. Now some random person is sitting in the defendant’s seat. The probability that he is guilty is ninety-nine percent and the probability that he is innocent is one percent. Should the court convict? Ninety-nine percent that he is guilty. Can he be convicted on that basis? On that consideration alone—there is no evidence against him. But I know he was in the yard, that I know, and I know that out of the hundred who were in the yard ninety-nine attacked the guard.
So he says, everyone’s initial reaction is: of course not. The court has no evidence that rules out the possibility that the defendant is the innocent prisoner. You cannot convict a person solely on the basis of statistical evidence. You can’t use statistics to convict someone. By the way, it’s not only an initial reaction; that’s how it is in every legal system in the world, at least according to those who wrote the articles, who are bigger legal scholars than I am. So the claim is that in all legal systems in the world, they do not convict in such a case.
By the way, this reminds you of the story of that woman with the two—remember? The one with the two children who died of crib death in Britain, and they convicted her on the basis of a statistical consideration, that the probability that two children would die of crib death is one in sixty million, and therefore clearly she murdered them, there’s no chance this was crib death, and on that basis they put her in prison. In the end it turned out that the statistical reasoning itself was flawed. But if it—if they really hadn’t found a flaw in the statistical reasoning, then there it seems they did convict on the basis of a statistical consideration alone. That’s very interesting, because from what I read in this literature that I looked through a bit, you don’t convict on the basis of statistical evidence alone. So that’s very interesting. Anyway.
In any case, he says as follows: the prison-yard uproar is a fictional example, but the same issue arises often in real courts, and then the law matches the intuition of the ordinary person. Statistical evidence alone is not sufficient. In both civil and criminal trials, defendants may be found guilty only on the basis of evidence that relates directly to them and does not replace them with some general category on the basis of which they are likely to be guilty. Okay, that is basically what is called statistical evidence. This may be an intuition, but the fact that the law forbids the use of statistical evidence is puzzling.
Think of a person who is convicted because an eyewitness saw him steal a necklace. Nowadays, thank God, courts know that eyewitness testimony can be mistaken, and therefore they examine carefully and make sure the testimony is reliable. Nevertheless, courts do not demand one hundred percent certainty and are satisfied with the point where the testimony can no longer reasonably be doubted. And from cases in which judges were persuaded to speak in numbers, one may infer that this means ninety-five percent certainty. Right, many times when you ask legal scholars “how much” in a criminal trial, then when they talk about conviction beyond a reasonable doubt, those who give numbers say ninety-five percent. Of course that has no real basis, but they want to give some sense of the level of certainty required in order to convict someone in a criminal trial, so they say ninety-five percent. I’ve also heard this number more than once, although I don’t really know how—it can’t actually be quantified—but that’s supposed to be the intensity of the feeling when you convict in a criminal case.
That is, the concluding sentence: in other words, we are often willing to convict on the basis of eyewitness testimony, but never on the basis of statistical evidence. You may wonder why we think that eyewitnesses whose reliability is estimated at ninety-five percent are less likely to mislead us than statistical data whose reliability is ninety-nine percent. Okay? There really is a very difficult question here. When you take an eyewitness, you tell me his reliability is ninety-five percent. There are five percent cases where an eyewitness is mistaken. Lawyers—I hear from Dafna, from my wife, all the time about witnesses. You can never know; with witnesses, you will never find two witnesses whose testimony is identical. We’re talking about witnesses, eyewitnesses. And apparently for lawyers this is daily bread: even eyewitness testimony, which to an ordinary layman might seem like—what, it’s obvious, if there are eyewitnesses then it’s closed—no. Eyewitnesses apparently make mistakes not infrequently. And let’s assume for the sake of discussion that it’s ninety-five percent, because eyewitnesses are enough to convict in a criminal trial. Let’s say ninety-five percent is the reliability of an eyewitness, and that’s enough for us. So why is the ninety-nine percent consideration, like that prisoner there in the prison, not enough for us to convict? The chance of error there is one percent, and that we do not use to convict. A chance of error of five percent, that we do convict on. Why? What’s the logic?
Now this is a very difficult question. And what’s even harder about it is that everyone’s intuition—everyone I at least have read and spoken to and so on—everyone agrees that this is right. That it is not right to convict the prisoner in such a situation, but it is right to convict on the basis of eyewitness testimony, even though you can never be sure. But you know, if we go only for certainty, then we’ll never know, we’ll never be able to do anything ever. So the intuition here seems very universal and sweeping; almost everyone says that this is how it seems to them intuitively. On the other hand, when you put the data side by side, it’s really embarrassing. How can it be that ninety-five percent is enough and ninety-nine percent is not enough? Something here is not logical.
So I want to—I’ll do this in two stages. First I’ll offer three explanations of my own, and you’ll see that they may actually even be one explanation, or very close to one another. But I think these are three angles that illuminate the point, which to me is the real point here. And next time, I assume only then will I get to it, I’ll come to the explanations offered in the literature. And I’ll try to show why in my opinion they’re not correct, they don’t hold water, they aren’t convincing. It’s word games.
I’ll just say in two sentences: Papineau himself in his article—this is also accepted in the philosophical literature, though everyone defines the concepts differently—distinguishes between knowledge or information and belief. Belief—I’m not talking right now about belief in the religious sense, but belief in something as opposed to knowledge. That is, when there are eyewitnesses that you stole a necklace, or murdered, or attacked the guard, or whatever it may be, although eyewitnesses can be wrong, that is information, just uncertain information. But the statistical consideration I’m talking about with the prisoners—I have no information about this prisoner. I’m simply making a statistical judgment; that is belief. I believe that he probably murdered, but I don’t have any information about this prisoner. I’m phrasing it very generally because these concepts of belief versus information receive different interpretations in the literature. And next time I’ll try to present at least some of them. But that is the accepted view in the philosophical world for explaining the difference between the two situations: this is belief and that is information.
