The Thought of Rabbi Gedaliah Nadel – Derash in Halakha – Lesson 4

This transcript was produced automatically using artificial intelligence. There may be inaccuracies in the transcribed content and in speaker identification.

🔗 Link to the original lecture

🔗 Link to the transcript on Sofer.AI

[Speaker I] That’s a novelty, that itself is a novelty; it’s not logic. If it’s a correct rule, one that’s always correct…

[Rabbi Michael Abraham] So that’s exactly…

[Speaker I] But it’s not logic by negation…

[Rabbi Michael Abraham] Right, but that’s exactly the point. So it’s the same with an a fortiori argument. In an a fortiori argument too there’s a novelty that isn’t logic, namely: if Tooth and Foot are more severe than Horn—well, actually Horn is more severe than Tooth and Foot in the public domain—then apparently it’s more severe than Tooth and Foot in every respect, in every area, including the damaged party’s courtyard. That’s exactly the point at which the a fortiori argument collapses and stops being logic. You can always translate it this way: if you put in the data we received from Sinai and call them premises, in the end it will always come out as a deductive argument, of course. But that’s exactly the problem. So fine, okay, that’s just an anecdote. I think either he didn’t quite understand what a syllogism is, or he didn’t mean what I understood him to mean—I don’t know. But it can’t be; in other words, it’s not a necessary logical inference. So there are refutations against it. What do the refutations attack? The refutations always attack the hierarchy reasoning. Sometimes something isn’t really a refutation; it’s just that one of the data points was simply wrong—it turns out one assumed datum was wrong. That’s not a refutation; you just found an error. You find that kind of error in mathematical proofs too. You made a mistake in one step, in moving from here to there—that’s not a refutation. A refutation means that your whole construction is internally correct, but when I put it to the test, I find a contradiction—like when we test a scientific theory. Right? So a scientific theory says that every object with mass falls to the earth. What would a refutation be in that context? I’d take, say, this thing, let go of it here, and it wouldn’t fall. That would be a refutation of the scientific law. Not that I found an error in the line of inference used by the scientist who formulated it—say Newton—but rather that I subjected it to an empirical test and found a counterexample. Now a refutation of an a fortiori argument is the same kind of thing. A refutation of an a fortiori argument is not finding a mistake in the chain of inference; rather, I show a case where we see the opposite, or a counterexample. In other words, you assume that Horn is more severe than Tooth and Foot? Fine, I’ll show you that somewhere else Horn is not more severe than Tooth and Foot, but lighter. For example, Horn incurs only half-payment for the first goring, while Tooth and Foot incur full payment from the first damage. Just as an example. By the way, the Talmud there doesn’t make that refutation, and that raises the question of why; but that, for instance, would be a refutation. That refutation doesn’t attack the way I carried out the a fortiori argument; it only shows me that it doesn’t withstand the empirical test. You want to claim a general theory: every object falls to the earth. You have a general theory: Horn is always more severe than Tooth and Foot. I test that theory empirically—that is, I examine the data—and I see that it’s not true, not always. There are aspects in which Horn is not more severe than Tooth and Foot.

[Speaker G] But then, I think the a fortiori argument is correct when all the conditions are the same. Exactly.

[Rabbi Michael Abraham] What do you mean?

[Speaker E] If there is the…

[Speaker G] The case where we’re talking about the same conditions and the same place.

[Rabbi Michael Abraham] But it’s not under the same conditions. Tooth and Foot versus Horn—that’s not the same conditions.

[Speaker G] No, but when you bring an example from somewhere else where Horn is less severe, then the conditions there are different. Meaning, you first have to see. No, I agree, I’m just saying—you know—if you run a hundred meters, if you walk a hundred meters in ten seconds, then you’ll run it in less than that. Fine? But then I’ll say, no, that’s not true, because if you run you’ll fall, because there are rocks there and you won’t be able to do the hundred meters running. But then you changed the place, you changed the conditions.

[Rabbi Michael Abraham] Of course. That’s exactly what a refutation does. Let me explain it, I’ll explain.

[Speaker G] The question is whether a refutation can show that I’d do it under the same conditions. I’m talking about a place where the road is smooth and you can run and you can walk.

[Rabbi Michael Abraham] That’s a trivial statement, but it won’t help you with the a fortiori argument. There’s no need to refute it; you won’t be able to infer anything from it.

[Speaker G] You mean you won’t learn anything new from it?

[Rabbi Michael Abraham] You won’t learn anything from it, because whatever you learn from it is always a generalization. That’s what a refutation does. What does a refutation do? Let’s now look more closely at our a fortiori argument. Let’s take this a fortiori argument again. I learned that Horn is more severe than Tooth and Foot from two data points in the public domain, right? Notice: there’s a generalization lying here, because empirically all I really know is that in the public domain Horn is more severe than Tooth and Foot. But I want to apply that to the damaged party’s courtyard. So I say: ah, if in the public domain it’s more severe, then apparently for all intents and purposes it’s more severe. If so, then in the damaged party’s courtyard too it’s more severe. That’s how the a fortiori argument is built, right? Now I find a refutation. I say that Horn, say, on the moon is lighter than in the damaged party’s courtyard. Just for the logic, okay? On the moon, Horn incurs no liability and Tooth and Foot do incur liability. Fine? So that’s a refutation. What does that refutation do? It says: it’s not true that Horn is always more severe than Tooth and Foot. Sometimes yes, in the same circumstances—meaning, in the public domain—but sometimes not. And now the question is: which “sometimes” does the damaged party’s courtyard belong to? Does it belong to the “sometimes” of the public domain or to the “sometimes” of the moon? And basically what lies behind this is the question: there is some particular aspect—because that’s what lies behind the dry data—there is some aspect of Horn in which it is more severe than Tooth and Foot; there is another aspect in which apparently it is lighter. The aspect in which it is lighter is relevant on the moon, and the aspect in which it is more severe is relevant in the public domain. And now I ask: in the damaged party’s courtyard, which of these two aspects is the relevant one? If it’s the aspect of the moon, then the a fortiori argument falls. If it’s the aspect of the public domain, then the a fortiori argument stands. But since I don’t know, I have no inference. Since you can’t infer a conclusion—maybe yes, maybe no. A refutation only has to show that the conclusion is not necessary; it doesn’t have to show that it’s false. Otherwise it wouldn’t be a refutation; it would be an anti-refutation—in other words, an anti-proof. Because then I’ve also reached some conclusion. A reverse proof. Exactly. And I’ve already reached a conclusion about what the law is for Horn in the damaged party’s courtyard—that too is an a fortiori argument; that’s the reverse, the lighter-to-heavier direction. In other words, I reached the opposite conclusion. But a refutation always means you can’t reach a conclusion; the question remains open. That’s called a refutation.

