Midrash and Rules of Interpretation – Lesson 3

Table of Contents

• Midrash and peshat: creative derashot and supportive derashot
• Dispute in derashot, law given to Moses at Sinai, and “it was learned as a tradition”
• A scriptural decree as reasoning strengthened by a verse
• “Punishments are not derived through logic alone”: source and reasoning together
• Asmakhta versus supportive derashah and gezerah shavah derashot
• The nature of the interpretive principles: arbitrary code or language
• The language-learning model: Moshe Koppel, Chomsky, and machine learning
• The forgetting during the mourning for Moses, Othniel ben Kenaz, and conceptualizing the rules
• Dynamic law given to Moses at Sinai and the development of the interpretive principles
• Pesachim 22b: Shimon HaAmsuni and Rabbi Akiva as a model of trial and error
• Popper, Thomas Kuhn, the Turing test, and analogies of language and law
• Decline of the generations, intuition, and the multiplication of rules

Summary

General Overview

The text presents the difficulty of seeing derashot as a source that actually creates Jewish law, because they often feel like a kind of “game” and are not convincing as binding proofs. It offers a solution according to which there are creative derashot and supportive derashot, as Maimonides writes. It explains that a supportive derashah is not an asmakhta, but a serious derashah for a Torah-level law known through tradition, and that it “takes hold of a verse” and prevents any additional law from being derived from it. The text then presents an approach according to which the interpretive principles are not an arbitrary code or a closed mathematical system, but a language parallel to peshat, transmitted at Sinai in a living and dynamic way. The formulated rules, by contrast, are a human product of reconstruction and conceptualization over the generations, which is why they develop and become subject to dispute. Finally, the text illustrates this through Talmudic passages and through the Talmud in Pesachim about Shimon HaAmsuni and Rabbi Akiva, tying the discussion to science, grammar, machine learning, the Turing test, “telephone,” and the question of binding authority when there is no absolute certainty.

Midrash and peshat: creative derashot and supportive derashot

The text assumes that the question whether derashot create law or merely support it determines whether there is any point to the interpretive effort when the law is already known, because many derashot are not persuasive as a binding source. Maimonides writes explicitly that there are creative derashot and supportive derashot, and this distinction solves the difficulty because the supportive ones are not mere “play.” A supportive derashah takes hold of the verse and thereby prevents it from being used to derive another law, and the Talmud works with the assumption that a verse is “already occupied by interpretation” and that two things are not derived from the same verse. The text presents two types of supportive derashot: anchoring a law given to Moses at Sinai in a verse, and reconstructing an interpretive source for a law transmitted by tradition when people no longer know from which derashah it originally came.

Dispute in derashot, law given to Moses at Sinai, and “it was learned as a tradition”

Maimonides is quoted as saying that no dispute ever arose regarding a law given to Moses at Sinai, and the text explains that disputes over derashot can arise because the law did not come down as Sinaitic law but rather as a law originally generated by derashah whose source was later lost. The Netziv, in his introduction to the She’iltot of Rav Achai Gaon, Ha’amek She’elah, distinguishes between a tradition from Sinai and a tradition of a law that arose over the generations and whose source was lost, and on that basis he interprets the Talmudic phrase “it was learned as a tradition” as indicating a binding tradition that is not necessarily a law given to Moses at Sinai, unlike Rashi. The text emphasizes that not knowing the original derashah creates room for different reconstructions and therefore for dispute, unlike the assumed certainty of a law given to Moses at Sinai.

A scriptural decree as reasoning strengthened by a verse

The text argues that a scriptural decree is not necessarily a law “without reason,” because reasons and explanations are found even for such laws. It explains that the verse does not make reasoning unnecessary, but gives it binding force when, without the verse, the reasoning would not have been strong enough to establish a law. Meiri, regarding the stubborn and rebellious son, explains “a son and not a daughter” on the grounds that “it is not the way of a daughter to rob people,” and the text sharpens the point that without the verse one would not necessarily have excluded daughters, or excluded other cases, on the basis of similar reasoning. The text formulates the definition that a scriptural decree means that without the verse we would not have done this on our own, even though once the verse exists, the logic can be understood.

“Punishments are not derived through logic alone”: source and reasoning together

Rabbi Yosef Engel is cited through the Talmudic Encyclopedia as offering three “approaches” to explaining the rule that punishments are not derived through logic alone, and the text argues that these are not three competing approaches but one source together with two lines of reasoning that explain it. The text brings two reasons: concern for a refutation of the a fortiori argument, meaning that it is not certain enough to justify punishment, and the distinction that the punishment for the lighter case does not necessarily fit the more severe case, because perhaps a different punishment is required. It cites scriptural sources such as “his sister, the daughter of his father or the daughter of his mother” and “if a man opens a pit or if a man digs a pit,” and explains that the source does not negate the existence of reasoning but joins it as binding justification. The text ties this to the broader claim that what comes from a verse is not devoid of reasoning, and therefore a scriptural source should not be seen as competing with reason but as a foundation accompanied by explanations.

Asmakhta versus supportive derashah and gezerah shavah derashot

The text distinguishes between an asmakhta, where a rabbinic law is attached to a verse for memory or support, and a supportive derashah, which is a full-fledged derashah for a Torah-level law that was known before the derashah. It cites Nachmanides and his students among the medieval Spanish authorities (Rishonim) concerning the rule that “a person may not derive a gezerah shavah on his own,” and raises the problem of how disputes in gezerah shavah can exist if everything was given at Sinai. Nachmanides answers that gezerah shavah does require some anchor from Sinai, but not necessarily in the sense that the whole content is a completely pre-packaged “fixed game” in which all the details were transmitted; sometimes what was transmitted from Sinai was a connection or hint without full specification, or a law was transmitted and the interpreter finds the fitting gezerah shavah. The text concludes that the interpreter is not absent from the process; there is a combination of tradition and interpretive creativity.