Now without even getting into the question of what belief is and what information is, in my opinion this distinction is very strange. Because even if you call this belief and that information, what difference does it make to me what you call it? Call it whatever you want. In the end, here you have a ninety-nine percent chance that you are right, and here you have a ninety-five percent chance that you are right. So what difference does it make to me whether you call it belief or information? Why do you convict on the basis of ninety-five percent and not convict on the basis of ninety-nine percent? What difference does the label make? Do names change anything here? Why are these labels interesting? You just want to make sure that the person you convicted was convicted correctly, that you didn’t make a mistake. That’s all. So in the case of the prisoners, the chance that you made a mistake is lower than in the case of eyewitnesses. So why not convict there, or why do convict in the eyewitness case? What difference do all these word games make?
So in general, all these distinctions between information and belief seem fishy to me, even before we get into the explanation of what information is and what belief is. I don’t understand what that label helps with. The labels, the labels. After I offer the explanations I have in mind, we’ll come back to this again, and maybe that’s what lies behind this conception of information versus belief, but that’s not what you’ll find in the literature. So that’s why I’m doing it in this order.
Okay, so first of all I want to offer—well maybe first of all I want to look at this from a halakhic perspective. So look at the Talmudic text in tractate Shevuot 34: As it was taught: Rabbi Shimon ben Shetach said, “May I not see consolation if I did not see one man running after another into a ruin, and I ran after him and found a sword in his hand, with blood dripping from it, and a murdered man twitching. I said to him: Wicked man, who killed this one? Either I or you. Only the two of us are here. The ruin is empty of people, so either you or I, those are the two possibilities.” Right? And of course what is Rabbi Shimon ben Shetach saying? Yes, but I know that I didn’t kill him, that I know. But what… And Rabbi Shimon ben Shetach was a judge, or head of the Sanhedrin. So what can I do, for your blood is not delivered into my hands? For the Torah said: “By the testimony of two witnesses or three witnesses shall the dead man be put to death.” Rather, the Omnipresent will exact punishment from you. They said: he did not move from there until a snake bit him and he died.
In other words, Rabbi Shimon ben Shetach basically knows with certainty that this person murdered. Right, the corpse is bleeding before our eyes, the man is holding a blood-soaked knife, there’s no one else anywhere nearby besides Rabbi Shimon ben Shetach himself. And he is, of course, the head of the Sanhedrin. “Seeing should be no less than hearing,” as the Talmudic text says in Rosh HaShanah regarding sanctifying the new month. If two witnesses who come before the judge present reality to him, that is enough to decide the case, then if the judge himself saw the reality, that’s only stronger. So if so, seemingly Rabbi Shimon ben Shetach should have judged him as a murderer, sentenced him to death. And why not? The Torah says what to do. The Torah says: “By the testimony of two witnesses or three witnesses shall the dead man be put to death.” Rather, the Omnipresent will exact… But I know that you are a murderer, it’s not that I have a doubt that you are a murderer. The Omnipresent will exact punishment from you—but I as a judge cannot… cannot do it.
Basically, this is the foundational Talmudic passage from which they derive Jewish law’s attitude toward circumstantial evidence. Toward what is called circumstantial evidence. Now, contrary to a common public image, circumstantial evidence is not a synonym for weak evidence. Circumstantial evidence can be very strong. The circumstantial evidence described here in the Talmudic passage is million-dollar evidence. It’s one hundred percent. I don’t know, a hundred, ninety-nine point nine percent. There’s nothing more. You see a person enter the ruin alive, you now see that person dead, you see another person holding a knife in his hand, there is no one else around. What else could have happened? It’s mathematical proof that he murdered. One hundred percent. But no one saw him. There are no witnesses who saw the act. Therefore this thing is circumstantial evidence. It is evidence… Circumstantial evidence in the literal sense means evidence by virtue of the circumstances. It does not mean weak evidence—on the contrary. Some would say this evidence is much stronger than the testimony of two witnesses, as we saw before. Two witnesses can be off by a few percent. Maybe they were mistaken in what they saw. Here I have mathematical proof. It’s one hundred percent.
So there is here wonderful circumstantial evidence, and it is circumstantial. And the Talmudic passage says that in the Torah it is written: “By the testimony of two witnesses or three witnesses shall the dead man be put to death,” and not apparently… only on the basis of witnesses, and not circumstantial evidence. In other words, the distinction is, at least as it appears in the plain sense of the passage, not in the quality of the evidence but in the category to which it belongs. Evidence… evidence by way of witness testimony is admissible evidence in capital cases. Circumstantial evidence is not admissible, unrelated to the question of how strong it is. Now this is very parallel, as I said before, to the examples we saw earlier. But in law…
[Speaker B] In today’s legal system, that same person with the knife would be cross-examined, and that would affect the ruling also based on how plausible his version is. Meaning, if he can’t explain it, then that itself becomes further evidence beyond the person who saw him.
[Rabbi Michael Abraham] Okay, but still, still in… First of all, in Jewish law it’s not like that.