[Speaker H] From an a fortiori argument you only learn through the refutation, because from an a fortiori argument that wasn’t refuted you learned nothing—everything was already hidden in the premises.

[Rabbi Michael Abraham] Right, right. From the refutation, yes—you learned that what you thought was hidden isn’t hidden. There’s another refinement in the premises. I think I said this once: the only argument from which we actually learn something is an argument in which we lost.

[Speaker G] That’s what Maimonides says, isn’t it? Right? He says that if a wise person defeated you…

[Rabbi Michael Abraham] Then first of all I gained from the argument that I learned something I had gotten wrong.

[Speaker D] Like my rabbi always used to tell us: an experiment that failed didn’t fail—it succeeded.

[Rabbi Michael Abraham] It taught us something else.

[Speaker D] An experiment that succeeded—we learned nothing.

[Rabbi Michael Abraham] So therefore, a refutation basically teaches us that the picture of the hierarchy is more complex than we thought. The hierarchy isn’t so simple. There is an aspect from which Horn is more severe than Tooth and Foot, and there is an aspect from which it isn’t. But since the relationship between Horn and Tooth and Foot is more complex, now who knows what is relevant in the damaged party’s courtyard. By the way, sometimes there is a line of reasoning that tells us which of the two relations is relevant to the damaged party’s courtyard, and then we’ll continue to derive the a fortiori argument despite the refutation. Because the reasoning tells us it’s more similar to what happens in the public domain than on the moon. So I’ll make the a fortiori argument. There are examples of this. I’ll make the a fortiori argument even though there is a refutation against it.

[Speaker J] But why, in a case where I’m missing this law of Horn in the damaged party’s courtyard, do I turn at all to an a fortiori argument? Why don’t I analyze it on its own terms? Why is Horn liable here and Tooth and Foot exempt there? Because Horn was intentional and Tooth and Foot weren’t intentional. And the public domain—if I obligate someone there—that’s much more severe than, or lighter than, the damaged party’s courtyard. Analyze it on its own terms, not because of comparisons and not because of an a fortiori argument—a logical analysis.

[Rabbi Michael Abraham] The simplest answer to that is that we do not derive law from the reason of the verse. Once you have data, and you start thinking why it is so—you do not derive law from the reason of the verse. It’s the dispute between Rabbi Shimon and Rabbi Yehuda; in Jewish law we rule like Rabbi Yehuda, that we do not derive law from the reason of the verse. But that’s a simplistic answer, and not a correct one.

[Speaker I] Most of the time, when there’s some specific exposition or some Torah midrash, there’s also some basic logic there.

[Rabbi Michael Abraham] I already said this, and I’ll say it again. I said: that answer is not correct, even though it’s the easiest thing to say. Why is it not correct? I’ll prove it from the case itself. The Rif, at the beginning of tractate Bava Kamma, explains why Tooth and Foot are exempt in the public domain. All in all, it’s a pretty reasonable explanation. It also appears in Maimonides in a slightly different version. Several medieval authorities (Rishonim) bring it, also the Rosh: that a person normally walks in the public domain with his animal, and other people need to understand that animals are walking there. And if you put produce there, then know that animals pass there. Nobody is supposed to keep his animal out of the public domain. It belongs to all of us. Let each person guard his own property. Fine? So that’s a very simple explanation. And the Rif brings it as the reason that in the public domain Tooth and Foot are exempt. And more than that—not only is he explaining the reason of the verse; the Rif is not a biblical commentator, the Rif is a man of Jewish law. When he brings this, the assumption is that it has a halakhic implication. And Tosafot indeed derive a halakhic implication from it. Tosafot claim that if there is a plank lying there extending from a private domain into the public domain, and the animal walks in the public domain and damages produce located in the damaged party’s courtyard by moving that plank, then because of the Rif’s rationale there too it would be exempt, because after all the animal is walking in the public domain and has the right to walk there. If this were some scriptural decree, right? Even though the produce is in a private domain. So never mind the specifics right now, but I’m showing you that the Rif did derive law from the reason of the verse in the very topic we’re dealing with. Not only in general—because of course people always derive law from the reason of the verse, as you said quite rightly—but here the Rif does it on this very topic. There really was Shlomo Meir—what, he derives law from the reason of the verse and so on—but people do it, they do it all the time. This Brisker naïveté that says we don’t ask why, only what—that’s nonsense. There’s no “what” without “why.” There is no such thing. If you don’t know why, how do you know what? What is explicitly written in the Torah—fine, I understand. But almost nothing is explicitly written in the Torah. Every small implication you draw, every inference, every interpretation, always presupposes some understanding of what is there—otherwise how do you do it? After all, you can generalize anything in a thousand ways. Every generalization assumes, implicitly or explicitly, some sort of understanding. There’s no such thing. You can’t define the “what” without asking the “why.” People always say there’s a difference between the definition and the reason. Without the reason there is no definition. That’s just an artificial distinction. No—you can’t do that. So what does it really mean not to derive law from the reason of the verse? Great question. I don’t have a good answer. Probably it means you don’t run too far with the reason of the verse, or something like that. The Rosh—Tosafot HaRosh—writes in tractate Bava Metzia on page 90 that if the reason is clear, then yes, we do derive law from the reason of the verse. So what did he basically mean? That “we do not derive law from the reason of the verse” means we don’t make overly far-fetched speculations with these analyses; but where it’s clear that this is the reason, then yes, you do derive law from the reason of the verse. So that’s why I said it’s not such a good answer. The better answer is that we can’t always make that analysis. Try making that analysis here, for example—it won’t be simple. It won’t be simple because when you… when you do the accounting in the public domain, the whole situation there is different from that in the damaged party’s courtyard. Because in the public domain the situation is that an animal really is walking. So I say: Tooth and Foot are damages that happen in the ordinary course of walking, so it is exempt, right. But Horn is intentional damage. Intentional damage may be something for which even in the public domain you should pay. Therefore in the public domain Horn is more severe than Tooth and Foot. Now let’s try to apply that to the damaged party’s courtyard. In the damaged party’s courtyard I don’t know what to do with that. In the damaged party’s courtyard you’re not allowed to be there either to eat or to gore. So is it more severe or lighter? I don’t know. Maybe in the damaged party’s courtyard it would be liable for half; maybe not liable at all; maybe liable for everything—I don’t know. I can argue it either way.