The nature of the interpretive principles: arbitrary code or language

The text cites Ralbag in the introduction to his commentary on the Torah as holding that everything is supportive, in order to avoid a situation in which “everyone can do whatever he wants,” and rejects that view by arguing that the Talmud is full of evidence that things did not work that way. It cites Maimonides in Millot HaHigayon as saying that the interpretive principles are not rules of logic, and raises the possibility of seeing them as an arbitrary code like atbash, but explains that if it were an arbitrary code it would have to be mathematically precise and unambiguous, with no disputes, contrary to reality. The central claim is that the interpretive principles are a parallel language with its own internal “grammar,” one that does not replace peshat but operates alongside it under the rule that “a verse never loses its plain meaning.” The text presents the midrashic language as a system that can be used intuitively and can generate disputes, like a natural language, and not as an algorithmic machine.

The language-learning model: Moshe Koppel, Chomsky, and machine learning

The text attributes to Moshe Koppel the model according to which at Sinai Moses our teacher learned two “languages” for reading the Torah: the peshat reading and the derash reading. The example given is “You shall fear the Lord your God,” where in the language of derash the word “et” comes to include something additional. It compares language learning in two ways: studying in an ulpan through explicit rules, versus the natural way children learn from use, and cites Noam Chomsky as arguing that there are innate patterns that make language acquisition possible from speech and not only from rules. The text parallels inference from rules to the method of machine learning, where examples are given and patterns are extracted from them that allow generalization to cases not previously studied. It concludes that in learning derash at Sinai, what was transmitted was not a formal list of rules like the thirteen principles, but the language itself through reading verses on both the peshat and derash levels.

The forgetting during the mourning for Moses, Othniel ben Kenaz, and conceptualizing the rules

The text quotes the Talmud in Temurah that during the mourning for Moses, “three thousand gezerah shavah inferences, a fortiori arguments, and seventeen hundred laws given to Moses at Sinai” were forgotten, and that Othniel ben Kenaz restored them through his dialectical analysis. It distinguishes between a law given to Moses at Sinai as closed information, which cannot be reconstructed if forgotten, and a language, which can be reconstructed by means of examples that remain and from which rules can be distilled. It brings an anecdote from the commentators on the Jerusalem Talmud, Pnei Moshe and Korban HaEdah, according to which a sin offering whose owners died is treated as rabbinic because the original Torah-level law was forgotten. It explains that Othniel ben Kenaz “creates an ulpan” by formulating rules in order to replace natural intuition that had weakened, so that the rules are a human product born of necessity.

Dynamic law given to Moses at Sinai and the development of the interpretive principles

The text presents a dispute between the medieval authorities (Rishonim), who say that the halakhic interpretive principles are themselves a law given to Moses at Sinai, and scholars, who argue that the system of principles developed historically from Hillel the Elder’s seven principles, to Rabbi Ishmael’s thirteen, to Rabbi Eliezer son of Rabbi Yose HaGelili’s thirty-two, and beyond. It argues that both sides are right: what was transmitted at Sinai was the midrashic language itself, while the conceptualization and formalization of the rules are human work and therefore develop and generate disputes. It explains that the farther one gets from Mount Sinai, the more one loses the “native speaker’s intuition,” and therefore the more one needs rules and sub-rules, like the thickening of models in Ptolemaic cosmology with epicycles until the shift to ellipses. It stresses that the disputes between Rabbi Akiva and Rabbi Ishmael are disputes about how to conceptualize the grammar of derash, not necessarily about the very existence of the language.

Pesachim 22b: Shimon HaAmsuni and Rabbi Akiva as a model of trial and error

The text quotes the passage in Pesachim 22b about Shimon HaAmsuni, who interpreted every instance of “et” in the Torah until he reached “You shall fear the Lord your God,” and then stopped. His students asked, “What will become of all the instances of ‘et’ that you interpreted?” and he answered, “Just as I received reward for the interpretation, so too I receive reward for withdrawing from it.” The text explains that Shimon HaAmsuni saw this verse as a case that did not allow any additional inclusion alongside fear of God, and therefore gave up the entire rule. From this the text concludes that the rule “et comes to include” is not a closed tradition but a general proposal tested against examples. Rabbi Akiva “interpreted: ‘You shall fear the Lord your God’—to include Torah scholars,” and is presented as someone who prefers to save a successful rule by means of the “least implausible” solution rather than abandon it, and the text describes this as an approach of preserving a strong theory in the face of exceptions. The text adds that the Talmud understands that Shimon HaAmsuni did not accept Rabbi Akiva’s solution, and that a tannaitic dispute remained over whether “et” is interpreted at all.

Popper, Thomas Kuhn, the Turing test, and analogies of language and law

The text uses criticism of Popper to explain that a single counterexample does not necessarily overturn a theory if the theory works in many cases, and attributes this direction to Thomas Kuhn and the idea of paradigms and a critical threshold of problems. It mentions the Turing test and explains that computers could not speak a living language through an algorithmic programming method of endless “ifs,” but can get closer to it through non-formal learning, similar to the way the brain learns. It brings the analogy of “telephone” to say that a tradition is not transmitted through a “hollow pipe” without changes, and cites Michael Polanyi on apprenticeship and professional knowledge that cannot be reduced to written rules. It recounts a meeting of rabbis and jurists and mentions Professor Englard the positivist as claiming that everything is professional and deductive, and formulates a position according to which there is discipline and there is “right and wrong” even without absolute mathematics, but the craft cannot be isolated from all human context and from a felt sense of use.

Decline of the generations, intuition, and the multiplication of rules

The text argues that there is not necessarily a decline of the generations in intelligence, but there is a decline in skill and in the natural sense of “use” in the midrashic language. Therefore later generations produce more rules, but are not necessarily closer to the practical truth of the original source. It cites a yeshiva saying that “if the Rosh says ‘it seems to me,’ he is probably right,” as an expression of trust in the intuition of someone close to the source, as opposed to someone relying on proofs and ulpan-style rules. It explains that the number of interpretive principles grows not because new content is being added, but because the conceptualizations become more detailed, and it illustrates this with the category of general and particular: with Hillel the Elder it is one principle, and with Rabbi Ishmael it becomes three principles that are really a more detailed account of the same phenomenon. It adds that in the days of Moses, and in Temurah, mainly a fortiori reasoning and gezerah shavah are mentioned, presenting them as the foundation of logical interpretive principles, as opposed to textual principles that continue to develop. The text concludes that when absolute accessible truth or full natural intuition is unavailable, binding authority rests on “a judge has only what his eyes can see,” on the existing tradition, and on the rules as the best approximation available, even at the price of possible human error, because “the Torah was not given to ministering angels.”