[Speaker B] Right, that’s what I’m saying, in Jewish law it’s not like that. And second, still…
[Rabbi Michael Abraham] You can see in this some kind of pattern. There’s statistical evidence here, and that’s exactly why the defendant’s version—or some flaw in his version—is required. Otherwise, why do you need the defendant’s version at all? You have one hundred percent evidence that he murdered. Why should you care about the different versions? Unless you need something more. So you understand that there’s still a problem with convicting based only on circumstantial logic. You apparently need something else to support the conviction in law as well, not just in Jewish law. But in Jewish law that is completely the case. Okay, now, so that’s actually very similar to the distinction we made earlier. Because when you take that person, that prisoner, after all, you don’t have anyone who saw him; it’s circumstances. He was in the courtyard, ninety-nine out of the hundred who were in the courtyard committed murder, so there’s a ninety-nine percent chance that he murdered. Here it’s ninety-nine, not a hundred, but still, in terms of statistical significance, that’s good enough; ninety-nine percent is good enough to convict. Okay? And still we do not convict. But if there are two witnesses who saw him, that is not circumstantial evidence. “By the testimony of two witnesses a matter shall be established,” and then he is executed, or judged as a murderer, whatever. It’s very parallel to the distinction Papino describes there. Now, Maimonides—the Tosafot really struggles with this, and Tosafot claims that it simply isn’t true. There is a problem with circumstantial evidence. He claims that maybe there was someone else there, whatever—you can’t be sure. So Tosafot does tie it to the quality of the evidence. Meaning, if there were circumstantial evidence at the same level of certainty as eyewitness testimony, then we would convict. That’s what Tosafot claims. So he does tie it to the quality of the evidence. But Maimonides learned otherwise. Let’s look at him inside. I’m starting with Maimonides in the Laws of the Sanhedrin at the beginning of chapter 20. But as background I want to show the beginning of chapter 24. In chapter 20 he speaks about criminal law, capital cases or lashes and the like. In chapter 24 he speaks about civil law, monetary cases. So in chapter 24 he says: “A judge may rule in monetary cases according to matters toward which his judgment inclines as true, and the matter is firmly fixed in his heart that it is so, even though there is no clear proof there. Needless to say, if he knew with certainty that the matter was so, that he judges according to what he knows.” What has Maimonides said up to this point? If the judge has some certainty that this is the truth, then in monetary cases that is enough. You can render judgment on the basis of circumstantial evidence, on the basis of whatever—as long as you know for certain that it is true. You don’t need witnesses. If the matter is firmly fixed in his heart, if it’s clear to him that this is the truth, “even though there is no clear proof there.” Even a strong intuition. “Needless to say, if he knew with certainty that the matter was so, he judges according to what he knows.” What does that mean? That’s our case, of Shimon ben Shetah, because there the judge knew with certainty that this was the situation; he just didn’t have witnesses. But the circumstantial evidence gave him certain knowledge. So Maimonides says that in monetary cases that is enough. You judge the person in monetary law according to circumstantial evidence or some sort of intuitive feeling. How so? Suppose a person became obligated to take an oath in religious court, and someone tells the judge—someone trusted by him, and whose word the judge relies upon—that this man is suspect with regard to oaths. The judge may transfer the oath to the opposing party, and he will swear and collect, since the judge relies on this person’s statement. Even if it were a woman or a slave who was trusted by him—as if one had told him that the person is not trustworthy regarding an oath—even if it was a woman or a slave, who are not trusted for testimony, not acceptable as witnesses. Since the matter seems strong and true in his heart, he relies on it and judges accordingly. Needless to say, if he himself knew that this person was suspect—that the person he was about to administer an oath to was suspected of being a liar who would swear falsely—then no. Therefore, if the document came before him, and so on. What is he saying? The case Maimonides is basing himself on here is also the Talmud we cited in tractate Shevuot, and there is another Talmudic passage about Rava, who wanted to administer an oath to a certain woman who was obligated to swear—she was being sued, and Rava wanted to administer the oath to her. Then Rava’s wife came and whispered to him: I know her, she’s a pathological liar, don’t make her swear. The religious court was in his courtyard, in one of the rooms of his house—that was the court. His wife came in and whispered in his ear: I know this woman who is standing here; she is a pathological liar, don’t make her swear. And what Rava did was transfer the oath to the opposing party. Because that’s what is done when someone who is suspect with regard to an oath becomes obligated to swear: you obligate the plaintiff to swear, and then he can swear and collect. Usually, the oath is taken by the defendant. He swears and is exempt. But if the one who is supposed to swear—the defendant—is suspect, then you transfer the oath to the opposing party, and then the plaintiff swears and collects. And that’s what Rava did there. Now you have to understand that this infringes on the defendant’s rights, because in effect you hand his fate over to the plaintiff. Now the plaintiff will decide whether to swear; everything depends on the plaintiff. If the plaintiff decides to swear, he will swear and get the money. And that’s what Rava decided to do, and in fact this is the source Maimonides brings for the idea that in monetary law—even though we are talking here about the testimony of a woman, and not only a woman, but the judge’s wife, a relative, a woman, not trustworthy in any way as a witness—she told Rava that the woman before him was a liar, and that was enough for Rava to change the judgment. To transfer the oath to the opposing party and infringe on that woman’s rights. I remember once Nahum Rakover used to run weekend seminars for jurists—Nahum Rakover, Thursday, Friday, Sabbath, these kinds of weekends. And once he asked me to lead one of those workshops, and we studied this topic. And when we studied it, the jurists there—especially the jurists there—really revolted, they were tearing their hair out. What do you mean, the judge’s wife comes in, whispers something in his ear—who knows, maybe she hates that defendant—and then he changes the judgment, infringes on the defendant’s rights? It’s unthinkable. Think about something like that happening today. The judge’s wife walks into court and whispers to him: listen, I know that guy, he’s a liar, don’t believe a word. The judge says okay, you’re not trustworthy, pay the other side money. The press would crucify that judge and his wife together. So they said: how can that be? What kind of primitive Jewish law is this? So I told them—and this is just an anecdote in parentheses—I said: look, put yourselves in Rava’s shoes. Rava is now standing before this woman, and the law says he has to administer an oath to her. Now his wife comes and tells him, and he knows his wife. He knows that in this case she isn’t lying—let’s assume for the sake of the discussion that he knows her well. So from his perspective it is clear that if he lets the defendant in front of him swear, she will keep money that is not hers. That is simply theft. Now he has a dilemma. Should he go by proper procedure? Say, what, my wife can’t interfere in these matters, and so on—and then knowingly give money to a liar? Or not? What is proper procedure? In the end I want the money to end up with its rightful owner. I do not want to give money to liars. Forget proper procedure. So if I believe my wife and she tells me that the woman before me is a liar, then what am I, an idiot banging my head against the wall? I’ll make her swear even though she’s a liar and leave the money with her even though I know she’s a thief? There is no logic in that. I told them that I think this very much depends on the question of how much trust you place in the system. Understand? If you trusted the judge sitting in judgment—if you trusted him—then I think that even today I would expect that if the judge’s wife tells him I’m lying, he should take the money away from me. Fine—if I trust him that he really knows his wife and knows there are no ulterior motives and plots here, but that this is really the case. True, she is his wife, but that’s what he has; he has no other witness. But he has this witness and he knows she is telling the truth. So he does not want a sinner to profit. So if you trust the judge and you are not afraid he will abuse this and screw over people he wants to screw over, or whom his wife wants to screw over, then apparently that is what really should be done. Except that the modern system—again, this is not slander, I’m saying it as a fact, a description—the modern system is not willing to place trust in people in that way, not even judges. And I understand it; it’s not criticism. I’m just saying, notice the background here. Because we are afraid there will be tricks; give a judge freedom to do what he wants, and he’ll do what he wants. You can’t rely on it; maybe this judge will be fine, but maybe that judge won’t be fine. You can’t build a system that way. You basically have to build behind a veil of ignorance, yes? You don’t know which judge will be before you. The rules have to be blind. Lady Justice holding the scales is blind because justice is supposed to be blind. It doesn’t matter whether this particular judge really is righteous and I’m sure he wouldn’t do such a thing; maybe the next judge would. Therefore it makes sense to set certain rules that are fixed rules; they don’t change from one judge to another. My trust in the judge, and so on—no, you can’t rely on what your wife whispers in your ear. That’s not how it works. There are rules of proper procedure or rules of legal process, and that’s how it should work. I understand that conduct, but one also has to understand the other side. The other side has a great deal of logic too. If the point, the basic starting point, is that you have full trust in the judges who sit in judgment, then I think this Talmudic passage is completely right. That is the correct way to run a legal system. Now the question is whether that trust is justified, and what things were like then. Maybe back then too it wasn’t appropriate to trust, but people trusted because they were naïve—whatever, I’m not getting into those discussions now. I mean on the principled level. On the principled level, if you trust the judge, then in my view this is the right way to act.