[Speaker J] Isn’t it obvious that your duty of care is much greater in the damaged party’s courtyard?

[Rabbi Michael Abraham] No, but I’m talking about Horn versus Tooth and Foot, not about the damaged party’s courtyard versus the public domain. I’m talking about Horn versus Tooth and Foot. I don’t know how to establish a hierarchical relation between them with respect to the damaged party’s courtyard. Again, this is only an example, so you’ll tell me, fine—then establish it along the axis of the damaged party’s courtyard versus the public domain, not along the axis of Horn versus Tooth and Foot.

[Speaker J] I wouldn’t go and make…

[Rabbi Michael Abraham] …that comparison, but that’s why, in that year when we studied the hermeneutical principles, I showed that it’s the same argument. I can translate one into the other. You can’t do one without the other. But for that I need the mathematics. But never mind; for purposes of the example, this argument is enough. When you want to examine the relation between Horn and Tooth and Foot, it’s not so simple. You can’t always really understand from what happens in the public domain what happens in the damaged party’s courtyard. And therefore you’re basically making some formal assumption that says: if this is more severe than that, then you may apply it also in the damaged party’s courtyard. If it doesn’t make sense, you don’t do it. Contrary to what the Briskers say, you don’t do it when it doesn’t make sense. But where it is reasonable—even though, again, if I had to do it without permission from Sinai to use an a fortiori argument, maybe I wouldn’t do it, because who knows, it may not be so. That’s why you need authorization from Sinai. In other words, the Holy One, blessed be He, tells me: yes, work with this logic; there is room to question it, and that’s okay. To punish by it—no; we do not derive punishments from an a fortiori argument. But to prohibit—yes. It’s good enough to prohibit. Okay, so the a fortiori argument works in a somewhat formal way. In the background there is some understanding—it’s not true otherwise. There are people who bring examples of a fortiori arguments. I think I told this story—I don’t remember if I told it—that in the Brisk Haggadah, at the end—I told this?—in the Brisk Haggadah, at the end, when it talks about the refrain “Who Knows One,” when they get to “thirteen who knows?” What are “thirteen who knows”? The thirteen hermeneutical principles. Right? Among the Hasidim, what are “thirteen who knows” in Hasidic Haggadot? The thirteen attributes of mercy—“The Lord, the Lord, compassionate and gracious.” Among the Briskers, “thirteen who knows” means the thirteen hermeneutical principles, of course. Right? So there he brings some teaching from Rabbi Chaim about a fortiori reasoning. And he claims that a fortiori reasoning is not a rational principle at all. It is a tradition from Sinai, and we use it in a completely formal, non-logical way—completely technical, mathematics. Why? He brings examples of it. He brings an example from the a fortiori argument of the “ninety”—they themselves carried the ninety by an a fortiori argument there; that’s an aggadic a fortiori argument, and there too there is logic—he’s mistaken about that. But one example often brought in this context is Rabbi Chiya’s first statement in tractate Bava Metzia on page 3: “A person’s own admission should not be stronger than the testimony of witnesses”—by an a fortiori argument. That is, if witnesses testify that you owe part of the sum, that obligates you in the oath of partial admission—certainly your own admission should as well. And there it seems very illogical. Never mind, I don’t want to get into every detail; I gave a class about it here after Sabbath, at Makom Menachem. That a fortiori argument seems very illogical. And that is always brought as an example that a fortiori reasoning is a formal principle; it is not ordinary human logic. That’s nonsense. Simply nonsense. You cannot say such a thing—it’s bupkes. You see in many places that there is logic to it, and where there is no logic, you don’t make an a fortiori argument. If you don’t understand the logic, think a little more. If you think a little more, you’ll find the logic there. And there is—I can show you the logic also in Rabbi Chiya’s first statement; that was the topic of the class I gave then. And also in the a fortiori argument of the ninety there is logic. There is logic in everything. If there is no logic, you don’t make the a fortiori argument. As I said earlier too: in a place where there is a refutation against an a fortiori argument, but my logic tells me that the damaged party’s courtyard is more similar to the public domain than to the moon, they will make the a fortiori argument despite the refutation. Because logic says that this is the closer case.

[Speaker C] And that’s not a refutation?

[Rabbi Michael Abraham] Exactly. But logic determines that; it’s not formalism. If it were formalism, the moment there is a refutation the a fortiori argument falls. You don’t go on thinking after you’ve already done the analogical mechanics.

[Speaker I] So that really reduces the weight of a fortiori reasoning as a hermeneutical principle. Meaning, not only do I not start without logic, but even when there is a refutation, sometimes I don’t stop.