Okay, let’s begin. Up to this point we’ve been talking about the meaning of derash as against peshat, and what the relation is between them. We talked about the question—at least a bit—the question of whether derashot are creative or merely supportive. I actually spoke about this explicitly at the end of the previous lecture: are derashot creative, or are they supportive? What’s the point of finding a supportive derash if I already know the law beforehand? So why play this whole game? Especially since the derashot themselves aren’t all that convincing. And if you want to show me that the accepted law is also correct, and for that you want to use a derash, it’s not really persuasive. So what’s the point? I said that if I understand that all derashot are supportive, then this is indeed a hard question—hard to know how to answer. So why are there still people who claim that all derashot are supportive? Because saying that derashot are creative is problematic, since very often it looks like some kind of game that you can’t really take seriously. So if the Jewish law is already known and you’re just playing around, okay, fine, not terrible—the law is correct. But on the other hand, then you say: okay, so what’s the whole story for? Then don’t do it. And therefore I said that the way out that can answer both questions is this: there are creative derashot and there are supportive derashot. Maimonides writes this explicitly: there are creative derashot and supportive derashot. And then what happens is that there is also a point to doing the supportive derashot, because once you found—once you supported one of the laws you received with a derash, that derash has already occupied the verse for me. So now I know that if I want to derive a new law from that verse, I can’t anymore, because it’s already taken. Therefore it’s very important to do the supportive derashot too, not only the creative derashot, because that also has to be taken into account when I come to make a creative derash. I can know that this verse is already unavailable—you can’t rely on it. That means that in effect you think your understanding takes up space with respect to everything else, and it isn’t tied to that. The Talmud talks about this. The Talmud brings a derash from a verse. The Talmud immediately asks: and what does the other one do with this? The assumption is that there is something in the verse saying, “Interpret me.” Now you have to interpret it. So if I already interpreted it, then nothing more can be interpreted from here. Someone else who doesn’t interpret it that way is asked: so what do you do with it? Then afterward, after he says what he does with it, they ask me: and from where do you derive his law? What’s the problem? From the same verse. No, no—because from this verse I already derived something, so I can’t derive something else. So you see that when we made some derash, it already occupied the verse for me, so you can’t use this verse for another purpose. We don’t derive two things from the same verse. So that means that this law could have been received from Sinai, as it were—a question I need to… and I’ll still speak about this more—but there are two kinds of supportive derashot. There is a supportive derash where the law is in fact a law given to Moses at Sinai, and I anchor it in a derash. There is another situation where a law reaches me through tradition, but they don’t tell me that it is a law given to Moses at Sinai. In that case it may be that in some previous generation they actually derived it from a derash, only what reached me was the law; I don’t know from which derash it came. So I look for where it could have come from, and I find a derash. What’s the difference? For example, Maimonides says that with things that are laws given to Moses at Sinai, no dispute ever arose. Now if I see that a dispute did arise about derashot, that could be because the derashot are supportive, say; it could also be because the source of the law is not from Sinai, but rather it came through a derash of the Sages—only that has been lost to us and we are trying to reconstruct it. Then one person may suggest it this way, another person another way, and a dispute is created. A law given to Moses at Sinai is something we receive, and regarding that there is no dispute. It doesn’t stand up all that well under scrutiny, but that’s what Maimonides writes.

Even though the one who invents the derash is not all that convincing. No, so I’m saying: if the derash is a creative derash, it has to be convincing. I’m talking about supportive ones. Then you can begin to hesitate, because it can’t be just a game. If it’s just a game, then it’s worthless. But if you already know the law, that can help me accept this derash—that maybe to produce a law from it I wouldn’t do that, because it’s not strong enough. I would do it, but there’s a paradox: I know that those who came before me did create with this derash. Right, because I assume they did create with it, only I can’t find it. I can’t create; I only need to… No, I can create; I’m just saying that maybe it isn’t strong enough in my eyes to rely on for creating a new law. Apparently they understood more than I do, or I don’t know what—earlier generations are earlier generations. So apparently that’s the derash they made. And from my point of view, that really is the derash. Even though I myself perhaps wouldn’t have… It’s like a Mishnah: the Talmud discusses where the law in the Mishnah comes from. Disputes begin: it comes from here, it comes from there, yes, no. And how did the tannaim say it and not disagree at all and none of that? Because for them it was clear. So why not for us? Because we don’t understand well enough how to interpret the way the tannaim did, and therefore there are disputes. But after I reach the conclusion that this is the best reconstruction I can find of what the tannaim did, then my assumption is that this really came from there, even though I myself am not sure I would have derived it from there. Then I say: fine, if it came from there, then I can’t derive another derash anymore, because it’s already taken. In other words, it can’t be a total game—not even a supportive derash. It has to be something that I do understand to be correct, perhaps just not strong enough for me to rely on in order to create a new law.