[Speaker B] Well, I also, as a jurist, revolt against this, but what I want to add is this. First of all, that is exactly the statement that justice must not only be done, it must also be seen to be done. And therefore you can’t go with this approach.
[Rabbi Michael Abraham] But I don’t agree, I don’t agree that you can’t. What you said—that justice has to be seen—I agree with; that’s proper procedure and legal process. But where there is trust in the judges, then justice is also seen, because it is done this way.
[Speaker B] When he believes that it’s objective.
[Rabbi Michael Abraham] But the trust—
[Speaker B] —is a subjective matter; you can’t show it through evidence.
[Rabbi Michael Abraham] Assume as a matter of fact, assume that the public trusts the judge, for the sake of argument. If that’s the case, then do you agree with me that it makes sense to act this way? It may be that this is not the case, and then we won’t do it, but on the principled level.
[Speaker B] No, no, I still don’t think so, because I think the word “seen” means precisely something objective. I want that when the judge gives a ruling, he should explain it not on the basis of my trusting what he is doing, but because I know what evidence he relied on. But I’ll say more than that: I could accept it if Rava had done something very simple. Maybe he couldn’t, but today he could. Suppose his wife whispered to him at home what she says, and then he puts her on the witness stand in court and she says what she says.
[Rabbi Michael Abraham] Yes, but in Jewish law that doesn’t work, because a woman is a relative of the judge.
[Speaker B] I understand, but you’re trying to project this also onto the situation supposedly today, so I’m saying—
[Rabbi Michael Abraham] No, I project it onto today’s situation in contexts relevant to today. If there were inadmissible evidence—there were inadmissible evidence for some reason, doesn’t matter right now—but you know it is correct evidence, then there is room to discuss such a policy. It depends on trust. Look, I’ll tell you more than that, Ezra: even in today’s court, when the judge says, “This witness strikes me as credible,” that carries some weight. Why? Where is the objectivity? Where is “justice must be seen”?
[Speaker B] No, the judge says that this witness strikes him as credible after both sides have already examined that witness, and then he says, after hearing all the cross-examinations, I believe him—and that is his right, because there is evidence. One more thing, just one more thing to say: this matter of trust also prevents you from appealing, because you have nothing to appeal on. You can’t now go to a higher court and say it wasn’t right that they trusted this judge. That’s not a matter for appeal. Therefore there has to be a ruling that says what the evidence is, and then I can go to the appellate court.
[Rabbi Michael Abraham] Completely wrong, completely wrong. If I trust the judge, that does not mean I cannot discuss his evidence. I trust that he is not lying, but he still may be mistaken.
[Speaker B] Fine, but if you’re already bringing evidence, then the matter is over, then everything is fine.
[Rabbi Michael Abraham] And in Jewish law you can—Jewish law also has evidence. In monetary matters one can also judge even without evidence, but there is also evidence, and on that one can appeal. By the way, even with us you generally cannot really appeal to a higher instance regarding the judge’s impression of the witnesses before him.
[Speaker B] Right, right, that’s true.