[Rabbi Michael Abraham] Right, right. It’s all subject to logic. Now, as I showed before, logic alone might not have given me this. In other words, if there were no tradition, maybe I wouldn’t use this logic, because truly, who knows—in the damaged party’s courtyard, can you infer from Tooth and Foot to Horn or not? So the Torah says: you may make such inferences. Not to punish, but to prohibit, yes. That you may do. Therefore there is a need here for tradition. But the tradition—and this is what I spoke about at the beginning of the chapter, where he discussed the relation between logic and exegesis—I said that exegesis is not mechanics. It’s clear that an exposition is a trigger, following which logic arrives, or what accompanies it is logic—even in verbal analogy. You say: you find a word here and a word there, right? “Under, under,” like what we brought earlier; or “giving, giving,” what we brought earlier regarding the best-quality payment for damages. After all, it’s obvious: okay, so the same word appears in these two passages—what do you do with that? In what respect do you compare? Does that mean that in all cases of damages everyone has to stand on one leg? In all damages? What—everywhere is it only with oxen, because there too it is written only about an ox? In what respect do I make the comparison from this passage to that passage? In respect to what logic says ought to be compared. Always logic.

[Speaker J] So maybe you should come to the thirteen hermeneutical principles only if you didn’t solve the matter through your regular logical reasoning; then you say: I have no choice, I’ll go to the formal mechanism—which also has to be logical—but that’s because I ran into…

[Rabbi Michael Abraham] No. Where there is a clear plain-sense interpretation, you obviously don’t need the hermeneutical principles. On the contrary, that is the interpretation; that is called the plain sense. Exposition is always something that is not according to the plain sense, even though it too has some logic in it, and I spoke about that—about this two-plane relation between exposition and plain sense. But yes, there is logic in that too; only here the logic indeed requires some trigger, which probably does need tradition. Now, when I discussed the principles, I spoke about logical principles and textual principles. A fortiori reasoning and verbal analogy are the two examples he deals with here. A fortiori reasoning is a logical principle; verbal analogy is a textual principle. The textual principles are a textual trigger for making an exposition, and it also tells me in what direction to make that exposition—generalization, learning from an example. But what exactly to do, why to generalize, or what to learn from the example—there is no hint. That is all the expositor, in the end. “You shall fear the Lord your God”—so the rule is that the word “et” comes to include something. The well-known Ben-Gurion rule that “et” comes to include. Fine. Include what? Maybe chairs. “You shall fear the Lord your God”—to include chairs, that they too should be feared. The verse does not say what to include. The logic of the expositor says… In fact Shimon HaAmsuni got stuck there. Right? He used to expound every “et” in the Torah until he came to “You shall fear the Lord your God.” Why? Because he says: should I compare something here to the Holy One, blessed be He? Impossible. There’s nothing—what could I…? I can’t find anything with my logic. So he almost thought he should abandon the whole method of deriving inclusions from “et.” “Just as I received reward for the exposition, so shall I receive reward for refraining,” and leave it. Fine? Because he found no logic. It’s not a technique. Then Rabbi Akiva came and said: there is something to derive—namely, the disciples of the sages. What is that “something”? It’s not something logical that I would have done without the exposition. Without the exposition I would not have said that if one must fear the Holy One, blessed be He, then one must also fear Torah scholars. Very far-fetched. Rather, the exposition, the rule, the trigger tells me: there is an “et,” you need to include something. First of all, that’s given. Now I can look for what is most logical to include, or what is least illogical, doesn’t matter, to include. And that is what I include. But without having that rule telling me that one must include something from “et,” I would not have derived the obligation to fear Torah scholars, or at least not necessarily. So there is a textual trigger from which one proceeds to exposition, but what one does with the exposition is always logic. And that is true also in the hermeneutical principles. In the textual principles that’s obvious. But even in the logical principles it’s the same; that’s what we’ve just seen with a fortiori reasoning, or with an archetypal rule, which is analogy. Right? A fortiori reasoning is also a kind of analogy, actually. So there too it is clear that some logic is involved. It’s not logical mechanics. Okay. So that’s what—even he basically writes these things. He says that, say, regarding the principle of a fortiori reasoning, let’s read the last paragraph on page 24, the last paragraph dealing with a fortiori reasoning. “There is much to analyze in the principle of a fortiori reasoning. A fortiori reasoning is not simply a plain ‘all the more so’; that is, it is not about two things A and B where the relation between them is that A includes B”—he means the rule “if two hundred includes one hundred.” “And if the law applies in B, then all the more so it applies in A,” like in ‘if two hundred includes one hundred.’ If that were the case, all the give-and-take in the Talmud concerning a fortiori arguments would be incomprehensible—neither how severity creates the a fortiori argument nor how a refutation can undermine it. Because a refutation couldn’t undermine it. Therefore these a fortiori arguments are not just a case of ‘if two hundred includes one hundred.’” What I claim is that even in “if two hundred includes one hundred,” a refutation can refute. Fine, but that’s… “Rather, a fortiori reasoning is a certain way of learning Torah that Moses our teacher received at Sinai.” There is some weight to tradition. Without the tradition, it may be that we wouldn’t do it, or at least not always do it. In other words, you need some anchor that puts a fortiori reasoning into my toolbox, telling me that I am allowed to use it. “And it is that between two matters in the commandments of the Torah, between which one can find a connection, though the similarity of their laws is not necessary, one creates an analogy by analyzing their components. If the severity in one as compared to the other can teach about similarity in the law one wishes to derive, one uses it to create an a fortiori argument.” I think the wording here is not precise. In other words, it would be correct regarding an archetypal rule. In an archetypal rule I make an analogy from one thing to another. There it is obvious that the two things must be similar to each other; there is no hint in the Torah why to make the analogy. Analogy is all from me. In other words, there is no hint here, no textual trigger to make an exposition. Some law is written in the Torah—say, “If one man’s ox gores his neighbor’s ox,” then one must pay. Fine? I say: a dog that bites must also be paid for. How do I know? I make an analogy from ox to dog. Fine? Where is that written? Is there any textual hint whatsoever? No. Therefore it is a logical principle, not a textual principle. But in a fortiori reasoning there are three data points; it does not say that one must make from them an a fortiori argument. There is no textual trigger saying: from these three data points make an a fortiori argument. No—the logic says so. The Torah gives only the data, right? As opposed to the textual principles, where the Torah gives you the trigger for the exposition. There, logic is not the trigger for the exposition; rather, this or that form of writing is the trigger for the exposition. Fine? So in an archetypal rule he is correct that the similarity here is a similarity that we decide upon by our reasoning, that A resembles B. In a fortiori reasoning there is a dimension of reasoning, but still it is not exactly an archetypal rule. A fortiori reasoning is a different principle. Because the relation of severity does not point to similarity. The relation of severity does contain some element of inclusion, let’s call it that—inclusion with a ‘kaf’—it’s not really inclusion; it’s not “the whole from the whole,” but it’s not merely simple similarity between two things. It’s a strong similarity, similarity with direction. Actually, if we want to translate an a fortiori argument into an archetypal rule, into analogy, the right way to do that is not to compare the data at all, but to compare the axes of the hierarchy. Let me show you how analogy actually works, how a fortiori reasoning is basically a simple analogy—nothing more than an analogy. Let’s go back to Tooth and Foot and Horn. In the public domain I know that Horn is more severe than Tooth and Foot, and from this I infer the conclusion that Horn is also more severe than Tooth and Foot in the damaged party’s courtyard. Then I apply that to Tooth and Foot in the damaged party’s courtyard and draw the conclusion. Here I made a straight analogy. The comparison regarding the hierarchical relation is a comparison on the same level; it is not itself an a fortiori move. What am I saying? I’m saying: if the hierarchical relation exists in the public domain, then it exists also in the damaged party’s courtyard. Here the comparison is level-to-level; it’s not an a fortiori argument—that the same hierarchical relation that exists here also exists there.