Something like this we discussed in another lecture—I already don’t remember which lecture—about what a decree of Scripture is. People think a decree of Scripture is something that has no reason. It comes out of the verse and has no reason. That’s not true. In many places we see that even for a decree of Scripture we find reasons. We explain why the decree of Scripture says this and not otherwise. Then people ask: if you have an explanation, a logical rationale, then why do you need the decree of Scripture? And why is it called a decree of Scripture? Why do you need the verse at all—would I have derived it from my own reasoning? The point is that very often my reasoning does exist, but without the verse it would not have been strong enough to build the law on it. Once the verse is there, the verse says: look, this reasoning is good enough, and you can build the law on it. Let me give you an example. Meiri, regarding the stubborn and rebellious son. The Talmud says: a stubborn and rebellious son—“a son,” not a daughter. And the Jerusalem Talmud says about this: “a decree of Scripture.” And in the Babylonian Talmud too it seems to be a decree of Scripture—less explicitly, but the medieval authorities bring that it is a decree of Scripture, okay? Now Meiri says: a son and not a daughter, because it is not the way of a daughter to rob people. A stubborn and rebellious son is judged that way because in the end he will stand at the crossroads and rob people, and it is not the way of a daughter to rob people. This is Meiri himself there—he has a reflective look, yes, he looks at himself, at what he said, and he asks: wait a second, but the verse—this is a decree of Scripture. So if I’m explaining it, then in what sense is it a decree of Scripture? So I think what he says there is a bit vague, his answer, and it seems to me that what he’s saying there is exactly this. He says: suppose there had not been a derivation from the verse, and I would say, look, there is a law of the stubborn and rebellious son, and the reason is that the way of the stubborn and rebellious son is that he will come to rob people; but daughters, by reasoning, less often come to that. Okay? Would I really exclude daughters? Not so simple. And maybe we should also exclude just some decent kid who isn’t likely to rob people in the end? Or someone who hasn’t dropped out of the educational system won’t have the law of the stubborn and rebellious son, because we assume he’ll be civilized in the end? Okay? But we don’t do such a thing, even though the reasoning is true reasoning. Why don’t we? Because there are lots of true rationales; that isn’t enough to build a law on. Right? But if there is a verse, a derash from a verse that tells me “a son and not a daughter”—now that doesn’t necessarily mean there is no rationale. It means that that rationale, which by itself perhaps you wouldn’t build on, the verse tells you: yes, you may build on it; it is a good rationale. Okay? And that can still be considered a decree of Scripture. A decree of Scripture does not mean it is unintelligible; it means that without the verse I would not have done it on my own. But once there is a verse, I can definitely understand the logic behind it.

In the Talmudic Encyclopedia and in other places—in fact this comes from Rabbi Yosef Engel—he brings three approaches in explaining the rule that we do not derive punishments from logical inference. One view is maybe… “We do not derive punishments from logical inference” means that when you begin with something by an a fortiori argument, you don’t derive the punishment. Suppose A teaches me through an a fortiori argument that B is also forbidden. A has a certain punishment. Should I punish B with the same punishment? Seemingly, why not? We learned it by an a fortiori argument. They say no: we do not derive punishments from logical inference. Why not? One possibility is that maybe there is a refutation of the a fortiori argument. And how would you know? Regarding the prohibition, if I didn’t find the refutation, I forbid it. But to punish? That’s not certain enough to punish. Maybe there is a refutation. After all, we’ve found refutations to many a fortiori arguments. If I didn’t find one, I still worry maybe there is one and I didn’t find it. That’s one possibility. A second possibility: who says the punishment for the lighter case is sufficient to punish the more severe case? Since it is more severe, perhaps it requires a greater punishment, so you… you can’t derive it and give the lighter punishment; that doesn’t work. Meaning, because that’s also a difference in kind. It’s not just… So one says: then at least give the lighter punishment. No, because a lighter punishment might be different from a heavier one—not only lighter, but different. And if he really deserves a more severe punishment and you don’t know what it is, then don’t give the punishment; the Holy One, blessed be He, will settle accounts with him. Fine? That’s the second explanation, and the third explanation is from a verse: “his sister, the daughter of his father or the daughter of his mother.” And the Talmud says: “the daughter of his father and his mother” didn’t need to be written, because if she is the daughter of his father alone and that is already enough, then obviously “the daughter of his father and his mother” certainly is the daughter of his father, okay? This teaches you that we do not derive punishments from logical inference. There is also “if a man opens a pit, or if a man digs a pit”—to teach, why do we need to write “if he digs”? After all, if one is liable for opening, then for digging all the more so. Rather, this teaches you that we do not derive punishments from logical inference. So it comes from verses. There are several verses that appear in rabbinic literature on this issue. And that flying fly… he definitely has his eye on you, but he’s not convinced to go out. Fine, he says if there’s a bigger opening he won’t be persuaded.

Now, what bothers me in this analysis of the three approaches to “we do not derive punishments from logical inference”? Because it’s not really three approaches. It’s two—two rationales and one source. When people say there are three approaches here, they assume that if there is a source, then there is no rationale, right? If there is a source, then there is no rationale—it’s a decree of Scripture. And that’s not true. You can say: this is the source—and what is the explanation? What is the rationale for why the source says such a thing? I have the two explanations. What are we trying to explain? Why a decree of Scripture is not necessarily something without rationale—in parentheses, in parentheses. Okay? Did he get out? Well done. The question is whether that counts as trapping on the Sabbath. In topology—in mathematics, in topology—every closed line generally divides the universe into two parts: the outside and the inside, okay? When you put a circle here and someone is standing outside, have you trapped him? You trapped him outside, because he can’t enter the inner part. True, the inner part is small, but topologically the outer part is also closed just like the inner part. It’s closed from the other side. So in fact, if I put a fence around this room, I trapped everyone outside, okay? Like with the… okay? Yes, topologically. Fine. Trapping has to be in one motion, of course; in the laws of trapping this doesn’t apply. You have to trap him in a place where you can seize him in one stroke—otherwise it’s not trapping.

Anyway, what I just want to say is that very often, something that comes from a verse—a decree of Scripture—does not mean it has no explanations, okay? I have a verse that says this is the law, and I have two explanations that tell me why that is the law. They do not compete. It’s not three approaches. It’s two approaches, one source, with two explanations of why—why the law is correct. In yeshivot they always think it’s three approaches because they assume that what comes from a verse has no explanation. Because if you offer explanations, that’s a different approach. But it isn’t. In other words, something can come from a verse—that’s one source—and two explanations, two rationales; not three, not three approaches.