[Rabbi Michael Abraham] That’s true. The parallel to today is greater than it seems at first glance. Fine, but that’s another discussion. In any case, for our purposes, that was just a side point; it’s not our topic. I only want to say this: Maimonides says that in monetary law you can basically follow impression, inadmissible evidence, as long as you are convinced that this is the truth—that is what you should do. What about “By the testimony of two witnesses a matter shall be established”? Yes, by the testimony of two witnesses a matter shall be established if you yourself do not have some certainty, some inner evidence of your own. Then you go with the laws of evidence, with formal law, and then you need two witnesses and so on. But if you have become convinced in one way or another that this is the truth, then you are exempt from all rules of procedure and evidence law and all those matters. That is the law in principle. By the way, just so the picture is more complete, in practice this is not done today. “All these matters are the essence of the law,” says Maimonides, “in practice.” The source is the Rif, really; Maimonides is just bringing it. “But when courts that were not worthy multiplied, and even if they were worthy in their deeds, they were not sufficiently wise and insightful”—I translate this, by the way, as a decline of the generations not in the judges, but in the community. The community no longer gives the judges the trust that it once did. And I think that is the point. The moment there is no trust, justice must be seen and not only done. Therefore it does not have to be that Maimonides thinks the judges are necessarily worse, but rather that the relation between the judges and the community in which they function is different; it is basically modern, like today, where you are not really willing to trust the person’s subjective decisions. And once that is the relationship, then indeed justice has to be seen and not just done. In any case, Maimonides says, most Jewish courts agreed that they would not transfer an oath except on the basis of clear evidence, and would not invalidate a document and deprive someone of his established legal claim on the testimony of a woman or some otherwise disqualified person, and likewise in all other laws. And the judge should not rule on the basis of his own confidence—just his own intuition—or on the basis of his own knowledge. By the way, “his own knowledge” means something he himself saw, like with the ruin of Shimon ben Shetah, where that is clear evidence, but justice does not appear, because you don’t have two witnesses who say it, and therefore you cannot judge on that basis, “so that every common person should not say, ‘My heart believes this one and my mind relies on that one.’” Again he ties it to the judges, saying that today judges are not ordained, so they may really be mere laymen. Laymen you certainly cannot give the privilege of acting in this way. As I said before, I think that by and large most judges today are not laymen. You can disagree with them, whatever, but they are not laymen; and people appoint scholars. I think the more significant point is the community’s attitude toward the judge, not the judge’s quality in itself. The community no longer gives him the innocent or naïve trust that people once gave judges, and therefore justice must be seen and not only done. In any case, for our purposes, this rule is canceled. That is the law in principle, but in practice today one does not act this way. So that’s just to complete the picture. He says: “Nevertheless, if one trustworthy person testified about one of these matters, and the judge became convinced that he was speaking the truth, he waits in the case and does not reject his testimony”—because it’s only one witness—“and he discusses the matter with the litigants until they admit the witness’s statement or reach a compromise or the judge withdraws from the case.” Meaning, you can try to use various maneuvers to bring the truth into effect, but you cannot decide the case without clear evidence. And therefore he then says in halakhah 3: this is a “deceitful case.” “If the judge knows in a case that it is deceitful, he should not say: I will decide it, and let the chain of responsibility hang on the necks of the witnesses.” Now this is the opposite case: witnesses come and testify that Reuven owes one hundred shekels to Shimon. Now my impression is that the witnesses are lying. This is the opposite—not that I have no witnesses and only some impression about who owes whom money, but that there are witnesses and my impression is that they are liars. Here Maimonides says this is what is called a deceitful case, and it does exist even today. You should not decide the case based on the witnesses. Do not decide it based on the witnesses, because this is a deceitful case. Now, if according to the procedural rules you found nothing—you did not prove the witnesses to be false, you found no contradiction in their testimony—you cannot do anything. So what should he do? “He should examine and interrogate extensively, with the searching and interrogation used in capital cases”—in monetary cases one does not need such searching and interrogation, but he should do it like they do in capital cases in order to test them thoroughly. “If according to his judgment it appears that there is no deceit, he decides the case on the basis of the testimony. But if his heart still troubles him”—he found no contradiction, but after all the examination and interrogation he is still not inwardly appeased. His intuition is still that something fishy is going on. “If there is deceit in it, or his mind does not rely on the words of the witnesses even though he cannot disqualify them, or if his opinion inclines that one of the litigants is a deceiver and has coached the witnesses even though they are valid witnesses and testified innocently,” and so on—then he does not decide the case. He withdraws. He says: “All these things and the like—it is forbidden for him to decide that case; rather, he should withdraw from this case and let someone whose heart is whole with the matter judge it. And these things are entrusted to the heart, and Scripture says: ‘For the judgment belongs to God.’” Fine. That means this does remain, and it is relevant today as well: if the judge has reached the conclusion that the evidence before him is good according to the formal laws of evidence, but something here stinks, then he should check again and again and conduct examinations and interrogations far beyond what is required. In monetary cases one does not conduct searching interrogation; that is a rabbinic enactment, so as not to lock the door before borrowers. In capital cases one does. He says: in monetary cases too, do searching and interrogation like in capital cases. If you still do not reach a conclusion, the searching and interrogation did not produce a contradiction, but your heart is still not settled—you still feel something here smells bad—then you say: listen, I am not willing to judge this case; I withdraw, find another judge. Let someone else judge the case if his heart is at peace with it. That is something a judge is permitted to do even today. And one cannot force a judge to decide a case with which he is not at peace, even though from the standpoint of evidence law the formal evidence is sufficient. He can say: look, I did what I could, and the rest is God’s business. But I—why should you enter into the hidden counsels of the Merciful One? No. A judge does have to enter into the hidden counsels of the Merciful One. And if he is not at peace with it, he should withdraw, he should not sign the judgment. By the way, in this context, it reminds me of a similar rule. I once spoke about this when we discussed a scriptural decree. It is commonly said that the disqualification of relatives is a scriptural decree. Witnesses who are relatives—either related to one another, or related to the judge, or related to one of the litigants—are disqualified. The disqualification of a relative. It is usually thought that this is a scriptural decree because a person does not sin for no reason, and everyone has a presumption of fitness. So the fact that the Torah disqualified relatives’ testimony is a scriptural decree. So say Maimonides, the Tur, the Shulchan Arukh—the source is a Talmudic passage in tractate Bava Batra that says it is a scriptural decree. Ordinarily people understand: what does “scriptural decree” mean? Basically that the relatives