[Speaker D] And that’s what the refutation attacks.

[Rabbi Michael Abraham] Exactly. No, exactly—and that’s what the refutation attacks. Exactly. When you look at a fortiori reasoning that way, then you see that it is really just analogy. So a fortiori reasoning is really analogy, even though superficially it seems that in an a fortiori argument there is some kind of push, some hierarchy—it’s not just resemblance between two things where neither has priority over the other. But in fact, when you look at the a fortiori argument more deeply, you see that a fortiori reasoning too is analogy.

[Speaker D] The Talmudic a fortiori argument is a kind of “the whole from the whole,” it’s not…

[Rabbi Michael Abraham] Right. That’s why I say: the Talmudic a fortiori argument. The Talmudic a fortiori argument is analogy. Correct. Okay, so that’s why I say: the wording here, in my opinion, is not entirely precise, but the basic principle is correct. “And if not all the components of the two matters point to the similarity, and one finds a refutation that can indicate that they are not similar, then the a fortiori argument is refuted.” What is the refutation? The refutation is exactly that there are two axes of comparison, and now you have to check whether the axis in the damaged party’s courtyard is similar to axis A or to axis B. He is basically saying that there is no similarity between the axes, or no necessity that there be such similarity. “One must examine this in every topic and passage on its own terms, and see the logic of this mode of learning.” And here this connects to what I said before. In other words, a fortiori reasoning is indeed logical. There is no purely formal a fortiori argument. That’s nonsense, simply nonsense; you can’t—it cannot be said. It’s part of that same, I don’t know, Brisker fiction, according to which people think it’s possible—that the Sages were equipped with some mathematical system with which they could derive things absolutely, and we simply lost the algorithm, so to speak. Therefore all we can do is cling to the Sages and see what they did. There is logic here; one can understand this logic. You have to work on it. We don’t understand it completely. Of course part of it has been lost to us because the tradition got stuck on this matter. But there is logic behind it. Okay, so that’s it regarding a fortiori reasoning. Now he moves on to verbal analogy. As I said, a fortiori reasoning is a logical principle; verbal analogy is a textual principle, okay? But as I said earlier, clearly there is a logical dimension both here and there. Even in verbal analogy there is a logical dimension on two planes. First, in the question of what exactly I compare—not everything do I transfer from one to the other. So with respect to which laws do I make the comparison? Even though the rule is that verbal analogy is not applied by halves, just as juxtaposition is not applied by halves, it is still clear that “not by halves” means not for all things indiscriminately, but for all things that are relevant. If something does not seem similar to you, you do not force it to be similar. If it is clear to you that here there will be a difference, then there is a verbal analogy of “these, these” between a slave and a woman—so what, now it means you can… no one among the Sages says such a thing. Why not? “Verbal analogy is not applied by halves.” If a slave and a woman were compared—and in general, according to that, you’d also have to write “woman” with an ayin, because “slave” is also written with an ayin. So according to that you’d have to correct all the Torah scrolls. Obviously we do not apply verbal analogy to things that don’t sound reasonable to us, things that don’t appear similar in the two contexts. So first of all there is a logical dimension in the question of what we do. Once there is a trigger to compare, what do we do with it? But there is another dimension, and it’s no less important—and he’ll talk about this too—the dimension of choosing which two words we make into a verbal analogy. Before the question of what to do with the verbal analogy, there is the very decision to make one. Not every two similar words in the Torah are a trigger for verbal analogy, and that’s an interesting question why. From what I once mentioned—I don’t remember—there is a book by a man named Chernik, a book on verbal analogy. It was published in Lod, by the institute there—Abrams, I think, I don’t remember—in the library there. So he published a book on verbal analogy, and there he argues that the early verbal analogies are only between words that are unique in the Bible. He does some kind of archaeology of periods; he’s an academic scholar, so he does some kind of historical archaeology of periods. In other words: which verbal analogies are early, which are later, he identifies this by, say, the sages involved, their generation, and so on. He does a historical analysis of the matter and classifies verbal analogies by periods. And he shows that the early verbal analogies are all between two words that appear only in those two places in the Bible. Not every two similar words we happen to find immediately become a verbal analogy. But interestingly, this develops later, and verbal analogies are made even with two words that are not unique. So how can that be? What—if really we received that only… if in the earlier period they did it only with unique words, then what are we inventing? I return to everything I said before: are the hermeneutical principles an invention or a development? I think it’s a development. And the secret here is that when you take two such words, apparently—I’m saying this tentatively; I don’t have an explanation for every pair of words, and I also haven’t worked on it enough to check seriously—but there are places where I did see this. There has to be a reason why I choose these two words and make a verbal analogy out of them. True, they may appear elsewhere in the Torah. In the later verbal analogies this was done also with words that appear elsewhere in the Torah, but even so, in these two places there is a reason why those words were chosen. It’s not just because here there are two similar words, period. There is a kind of interpretive understanding of where this is appropriate. Now, part of this is explicit in the Talmud itself: the concept of “available” wording. Available from both sides, available from one side—the Talmud distinguishes between different kinds of verbal analogies. “Available” means