So what’s the rationale of “the daughter of his father and his mother”? What’s the explanation? The Talmud says: “the daughter of his father and his mother”—the Torah, why does it need to say “the daughter of his father and his mother”? If you already said “the daughter of his father,” then “the daughter of his father and his mother” is all the more so—it’s obvious that someone who is the daughter of his father and mother is also the daughter of his father. So why was it written? Rather, to teach you that we do not derive punishments from logical inference. After all, it’s an a fortiori argument: if “the daughter of his father,” then “the daughter of his father and his mother” is all the more so. So why write it? To teach you that an a fortiori argument alone is not enough—that we do not derive punishments from logical inference. And there are several such derashot in the words of the Sages.

So what’s the difference between this and a supportive derash? And this… There the assumption is usually that this is a rabbinic law. I’m not talking about an asmachta. A supportive derash is not an asmachta. There it’s… there it’s not a derash at all. Yes, it isn’t a derash but some sort of memory aid or whatever, yes. We’ll talk more about that too, but for now let’s assume that is the accepted view. When I speak about a supportive derash, this is fully a derash, a complete derash. An asmachta usually means a rabbinic law that I hang on a verse, but it is not really derived law. Here I am talking about a Torah-level law, and I attach it to a certain derash from the verses. Only the law was already known before I made the derash, and therefore it is called a supportive derash, but it is not an asmachta. Okay? Fine.

So the claim in effect is that a supportive derash is also a serious derash; it isn’t that it’s not a derash. There too you may not be entirely certain, and still, since the law is given, then if there is a verbal analogy… Nachmanides and his students among the medieval authorities of Spain say that after all we say that a person does not derive a verbal analogy on his own, unless he received it from his teacher. So the claim is that seemingly all verbal analogies should have been laws given to Moses at Sinai, right? But Nachmanides says: but we have disputes about verbal analogies—in Bava Kamma 9, there regarding the best land, the sugya of best land, the dispute between Rabbi Akiva and Rabbi Ishmael, and those are verbal analogies. So how can there be disputes about verbal analogies if it was all given at Sinai—if it’s all tradition? Nachmanides says: it is not really the case that everything was given at Sinai. It is not true that all verbal-analogy derashot are supportive derashot—that is what would seem to follow, but no, not true. With a verbal analogy there has to be some anchor that was given at Sinai—either at Sinai they say there is a verbal analogy between these two verses but they don’t tell me what is learned from it, or at Sinai they give me the law and I find the verbal analogy from which the law could be derived, or there is some such hint or some other hint—but there has to be some hint; you don’t do it entirely on your own. But it’s not that the whole thing is a rigged game, where they gave you everything at Sinai and you just pass it from rabbi to student, but nobody ever actually interpreted that derash. Okay? So very often there are situations that help me reach the halakhic conclusion—so there is some sort of help from above, a derash or something like that, but in essence either it’s a law that comes through tradition, yet I too am making a derash. It isn’t that the interpreter has no role at all in creating this law. It’s some combination of the two.

By the way, the one who makes this distinction between two kinds of tradition—tradition from Sinai, and tradition in the sense of a law that was created over the generations and whose source I lost—is the Netziv in his introduction to the She’iltot of Rav Achai Gaon. He has a commentary called Ha’amek She’eilah, and the introduction to Ha’amek She’eilah is called Kadmat Ha’emek. That introduction is really a work in its own right. And there the Netziv says that when the Talmud says “it is a learned tradition,” sometimes the Talmud says “it is a learned tradition”—Rashi always says there: that means a law given to Moses at Sinai. “It is a learned tradition” means a law given to Moses at Sinai. The Netziv proves from Maimonides that this is not correct. Maimonides does not treat the phrase “it is a learned tradition” as if it meant a law given to Moses at Sinai. So what does it mean? The Netziv suggests exactly this: “it is a learned tradition” means we received some tradition from earlier generations; we know that this is the law. It is not a law given to Moses at Sinai. It is a law created by the Sages, but we have already lost the source—we no longer know from where the Sages derived it, on what derash it rested, where it came from. So the Talmud says “it is a learned tradition.” This is a valid law even if you didn’t find a derash—but not that it is a law given to Moses at Sinai. There was a derash, and it disappeared from us. Okay.

Fine, let me continue. Up to this point I spoke about the relationship between derash and peshat. Now I want to focus for a moment on derash itself. How does this story work? What are these tools of derash? What do they mean? There are claims that say that in effect this is some sort of… that is, for someone who says everything is supportive, the question is less critical: okay, all the laws are known, we are playing games. Who says that? Gersonides, for example, in the introduction to his Torah commentary says that everything is supportive, for exactly that reason—because otherwise anyone can do whatever he wants there. Therefore, if you take seriously the laws that come out of derashot, that means these laws were handed down to us by tradition and were not created from the derashot themselves. That cannot be true—I already said this—it simply cannot be true. The Talmud is full of places where it is clear it didn’t work like that.

There are other possibilities. In Maimonides, for example, in his Treatise on Logic, he talks there about logic, and he says, “and we have other rules whose concern is not this matter,” or something like that—I don’t remember his exact wording. He means the interpretive principles of derash. He says: this is something else; these are not logical rules. What does that mean? A simple view is that this is some kind of code—an arbitrary code. What do I mean by arbitrary code? Think of it like my sending you a letter written in Atbash. Aleph in place of tav, tav in place of aleph, bet and shin, and so on—a kind of code. Okay? Now you read the gibberish I send you. If you have the key, if you are the decoder, you can write out the sentences. Right? Now is there any significance to the fact that specifically aleph switches with tav and vice versa? I could also have switched aleph with vav. It makes no difference; there is no significance to the code in itself. It is an arbitrary code; we just need to agree between us. So likewise, say, the principles of derash—some sort of arbitrary code. The Holy One, blessed be He, agrees with us: know that if there are two words in two places, we make a verbal analogy. He could just as well have said that when the words are different, that’s when we make a verbal analogy, not when they are the same. I’m saying there need be no inherent logic to it at all; it’s only an agreed code. We both agree. Therefore Maimonides says that the principles of derash are something else—they are not rules of logic. There is no logic behind the principles of derash. It is an arbitrary code.