are telling the truth, but there is a scriptural decree not to rely on their testimony. Right? The problem is admissibility, not credibility. The testimony is inadmissible not because it is unreliable. So I have often asked people—we are so used to gliding over this statement that it is a scriptural decree without much thought. Why? Because think about two witnesses who are relatives testifying that Reuven murdered Shimon. At most I say: they are relatives, I have no proof he murdered Shimon, what can I do? I can’t punish him. Not ideal, but fine, that’s what there is. But let’s try thinking of the opposite case. Two witnesses come and testify that Reuven murdered Shimon. Now two relatives come and prove the first witnesses to be false. Okay? Now if I disqualify the testimony of the relatives, what is the law that results? Reuven is executed as a murderer, right? There are two witnesses that he murdered, and the rebuttal is disregarded because the rebutting witnesses are relatives. But notice: if this is a scriptural decree, then the rebutting witnesses are actually truth-tellers. There is just a scriptural decree not to accept their testimony. That means that when they rebutted the testimony of the first witnesses, then in truth the first witnesses are liars, and Reuven is not a murderer. But there is a scriptural decree not to accept the testimony of relatives, and therefore we will execute Reuven even though he is not a murderer at all. I asked people: what do you think is done in such a case? So they said: what do you mean? It’s a scriptural decree, so obviously you have to execute Reuven because he’s a murderer. Which is unbelievable. If this really is a scriptural decree in that sense—that these witnesses are truth-tellers, only their testimony is inadmissible—then obviously any court that executes Reuven ought itself to be punished. Certainly it should be removed from the bench. You are executing an innocent person—are you out of your mind? What can we do, though—the Torah said not to accept relatives’ testimony. That is exactly what a deceitful case is for. Withdraw from the case. After all, you know he did not murder. True, according to the formal laws of evidence you have no refutation of the testimony that Reuven murdered. But here you actually know—it’s not just a gut feeling. You have truthful witnesses, only they are relatives and their testimony is inadmissible, and they are speaking the truth. In terms of the truth, you know he did not murder. How can you execute someone like that? Obviously such a thing is unthinkable. Right. So a deceitful case—what we just saw in Maimonides—solves that issue too. A deceitful case basically says: listen, judge, you are not supposed to issue a ruling that you are not at peace with. If you are not at peace with it, then even if the formal laws of evidence are perfectly buttoned up, if you are not at peace with it, do not sign the ruling. True, you are not allowed to sign the opposite ruling. It would not be an acquittal. It would be what today is called acquittal due to doubt, or something like that—or maybe not even that, but rather withdrawal, withdrawal from the case, replacing the panel. It would not be acquittal. You cannot acquit against the laws of evidence. Okay? And I am talking now, of course, about monetary law, not criminal law, because up to now that’s what we’ve been discussing. You cannot force me to sign a ruling that I am not at peace with. That is what is called a deceitful case. Fine. So up to this point we have seen monetary law. What happens in capital law? So that was Maimonides in chapter 24 of the Laws of the Sanhedrin. Now this is Maimonides in chapter 20. In chapter 20, halakhah 1, it says as follows: “A court does not punish on the basis of estimation, but only on the basis of witnesses with clear evidence.” The background is of course what we saw in monetary law. In monetary law, estimation—in principle, not nowadays, but in principle—estimation is sufficient to decide the case. Nowadays they canceled that, but in principle it is enough. In capital cases, even in principle—even the courts of old, which judged monetary matters on the basis of estimation—this was only in monetary matters, only civil law, not criminal law. In criminal law, when you come to punish someone—lashes, death, and so on—no. Only on the basis of witnesses. Not on estimation. Circumstantial evidence is not accepted, certainly not impression. And he now brings the Talmudic case we saw in tractate Shevuot: “Even if witnesses saw one person chasing another, and warned him, and then lost sight of them, or they entered after him into a ruin and found him slain and convulsing, with 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 concerning this and the like it is said: ‘Do not kill the innocent and the righteous.’” Astonishing, the verse he brings—“Do not kill the innocent and the righteous.” The man is a murderer! What does “Do not kill the innocent and the righteous” mean? And Maimonides explicitly does not make the distinction here between weaker evidence and stronger evidence, unlike Tosafot. Tosafot says that there the circumstantial evidence was not one hundred percent. But Maimonides says no: there is a difference between circumstantial evidence and witnesses. It makes no difference how strong the circumstantial evidence is. So the verse “Do not kill the innocent and the righteous” is very strange, because here you are not killing him even though it is obvious that he is neither innocent nor righteous, but you have no witnesses. And therefore, if two people testified against him that he worshipped idols—one saw him worship the sun and warned him, and the other saw him worship the moon and warned him—they do not combine, as it is said: “Do not kill the innocent and the righteous.” Since there is some aspect by which he can be declared innocent and righteous, do not kill him. Do you understand how he reads the verse? “Do not kill the innocent and the righteous” does not mean do not kill one who is innocent and righteous, but do not kill one for whom there is some way to clear and vindicate him, even if only formally, even though in reality he is obviously not innocent and righteous. “Innocent and righteous” here is figurative language. Meaning: if there is some basis to clear and vindicate him formally—obviously he is not innocent and righteous, but formally he is—then you cannot kill him. That is how Maimonides reads the verse. So in the end—and this is the Talmudic passage we saw in tractate Shevuot, the one we saw earlier—if so, Maimonides indeed rules that unlike monetary law, where in principle—never mind what happens nowadays—in principle all I want is the maximum degree of certainty. And if the maximum degree of certainty comes from circumstantial evidence, that is also fine; I do not care whether it is witnesses or not. All that matters is that it be clear that this is the ruling. By the way, in civil law today, for example, it is accepted that in civil law you do not need proof beyond a reasonable doubt; it is enough that the balance tilt in favor of one side—fifty-one percent, not ninety-five percent, right? Because if we demand ninety-five percent, it follows that we are justifying the other side on the basis of five percent, which is not logical. Therefore in civil law, the moment you justify this side, you harm the other side. It’s not just one person, as in criminal law. So there it is clear that you do not require proof beyond a reasonable doubt; you require that the scale tilt more in favor of one side. Fine, that has significance for possession, but it doesn’t matter right now. It is understood that the burden of proof lies on the plaintiff. But if in the end we are left with evidence that is not one hundred percent, but it still tips the scale, then indeed he gets the money, even though he is the plaintiff.
[Speaker B] But not only that—if the evidence cancels itself out, meaning he is in a situation where there is no fifty-one percent, since the burden of proof is on the plaintiff, his claim will be rejected.
[Rabbi Michael Abraham] Yes, of course, that’s what I’m saying. That’s why you need fifty-one percent. Fifty percent is not enough. There is a certain advantage to the person in possession of the money. But that advantage, unlike in Jewish law, fades away with every additional percentage. Meaning, once there is fifty-one percent in favor of the plaintiff, he gets the money.
[Speaker B] Okay? Possession can shift the burden of proof.