the word is superfluous. So what does that really mean? That in the early verbal analogies they took only two words that were unique. Fine—but that itself calls for interpretation: why did they use precisely these two words in precisely these two contexts? Apparently because the verses came to compare. But later sages say: fine, I can apply the same logic also to a word that appears elsewhere in the Bible, provided that in these two places it is “available,” it is superfluous. That too calls for interpretation. What difference does it make? It’s the same idea. So in effect I am expanding the logic of verbal analogy, but I’m expanding it on the basis of the original logic; I’m not inventing. It’s not an invention. It’s the claim that I can apply the same logic to further cases. And therefore what this really means is that there has to be some reason in a verbal analogy why these words were chosen, especially when the words are not unique. If they are unique, perhaps even that isn’t needed, because that in itself is a sufficient reason. But if the words are not unique, there has to be some reason. Now the Talmud itself says this. It says that one makes a verbal analogy only when the words are “available.” If they are available from one side, then one does it—but one can still refute it; if there is a refutation, we reject the verbal analogy when it is available only from one side. But if it is available from both sides, then there are no refutations either; then we make the verbal analogy and that’s it—there is no refutation against it. But already there you see that the availability, the extra word, is really the trigger for the exposition. It’s not just that you take two words, some mechanical thing. But then—one moment—how do you know whether it is available or not? How do you know whether the word is superfluous or not? Logic. You look and see whether it adds something, or whether you can propose a wording that omits that word and still tells you everything the verse says. That’s interpretive logic. Now there can be a dispute about whether that word is available or not, and there are disputes about whether a certain word is available. Someone who derives something from that word, from his perspective it’s not available. There are such Talmudic passages. So what does he do with—how does he make a verbal analogy if it’s not available? Meaning, you see that the logic of the expositor, the expositor’s overall picture, affects the question of how you make a verbal analogy. Now this is especially interesting with verbal analogy, because verbal analogy is the principle about which the Talmud says that a person may not derive it on his own unless he received it from his teacher. A person does not derive it on his own. An a fortiori argument—a person derives on his own. A verbal analogy—a person does not derive on his own unless he received it from his teacher. There is a dispute between Rashi and Tosafot about the other eleven principles: are they like a fortiori reasoning or like verbal analogy? Usually people assume they are like a fortiori reasoning—that all the principles we do on our own except verbal analogy. Verbal analogy is unique, but that is a dispute between Rashi and Tosafot. In any case, even verbal analogy, about which it is certainly said that a person does not do it on his own, we still see that logic is involved in it. And there are disputes, and Nachmanides in fact comments on this—one moment—Nachmanides indeed comments on it: because there are disputes about verbal analogy, Nachmanides and his students, and several medieval authorities from his school, all write that it is not true that we received everything from Sinai. Rather, we received some hint from Sinai that here one should make a verbal analogy, or that we received the result and the exposition merely leans on it, or that we received that these words are for verbal analogy but were not told what exactly to do with them. There is some hint that is supposed to help us, and that connects with what I said earlier. Because really, without that, what would we do—just compare any two words in the Torah? From that we could make a whole mess. The Sages clearly never came close to doing that. So by what did they decide? So the Talmud says they received by tradition what should be done—but what does “by tradition” mean? It means they understood that a certain word was extra and designated for verbal analogy. That fits the picture I described earlier. There is no need to say that there really was an explicit tradition: from these two words make a verbal analogy. Rather, they received a tradition of how to analyze the text and how to discover whether this word is superfluous or not superfluous. That is the skill transmitted by tradition. It doesn’t mean exact details, but that they were skilled in analyzing the text to see whether these words are superfluous, available, or not available, and accordingly they made verbal analogies. That’s all.

[Speaker I] Not always—not always is it necessarily true that the Talmud calls it “available,” but maybe that’s even very late. Just in general you also see that even if the words are not superfluous syntactically, it may still be that a certain phrasing sounds similar to what we saw elsewhere. Right. But that doesn’t necessarily mean the word is “available” syntactically. One hundred percent.

[Rabbi Michael Abraham] That’s exactly what I said earlier: “available” is not really “available.” You need some trigger in the text that tells you this word is meant for verbal analogy. It isn’t enough that there are two identical words. “Availability” is only an expression that describes this. If it’s superfluous, that’s a simple trigger. And of course, if there is some understanding that these two passages are similar and you also see a key word that is similar in both, then that’s the direction.

[Speaker G] That’s the direction. I think that in a great many Talmudic passages, when you see the verbal analogy and afterward they try to convince you that the word is “available” or not “available,” and they propose all sorts of texts, it looks like it totally doesn’t work. One hundred percent true. It looks as though there was some reasoning, they were looking for something—here there was some missing logic, some kind of test case that had to be dealt with, and the closest case was a slave. Right. So they said, fine, let’s start with a slave, and now they had to force the text so there’d be a word here and a word there.