Of course, on the other side of that same conception, it means that this code has to be very unambiguous. Because if you leave it to the logic of the interpreter, and there is no logic in it, then what is he going to do with it? Can I discover an Atbash code without there being any logic to it and without someone telling me that it’s Atbash? Impossible. In other words, you have to tell me: look, I’m writing this in Atbash. Both sides agree. Yes, there’s a word that is supposed to be interpreted through Atbash and a word that isn’t—how do I know what is and what isn’t? Exactly. So if you view it as an arbitrary code, it has to be totally mathematical. Because you have to give me the code so that we know who is synchronized—and that is completely mathematical. But that’s not how it works. We see the Sages arguing: here they interpret, here they don’t interpret, this one interprets it this way, that one that way. It’s not mathematics. There are disputes in the principles of derash, in the principles themselves—not only in their application, but even in which principles there are. You can’t say this is really mathematics, and if it were totally arbitrary then in effect you couldn’t believe anything here. Because if it’s totally arbitrary, you have to give me the full key—let a computer do it. The computer, really, should do it. It ought to be something straightforward. Okay? That is not what appears from the Talmud.

A third possibility—and I think this is the correct one, and I’ll explain it a bit more—is that this is really a kind of language. Another language. Think of these rules as grammar rules. When we write or read the text, there is the plain reading—that is the regular language—we read the text and understand what it says. There is another, parallel reading, and we spoke about the parallels between derash and peshat. This other reading, the reading of derash, also has its own internal logic. It is a kind of grammar of another language. That is how one reads it in the world of derash. It does not replace the peshat—we talked about that. The plain reading remains in place; a verse never loses its plain meaning. Alongside that, there is the same reading in the world of derash. You receive another language with different rules, and it is a kind of language. It has an internal logic; it is not mathematics. We can use it—and there will sometimes be arguments—but we understand how to read the matter.

To explain this more, let me continue that analogy to language. Yes, the first time I heard this was from my friend Moshe Koppel, professor of computer science here—today slandered all over the world; he founded the Kohelet Forum and is in the middle of the whole mess. Anyway, he basically argued as follows. He said that the principles of derash—or I already don’t know what is his and what are my own extensions, doesn’t matter—the model basically works like this. Suppose Moses sat and studied Torah with the Holy One, blessed be He, on Mount Sinai. Okay? In a pictorial way. He’s sitting there with a shtender, the Holy One, blessed be He, beside him, they read the Torah, the Holy One explains it to him, and that’s how He teaches him Torah. Now on the assumption that the world of derash and the principles of derash are a law given to Moses at Sinai—the agreement of all the medieval authorities is that halakhic derash is a law given to Moses at Sinai, the principles, not the halakhic results but the tools themselves were given to Moses at Sinai—then in effect the Holy One sits with Moses and reads, say, the verse “You shall fear the Lord your God”: in peshat, what is needed is to fear the Holy One, blessed be He, and besides that, know that in the language of derash, “et” comes to include. So in the language of derash this means to fear Torah scholars—“You shall fear the Lord your God” includes Torah scholars. So He reads the plain meaning of the verse, and after that He reads the midrashic meaning of the verse, and both readings exist—and that’s how He teaches. Why not also the Aramaic translation? Yes, exactly. And that’s how He teaches Moses our teacher the two languages: the language of peshat, biblical Hebrew, let’s call it that, and the midrashic language. And that is a language like any other.

Now the claim is that because this is a language, think about it for a moment and you’ll see—we learn a language in two ways. One way is like learning in an ulpan: they teach you grammar rules, they teach you some vocabulary, grammar rules, how to build sentences, and so on. If someone knows nothing, that is supposedly the way to teach him. Okay? Children don’t learn that way. Children simply start having people talk to them. This is Noam Chomsky, with linguistics, who basically argues that a human being cannot be born a tabula rasa. If he were born a tabula rasa, a blank slate, it would be impossible to teach him language. You start talking to him—and then what? How does he decode it? How does he know how to apply it to other sentences? Clearly he has certain templates we are born with, templates that help us understand or distill the rules out of the natural speech people use with us. Parents speak with the child, or friends, or the kindergarten teacher, or teachers, and so on—they speak with the child and he learns from that, and he also says sentences he never heard before. That means he knows how to build sentences from other sentences, just as, by the way, software is taught in machine learning. What about a language system based on C? Right. Chomsky’s claim is that it’s universal. In all languages it is the same system. The languages are just different implementations, but linguistic logic is universal among all human beings. That is his claim. There are some disputes about that, but that is the claim.

And for our purposes—I’m not entering the disputes in linguistics right now—what I want to say is that there are basically two ways to learn a language. Now the accepted assumption is that if the principles of derash are a law given to Moses at Sinai, then when Moses our teacher ascended on high, the Holy One, blessed be He, told him: there are thirteen principles—an a fortiori inference, verbal analogy, deriving a principle from one text, deriving a principle from two texts, general and particular, particular and general—yes, the thirteen principles of Rabbi Ishmael—and then He explained to him what to do with each principle, and so on. I claim this is incorrect. Moses our teacher neither dreamed of nor knew all the principles of derash—and this is still a law given to Moses at Sinai. He knew nothing. What did he know? He learned the language. Meaning, they read the verses with him. He doesn’t tell him “‘et’ comes to include.” He says: “You shall fear the Lord your God”—in peshat this means fear the Holy One, blessed be He; in derash it means fear Torah scholars, okay? He says to him “for her” from “a woman”—yes, a verbal analogy. He does not say: this is the principle of verbal analogy—when there is a word here and a word there, what we do is such-and-such. He simply reads with him, just as parents teach language to their child. He simply reads the Torah with him in the way of derash just as he reads it with him in the way of peshat. There are no rules there. How do children know grammar rules? They don’t. If you ask them about letters begadkefat at the beginning of a word, they have no idea what you are talking about. But they pronounce those letters with a dagesh at the beginning of a word. They speak correctly, but they can’t formulate for you the rule behind it. Okay? Or predicate, subject, direct object, indirect object—children don’t know what that is, but they speak correctly. It’s not that they don’t speak correctly. They don’t know how to tell you what the rules are or conceptualize—yes, present to you—the rules that govern correct speech in the language, but they do speak correctly. Okay?