[Rabbi Michael Abraham] Yes. So possession determines that he has to bring the proof, and when he brings proof, even if it is only fifty-one percent proof, that is enough. Correct. Yes. So in Jewish law now, some wanted to claim it that way. Tosafot basically understood it that way. Tosafot says: why is circumstantial evidence not enough in Jewish law? Because it is not strong enough. But if it were strong enough, I would not care whether it is circumstantial or testimony. But Maimonides makes a categorical distinction. In capital cases, unlike monetary cases, witnesses convict; circumstantial evidence, no matter how strong it is, does not convict. And the difference is not in the strength of the evidence, but that there is some categorical difference, at least in criminal law, between—yes. By the way, what would happen in monetary cases if there were prisoners there in prison—that is an interesting question, I don’t know. We would need to investigate. What would happen in a case like prisoners in prison, but where the proceeding is not criminal but civil, meaning monetary? A hundred people stole money, I don’t know, and divided it among themselves, ninety-nine—okay. And now the guard sues for one shekel; they stole ninety-nine shekels from him. Now he sues each of the attackers for one shekel. The question is whether you can demand one shekel from each one. Because in monetary law all you need is fifty-one percent. Statistical evidence—even though it is statistical—there is fifty-one percent here. Question. I don’t know, I don’t know what happens in the legal world in this context. Ezra, do you know? You’re muted. You’re muted.
[Speaker B] I just think—I haven’t dealt with this in practice—but I think the problem with statistical evidence is that before the judge there is no possibility of conducting a cross-examination of that specific person. Because you’ll bring the Central Bureau of Statistics, it will say it, and the judge can’t do anything with it. But there is still ninety-nine percent here.
[Rabbi Michael Abraham] What? But there is still ninety-nine percent here.
[Speaker B] Yes, but he will have a problem ruling in someone’s favor on the basis of statistics, because he has to create a situation where there is some evidence regarding this specific person.
[Rabbi Michael Abraham] And the question is why. Why aren’t these ninety-nine percent that are here enough? It’s like what you said earlier, that you need the version of the one holding the dripping knife. There too it’s exactly the same thing: after all, you have one hundred percent, so why should I care about his version? The very fact that you need an additional version means there is some weakness in this kind of conviction.
[Speaker B] So I’m saying that with statistics, its weakness is that you are unable to carry out a cross-examination regarding the individual person.
[Rabbi Michael Abraham] What? Without cross-examination, I have ninety-nine percent.
[Speaker B] I don’t think a judge could do that.
[Rabbi Michael Abraham] No, obviously not, that’s obvious. We’re only looking for the explanation why.
[Speaker B] Isn’t it enough as an explanation that the judge wants to be in a position where he can hear the person’s own version?
[Rabbi Michael Abraham] If there’s ninety-nine percent, then we’re done. What do you mean? The whole point of hearing the opposite version and examining and probing and checking is to get to ninety-nine percent. Now here I have the ninety-nine percent, so what is the problem?
[Speaker B] No, not in order to get to ninety-nine percent, but in order to get to what is called finding the truth, in the sense of whether this man is telling the truth or not, whether the witnesses testifying against him are telling the truth or not.
[Rabbi Michael Abraham] But I have ninety-nine percent that he is not telling the truth—that he is a murderer. I’ve reached it, I’ve reached the truth at ninety-nine percent. That’s it. What could be better than that? To me, that’s not an explanation.
[Speaker B] The requirement of the possibility of cross-examination is a requirement that turns the whole matter into something substantive. If I can’t cross-examine, I’m not willing to rely on evidence that doesn’t allow me to—
[Rabbi Michael Abraham] But the evidence is ninety-nine percent, so—
[Speaker B] —it allows me to.
[Rabbi Michael Abraham] Why not rely on it? Why do you also want the possibility of cross-examination? You have ninety-nine percent.
[Speaker B] Because the issue of being able to question this specific person is a substantive element in legal procedure. You can bring statistics as supporting evidence. That’s words. That’s words.
[Rabbi Michael Abraham] I have ninety-nine percent. You’re assuming again—and I agree with the result, the bottom line is agreed. We don’t convict. I just don’t think this explanation holds water. What does it help? If I have ninety-nine percent, why should I care whether it is possible to cross-examine him or not?
[Speaker B] So again I say: cross-examination is part of the search for truth. Statistics is not the search for truth. What is the search for truth?
[Rabbi Michael Abraham] Ninety-nine percent truth is enough for criminal law. So I have that. That’s it. That is finding the truth—I found the truth to ninety-nine percent. Again, there is a very strong intuition that this—
[Speaker B] No, but it’s not just intuition, because the statistics piece is, somewhere in the end, words. Let me put it differently. A witness will come and testify that someone committed murder, okay? And that witness couldn’t make it to court and they couldn’t—or he already gave testimony but they couldn’t cross-examine him—they won’t accept it, because the issue of cross-examination is a basis for the judge’s decision. With statistics, you cannot conduct a cross-examination.
[Rabbi Michael Abraham] So you need to distinguish between two possibilities. If you really have a genuine concern that this witness is lying or mistaken or whatever, and cross-examination can uncover that, then I agree completely, no problem, because then you really don’t have ninety-nine percent as long as you haven’t done the cross-examination. But if you do have the ninety-nine percent, then I would ask the same question about the witness too. So you still haven’t answered anything; that’s an English-English dictionary.
[Speaker B] No, even if the judge decides that he believes the direct testimony, if there is no cross-examination he still won’t be able to rule.
[Rabbi Michael Abraham] I ask the same question there too. So you still haven’t answered. If so, then I’m bothered by that case too. In my opinion, that’s not an answer. If there is ninety-nine percent, ninety-nine percent is beyond a reasonable doubt. No, no.
[Speaker B] In Israeli law, the issue of the possibility of—
[Rabbi Michael Abraham] —cross-examining—
[Speaker B] —cross-examination is substantive.
[Rabbi Michael Abraham] I know. I’m asking why. What? You’re repeating the bottom line, which I agree with; I’m asking why.
[Speaker B] First of all, you’re allowed to set that rule, the evidentiary rule that says I want every testimony to be open to cross-examination. Why? Because once you don’t allow the possibility of examination, you haven’t exhausted the process of finding the truth. It’s part of finding the truth.
[Rabbi Michael Abraham] Why haven’t you exhausted it? If I got to ninety-nine percent, ninety-nine percent is enough certainty, so I exhausted the search for truth, that’s all. Again—
[Speaker B] If I had cross-examination—
[Rabbi Michael Abraham] —then I’d have ninety-nine and a half percent. Why should I care?
[Speaker B] No, let me do it differently. The prosecutor in the criminal trial will bring the statistical evidence, okay?
[Rabbi Michael Abraham] Fine.