[Rabbi Michael Abraham] But “force” I’m not so sure.

[Speaker G] Why? Because when you read the sentence sometimes, the strange wording of the verse…

[Rabbi Michael Abraham] Although one has to be careful—we once talked about the difference between the holy tongue and our language. We are prone to mistakes here. Precisely those who don’t know Hebrew are better at this. Those who do know Hebrew sometimes see the formulations of the Sages and it looks unreasonable. But I think that sometimes, when you think in biblical terms, it can sound reasonable. We’re just captive to…

[Speaker G] They drop the “heh” or want it without the “vav.”

[Speaker I] Is it “hen” without a yod? “Hen” without a yod?

[Rabbi Michael Abraham] Yes, whatever it may be for us. No, no—again. I didn’t say that in every place it’s like that. I said one has to be careful. By the way, even there, with “hen” without a yod, I don’t know whether “mehen” really is “hen” without a yod.

[Speaker I] The only word the Talmud found is “from them, his yevamah,” which is something else entirely.

[Rabbi Michael Abraham] All right, but it could be that in biblical Hebrew, the way those people understood it, that really is what it means. Why not? It’s a unique example, fine. It doesn’t matter right now, but I’m saying you have to be careful about that. I don’t know whether that will always give you the solution, but you have to be careful about it. I don’t think that if the Sages forced it, then it was just arbitrary—so why did they force it? For no reason? Obviously they had some reason why they decided to derive it from here. Because of the logic. Exactly. I definitely accept that broader understanding, and he’ll write that too: that it doesn’t have to be specifically a free expression in the simple sense of the term. On the contrary, a free expression is only one example of the fact that you’re not just taking two words and doing mechanics; the text has to signal to you that some comparison is supposed to be made here. Sometimes it signals that by some extra word, sometimes you understand that the passages are similar, sometimes your own logic tells you that the passages should be compared because the legal logic says they are similar. For all kinds of reasons, it doesn’t matter right now. So that, I think, is the broader meaning of the concept of a free expression. I completely accept that. That’s exactly what I wrote. Now let’s read it in his words, because overall this is more or less what he said—what I’ve been describing until now. A gezerah shavah is also only between two similar subjects. That itself already says something. If it were just mechanics—take two words and that’s it—no. You have to understand whether it’s really similar or not. If it’s not similar, then no. It’s not just logical mechanics. Usually, for each topic the Torah uses a different style, and if there is similarity in language, that teaches similarity in content. When the subjects are close and can be compared to one another. Between subjects that are different and far from one another, you never find a gezerah shavah. Nor can every two identical words serve for a gezerah shavah. If it says “with” here and “with” there, we won’t derive a gezerah shavah from that. Between connective words there is never a gezerah shavah, because that would turn the entire Torah into one single law. By the way, I’m not entirely sure about that; I haven’t checked it. But in places where there’s some indication that a connective word is extra, it could be that they would make a gezerah shavah even from a connective word. I don’t think—I mean, there has to be some logic or some hint that we’re supposed to make a gezerah shavah. I’m not sure that with connective words it can’t happen. Clearly, to do it automatically on connective words is madness. That would mean that all the “withs” in the Torah have to be the same thing. Obviously not. Okay, fine, but I don’t think we need to go so far as to say there are no gezerot shavot on connective words. That’s too far; it needs checking. I haven’t checked it, but I don’t think it has to be that way. Because that would turn the entire Torah into one single law. A gezerah shavah exists only where we notice that the Torah uses the same style in two places, in a way that can indicate a comparison between them. According to one view in the Talmud—see Niddah 22—a gezerah shavah is expounded only when the expression is free, and that draws attention to it. According to another view, when it is not free, we do expound it but can challenge it—that is, if there is a refutation showing that not all the components of the matters are similar, then we do not learn one from the other. And again, this is a question: if it can be refuted, then where is the tradition? A person does not expound a gezerah shavah unless he received it from his rabbi. So how can you refute it? How can one refute a gezerah shavah? We received it by tradition from Sinai, didn’t we? Well, you could say that only when it is free is it tradition, because then it cannot be refuted. But if it is free, then why do you need tradition? It’s free, so you know it is to be expounded. On the contrary, the tradition is required specifically in a case where…

[Speaker E] Even in a tradition there can be cracks—that someone missed it, or made a mistake, or things like that—so if it’s refuted, apparently the tradition wasn’t correct.

[Rabbi Michael Abraham] Fine, you can say that, but then what is the point of tradition? If the tradition itself is suspected of not being correct, then what? So it’s just one more support, one more support—both logical and… okay. And there is a distinction between being free on one side and being free on both sides. Even when we expound a gezerah shavah, we do not compare for every purpose. Exactly what I said: these are two levels. One level is choosing the pair of words on which I make the gezerah shavah. I chose the two words; now the question is what to do with the gezerah shavah. With regard to what matter do I compare these two things? So there too, logic comes in. Meaning, it comes in on both levels. True, there is a rule that there is no gezerah shavah by halves, but Tosafot prove in several places that a gezerah shavah does not compare for every purpose. Because the comparison has to be only regarding things in which it is possible and appropriate to compare the two subjects. And in general, with a gezerah shavah this is always a problem, because you want to compare two subjects, and yet you always say that at the starting point they are different from each other—so why compare them? Now if they are different, the question is which one do you take toward which? Do you take the one that is liable over to the one that is exempt and learn that the latter is liable? Or do you take the one that is exempt over to the one that is liable and learn that the latter is exempt?

[Speaker G] That’s how the Talmud asks it. Why did you choose to learn it this way—why not learn it the other way around?