Now the question is how Moses our teacher learned the principles of derash from the Holy One, blessed be He. Did he learn them in the first way, like in an ulpan? Or did he learn them in the second way, like a child learns at home from his parents? And I want to claim it was the second way he learned. What difference does it make? I’ll explain in a minute. But first I want you to understand: I want to claim he learned in the second way. Now why do I say this? First why, and then I’ll say what the implications are. As I said before, all the medieval authorities agree that the principles of derash are a law given to Moses at Sinai. I’m talking now about halakhic derash; aggadic derash, I said, is something more… a law given to Moses at Sinai. Finkelstein, in the introduction to his commentary on the Sifra—he’s a scholar—in the introduction to his commentary on the Sifra he wants to argue that Maimonides did not hold this way. Nonsense. All his proofs are not proofs; it isn’t true. It is clear that this is a law given to Moses at Sinai, even according to Maimonides. Not because it is heresy to say otherwise—just because it’s not true. Look at Maimonides and you’ll see it isn’t true.

That’s on one side. On the other side, all scholars agree that it is not true. The scholars argue it cannot be a law given to Moses at Sinai. There are all sorts of indications of this. For example, the system of the principles of derash develops over history. Hillel the Elder had seven principles; Rabbi Ishmael had thirteen; Rabbi Eliezer son of Rabbi Yose the Galilean had thirty-two. And there are more, by the way—thirty-two isn’t everything. The Talmud is full of all kinds of principles. In other words, this whole business develops and expands. So how is that a law given to Moses at Sinai? Especially according to Maimonides, who says that in laws given to Moses at Sinai no dispute ever arose. How is this a law given to Moses at Sinai? Well, one could say yes. Like perhaps Rabbi Kook thinks—that it is like an axiom in geometry. There is an axiom, and if you accept it, then all the formulas, all the proofs built on it, are in effect included in it. I ask: but is the axiom itself a law given to Moses at Sinai? Well, one could say yes: that there are several principles that are the axioms. If so, then how can there be a dispute between Rabbi Akiva and Rabbi Ishmael? Because in geometry there is no dispute, unless your assumptions differ—curved space and so on, fine. But if the assumptions are these—exactly, that’s the point—it cannot be like that. Therefore I claim that once there is a dispute, it is clear that there is something here that was not given to Moses at Sinai—not in a closed mathematical form.

Okay. I claim both sides are right in this dispute. The scholars are right that it was not given to Moses at Sinai as-is, and the medieval authorities are right that it is a law given to Moses at Sinai. It is a dynamic law given to Moses at Sinai. What do I mean by that? Exactly what I said before. When Moses our teacher ascended on high, he learned the world of derash not in the style of an ulpan, but the way a child learns a mother tongue. The Holy One, blessed be He, did not tell him grammar rules and all that, and general and particular, and verbal analogy, and deriving a principle from one text and from two texts, and all things of that kind. “Anything that was included in a general rule and then left the general rule in order to teach”—Moses our teacher had never heard of any of that, never dreamed of any of these principles of derash. But he did speak the language. He spoke it correctly. If he read the verses, he could tell you what comes out of the derash, from the midrashic reading of these verses. He would not know how to tell you the rules according to which he worked. Just as a child speaks the language correctly but cannot formulate the grammar rules for you. He learns that in school. Before that he doesn’t know, but he speaks correctly. It’s not that thanks to school he speaks more correctly; he doesn’t speak more correctly. Sometimes, yes. Usually not. He learns why that is the correct speech, but naturally he knows how to speak correctly.

Think, for example, of the story about Ephraim Kishon. Kishon was a genius. Really a genius. A humorist—there are already generations that don’t know who he was. But he was a humorist known all over the world, the most successful Israeli writer in the world—at least he was; I don’t know if that is still true today. He was translated into many languages, wrote humorous sketches. Truly a towering genius, no argument at all. Now when he came to Israel, he learned the dictionary by heart, and he coined words himself in Hebrew. Meaning, the man was a virtuoso in Hebrew. He used standard words—not words he invented that nobody knows. He makes fun of… yes, he uses academy words that nobody knows, and he puts them into his sketches and you die laughing; you don’t know where it came from. And those are words of the Academy of the Hebrew Language. Just nobody knows them. “Mirka” instead of “screen”—today we already know what that is, but think of that kind of thing. All kinds of things—“anfil’ot” instead of socks, I don’t know, things like that. You die laughing there. Anyway, there are also words he coined himself, among other things for the Gashash troupe—he wrote sketches for them too. Never mind.

So when Kishon was in ulpan, say—I’m just imagining, I don’t know his exact biography—he learned the dictionary by heart. It took him some time and he learned the dictionary by heart, that’s how he learned Hebrew, he says so. Now he comes to ulpan and they teach him grammar rules and how things work and so on. Then, say, he goes outside, outside the ulpan, okay, and he might start—I’m imagining—correcting all the natives on the street, like: “you’re speaking incorrect Hebrew, because there are rules, and the words should be this way and that.” And all the native Hebrew speakers, yes, clearly have mistakes, distortions, and so on. Now who was right? Obviously they were, not he was. The native speaker, in arguments like this, is usually right. Notice that. The one who comes from the ulpan is mistaken. Why? Because language does not fit into a closed set of rigid rules. You’ll use those rules, and you’ll be right according to the rules—but that isn’t how people speak. That’s not correct Hebrew. So how do I know how to speak correct Hebrew? Because I’m not working with rules. I have some sort of feel—this is the right way to speak, this is the natural rhythm of the language. And I understand that this is how people speak and this is how they don’t. I don’t know how to explain to myself why, but that’s the way it is. And if there are deviations from the rules and you come and say to me, “You’re from the ulpan, you learned it from the rules, you’re calculating: wait, this speech isn’t standard”—usually you’re wrong and I’m right. That is usually the situation. Always—usually that’s the situation. Because grammar tries to define the language exactly. Exactly. And not the other way around. Exactly so. That’s my claim.