[Speaker B] The other side will get up and say: listen, I’m unable—there’s nothing for me to cross-examine in statistical evidence. The judge will come and say: since it is impossible to cross-examine, as far as I’m concerned the truth has not been exhausted. And that’s it. Words! Words! If you have ninety-nine percent—
[Rabbi Michael Abraham] —that he is guilty, then the truth has been exhausted, that’s all. Just words.
[Speaker B] But that’s only one side of the coin—the burden of proof.
[Rabbi Michael Abraham] No, it’s not only one side of the coin, because it is really ninety-nine percent. If you say that truly, without cross-examination, there wouldn’t be ninety-nine percent here, then you’re right. But I’m saying: if it is ninety-nine percent, and it’s clear that here it is ninety-nine percent, then why should I care that there was no cross-examination? That’s exactly the point. Unless you say it’s a formal rule, and a formal rule explains nothing.
[Speaker B] No, no, it’s not just a formal rule. I really feel it’s also right that for every piece of evidence brought in there should be a possibility to—what? I also feel it. The question is why. I’m asking why. Because you know what? It may be that if there were a possibility of cross-examination, then maybe he could expose flaws in those statistics—not in the sense that the statistical truth is wrong, but that it doesn’t fit this case, for example—but I have no one to examine.
[Rabbi Michael Abraham] So that, you can raise yourself. Meaning, if there really is a possibility that this person is not the culprit at a rate of one in a hundred, then we really are not at ninety-nine percent. But if the ninety-nine percent calculation is correct—and obviously in this case it is correct, because there are a hundred people, only one of whom did not murder—take some random person from among them. If you took some random person from among them, what is the chance that you happened to pick the one who didn’t murder? One percent. That’s it. It’s mathematics. Again, my intuition is like yours; we all agree that we do not convict. I’m looking for the explanation, what is the explanation. Okay. So in short, the point is that in Maimonides, what is written there is really exactly Papineau’s distinction. I’m talking only about criminal law right now, but I think he is also talking about criminal law. For criminal law alone, it is exactly the same distinction. Circumstantial evidence is basically statistical evidence. The circumstances say that there is a ninety-nine percent chance that he is guilty. As opposed to what is non-circumstantial evidence? That is testimony—evidence where there is direct evidence about the case, where there are witnesses who saw it. Meaning, the distinction between circumstantial and non-circumstantial evidence is the same distinction as the one between witnesses and statistics. And we see that in Jewish law too it is like that, just as in ordinary legal systems. Now maybe one more sentence before—I see our time is already up—just by way of introduction. I just want to introduce a concept that will come up again later. There is a pamphlet on migo that also appears on the website, and in the last part—the second part—of that pamphlet, it talks about what I call legal reasonings. In procedural law, and even in the laws of evidence, there are sometimes considerations whose basis is not only the desire to get at the truth, or at least in this particular case. There are considerations that one can call legal considerations. For example: the burden of proof is on the one seeking to extract from another. There are people who think that this is simply because most likely the person holding the money is really its owner, and we already spoke about that on one of the previous occasions—that that is not necessarily true. There is no indication whatsoever that it is true. I don’t think there are statistics showing that in most monetary claims the defendant is right and not the plaintiff. Of course that is how it comes out, because the burden of proof is on the plaintiff, so if he did not manage to meet the burden of proof, then the money is left with the defendant, and de facto it comes out that the defendant wins in most cases. But I’m asking about that itself: why do we prefer the defendant and place the burden of proof on the plaintiff? So here there is a legal logic, but it is not a probabilistic logic. And it is not the result of the probabilistic assumption that most likely the possessor is really the owner of the money; rather, there is a legal logic to acting this way. That is what I call legal reasonings. Migo, strength of claim, all kinds of things of that type—right, like that Chazon Ish, yes, exactly that way—that an implausible claim changes the presumption of possession. There is an entire sector of reasonings that are legal reasonings, and they are not halakhic / of Jewish law reasonings: that this is the proper way to act even though statistically there is no bias in favor of that side, but legally it is still right to rule in his favor or to obligate the other party in law. And someone could come and say that perhaps behind the distinction between circumstantial or statistical evidence and testimony, there is some kind of legal reasoning, a legal intuition that this is the proper way to proceed, because we feel that this is right, as Ezra came back and argued earlier. There is such a feeling, that this is right. But then, although mathematically, statistically, it doesn’t hold water—because eyewitness testimony is ninety-five percent and here it is ninety-nine percent—there is still an intuitive feeling that this is the proper way to proceed. Not because the chance that he is a murderer is really high—the chance that he isn’t is very small, one percent. But still, legally we have some tendency to think, for example, that without cross-examination or something like that you cannot convict a person—some legal reasoning. So that is perhaps the line of approach that you meant before, Ezra; that may be a natural approach. But I would like—and we’ll probably do this next time—to bring explanations here that will pour content into this legal reasoning, and after I pour content into this legal reasoning I will want to argue that maybe there is actually a probabilistic reasoning here, not a legal one, and that, as far as I know, no one has yet argued in these discussions—at least not in the literature I’ve read. And I’ll want to argue that that is what stands there behind these things, but we’ll do that next time.
[Speaker B] Just one sentence: behind all this, behind this whole debate, there is one thing. When you relate to a legal system as a system that is supposed to bring justice—and there is no such thing—the system has to bring law, and therefore it works with legal rules. It does not bring justice; you are trying to bring justice.
[Rabbi Michael Abraham] No, that’s what I said. Those are the last sentences I just said now. That’s what I call legal reasonings.
[Speaker B] No, okay, but I’m saying that this is not procedural, it’s substantive; it’s substantive because that is how law is built.
[Rabbi Michael Abraham] Fine, that’s what I call procedure—what difference does it make what you call it. But if you bring substantive explanations—I will want to argue that there are substantive explanations here. We’ll be glad to hear.
[Speaker B] No, not explanations—I call it procedural; it doesn’t matter what you call it.
[Rabbi Michael Abraham] Okay, so we’ll talk about that further.
[Speaker B] Good, thank you very much. Thank you very much, Sabbath peace.
[Rabbi Michael Abraham] Sabbath peace, peace.