[Rabbi Michael Abraham] I don’t remember it asking that; maybe there are particular cases, but generally it doesn’t ask that. There are places where this is also a dispute between Rashi and Tosafot—or Rashbam and Tosafot, I think—about what “there is no gezerah shavah by halves” means. Rashbam claims, I think, if I remember correctly who is who—Rashbam or Tosafot—one of them says that “there is no gezerah shavah by halves” means that just as you learn from A to B, so you must learn from B to A. And the other says that “there is no gezerah shavah by halves” means you take all of A over to B, not just part of A. Now usually this dilemma doesn’t really exist, because when in one case there is liability and in the other exemption, the exempt case is not one where I have a verse exempting it; rather, I have no verse obligating it. So basically the law is that it is exempt. True, if I don’t fill it in, it will be exempt. But if I have a gezerah shavah from elsewhere where I have a source saying there is liability, then clearly the direction is to take the liability into the place of exemption, not the exemption into the place of liability. If I have a special source exempting it there, then indeed the situation will be the opposite. All right? If there is a source exempting here and a source obligating there, I simply won’t apply the gezerah shavah. All right? So therefore, usually there is no dilemma about which direction to take the gezerah shavah. The direction is always from the known case to the exempt case. And this is the… It is well known as a rule that a person does not derive a gezerah shavah on his own. Does that mean that gezerot shavot were handed over to Moses at Sinai? Absolutely not. There are many disputes about gezerot shavot; this is Nachmanides’ argument that I mentioned. And if they had been received from the Holy One, blessed be He, or had been received from Moses our teacher who learned them, there would not have been disputes about them. At the moment I accept Yossi’s comment from before: this argument is very questionable. The fact that something is a law given to Moses at Sinai does not mean there cannot be disputes about it. True, Maimonides writes… and he explains—some of them he tries to explain, and still in another part he gets stuck on Maimonides. This determination of Maimonides is problematic.

[Speaker G] He didn’t mean—Maimonides didn’t mean—that no dispute ever arose, as if regarding what is accepted as a tradition, about that they don’t argue. That’s not what he meant. There can be a dispute—no?—if there was some corruption along the way, then you may find a dispute about something that someone says is a law given to Moses at Sinai.

[Rabbi Michael Abraham] But if there was no corruption and everyone agrees, then of course there’s no dispute.

[Speaker G] So what did he say? No—but there also can’t be a dispute, no? It’s not that someone can now come and say, listen, I don’t accept this law.

[Rabbi Michael Abraham] No, that’s an empirical claim, not a normative one. According to your understanding, what he said here is a normative statement—that one is forbidden to dissent from a law given to Moses at Sinai. Maimonides presents it as a factual claim: that with a law given to Moses at Sinai, no dispute ever arose. That’s the claim. The Talmud does not say that one must receive the gezerah shavah from Sinai; it says only that a person does not derive a gezerah shavah on his own. And Rashi adds: unless it was received and came from Sinai. So in Rashi, yes, he always says that. Rashi also says “he learned it as a tradition”; Rashi always adds that it is a law given to Moses at Sinai. For Rashi, “he learned it as a tradition” means a law given to Moses at Sinai. The meaning is that the question of when equal style really comes to teach a comparison of the subjects and when it does not is not so simple. Therefore, if a sage does not have a tradition of many generations in which everyone learned this way—that such-and-such an expression appearing identically in two places indicates a comparison between them—he should not innovate such a derivation on his own. You have to have the textual sensitivity to know whether these really are two words for which it is proper to make a gezerah shavah, or not. Of course this is not Rashi’s plain meaning. Rashi’s plain meaning is that it came from Sinai. What he wants to say is that there has to be a tradition established by authorized Sages and so on. There is something like this in the introduction to Ha‘amek; the Netziv writes in the introduction to Ha‘amek a question, and there he writes that in the Talmud, regarding “he learned it as a tradition,” Rashi always says it means a law given to Moses at Sinai. In Maimonides it sounds as though “he learned it as a tradition” means some kind of ancient tradition, not necessarily a law given to Moses at Sinai. And so he indeed explains that there is a dispute between Rashi and Maimonides about what “he learned it as a tradition” means. Something similar is what he wants to say here about a law given to Moses at Sinai itself—not about “he learned it as a tradition.” There the discussion is whether “he learned it as a tradition” means a law given to Moses at Sinai. He claims that even “a law given to Moses at Sinai”—Rabbi Gedaliah claims that even a law given to Moses at Sinai is not literally a law given to Moses at Sinai, but rather only a tradition accepted from outstanding Sages. He means this regarding gezerah shavah, not every law given to Moses at Sinai. And when Rashi says that it has to come from Sinai, the meaning is: it has to come from an ancient tradition from authorized Sages, and so on. For example, the Talmud says regarding the twenty-four primary categories of damages: all are like primary categories in that they pay from the best quality. What is the reason? It is derived from “in place of,” “giving,” “shall pay,” “money.” Can we be sure that the expressions “in place of,” “giving,” and so on all indicate a comparison specifically in the manner of payment? That it should be from the best quality specifically? That it should be from the best quality, yes. In damage payments one must pay from “the best of his field and the best of his vineyard,” and movable property too is called best quality. Never mind—but not every kind of payment is acceptable. It has to be quality payment, readily marketable. So why decide that this comparison is specifically about the manner of payment? Maybe the comparison is about completely different things. It is not so simple. We would not rely on ourselves in such a matter. But if many generations of great Sages, who knew the ways of language, held this way—not necessarily from Moses at Sinai—we can rely on them. And if we look carefully, we will understand that the matter is correct: that it really does make sense that these expressions indicate the manner of payment. Therefore this gezerah shavah is indeed applied to the manner of payment. And that is what the logic says. And if we have lost that logic, then we need to rely on Sages who still had that linguistic sensitivity to biblical Hebrew, who knew how to identify these two words and expound them. And I think it would be very worthwhile—this is one of the conclusions that emerges from his approach, and I completely agree with him on this point—that it would be proper to renew a bit of that linguistic sensitivity and try to see how the Sages actually worked, and see whether it’s possible to create additional gezerot shavot ourselves. Okay, we’ll stop here.