In other words, how did grammar develop? It’s not that some sages of some academy sat down and defined the rules and that’s how a language was created. Nonsense. No language was ever created that way. It worked the other way around. People started speaking, somehow a language came into being through speech itself—of course a process of years, fine, though actually not all that many years, but still. Afterward people came and tried to fit the forms of speech into rules. Right, that’s how it usually works. And that’s how the grammar was created that one learns in school. But language is not subordinate to grammar; grammar is subordinate to language. Therefore very often grammatical rules do not really succeed in capturing correct speech. Because it doesn’t work by rules; there are many exceptions. We say: here’s roughly a rule that generally works—okay, that’s a rule. But there are exceptions. That’s what is always so frustrating when learning a language, learning grammar. You try to use the rules, and then suddenly it doesn’t work in many places. You say: there are always exceptions to the rules. Good grief—who built such a stupid language? Give me the right rules! Why give rules that have exceptions? And the answer is: because there are no right rules. There’s no such thing. By definition, the rules are not right. The rules are trying to approximate correct speech. The best approximation I can make—I try to formulate rules that will work in as many cases as possible—but it’s always… it never really works fully. As rules, they are not the true thing; natural speech is the true thing. The rules are an attempt to compress, yes, to formalize, proper speech into a set of rules—but it is always a kind of forcing. It can never really work as a…

Think again of current events. Think of trying to formulate, in a set of rules, the values of a society—to create a constitution. It will never quite work. In court, you always suddenly encounter cases where a simple application of the rules does not lead you to the right result. And then the court has to do things, and then of course disputes arise whether it was right or not—that’s not my issue right now—but it stems from the fact that you cannot fit a rich, living, complex system from life into a set of mathematical rules. A computer cannot be a judge. And a computer cannot speak as we speak. Though today with artificial intelligence it’s already different, because it truly isn’t built in an algorithmic way; with neural networks and things like that, it can work—or at least work better. Yes.

Anyway, the claim is that when Moses our teacher learned the language of derash, it was simply language acquisition. The Holy One, blessed be He, read the verses with him: this is the derash, this is the peshat. He did not tell him by what rules He derived it. Okay? Now Moses our teacher passes this on to the Jewish people—and again, he teaches them the language. Like a parent teaches a child, or a teacher a student. Moses our teacher dies. Then “the children of Israel wept for him forty days and forty nights,” as it says in the Torah. The Talmud in Temurah says that in that time three thousand verbal analogies and a fortiori inferences and seventeen hundred laws given to Moses at Sinai were forgotten. Temurah 15, I think, 14–15. They were forgotten from us, until Othniel son of Kenaz restored them through his analytical reasoning—the a fortiori and the verbal analogy. The laws given to Moses at Sinai cannot be restored. Because that is information. If you lost it, how can you reconstruct information you lost? There is no way to reconstruct it. The Talmud says there… meaning, it does not bring there a reconstruction of a law given to Moses at Sinai; the reconstruction is only for the world of derash. By the way, because of this, commentators on the Jerusalem Talmud such as Pnei Moshe and Korban Ha’edah, I think, say that the rule of a sin-offering whose owner died—even though this is a Torah law, that if the owner of a sin-offering dies it is left to graze until it develops a blemish—there are all sorts of discussions of what one does with a sacrifice whose owners died, okay? That is a law given to Moses at Sinai. It is one of the laws forgotten during the mourning for Moses, and Pnei Moshe argues that our law today regarding a sin-offering whose owner died is rabbinic law. Why? Because the original Torah law we forgot during Moses’ mourning. Now we say: fine, it seems to us this is what should be done—but it is not really that we have the information we received from Moses our teacher. They reconstructed it somehow, or reasoned it out, or I don’t know exactly how—and therefore he says that in essence this is a rabbinic law, not a Torah law.

That’s an anecdote, but what I’m saying is: what is the difference between a law given to Moses at Sinai and laws that come out of derashot? Why can laws that come out of derashot be reconstructed, whereas a law given to Moses at Sinai cannot? Because a law given to Moses at Sinai is a closed collection of details. This is the law, this is the law, this is the law—a list. You forgot it. You have a grocery list—if you forgot it, can you reconstruct what was on the list? No idea. Forgotten is forgotten. That’s it, there’s no way. But with language there is a way. Exactly like machine learning, by the way. What do you do? You take the examples that you did not forget, okay? And let’s try to see—map them, exactly—what they have in common. Distill, conceptualize from them, a set of rules. Fine? And if I succeed, then I carry that over and use it to reconstruct what I forgot. Exactly what we do in grammar, by the way. What do we do? We learn from our parents to speak our mother tongue, right? We encountered I don’t know how many thousands of sentences. From those sentences, not consciously of course, we build some grammatical patterns, and now we can speak sentences we never heard—construct new sentences. How do we do it? Because without noticing, we basically distill from the examples we learned some set of rules or correct ways of speaking, and we know how to build with that new sentences we did not learn. Exactly like machine learning, by the way. What do you do in machine learning? You give the machine examples, and from those examples it learns how the whole thing works, and then it can tell you what the answer will be in examples it has never encountered. Because this is an attempt to imitate, to simulate—yes, a simulation—what human beings do. That is the whole idea of machine learning. How do we do it, but not through a closed positivistic algorithm—rather the opposite, through learning, the way a human being learns? That is the whole idea of machine learning. The whole breakthrough came from there: that if we imitate the way the human brain works, we can perform tasks that by the classical ordinary route there was no chance of performing. To make a computer speak a living language. To speak with me—to pass the Turing test. Right? What is the Turing test? You know it? Turing said—Alan Turing—who does? Exactly. He was also one of the fathers of computability theory and computer science, and he proposed a test: when do you know that a computer is already a human being? From what point should you treat computers as human beings? So he says: if you put in two rooms, in one room a computer, in one room a person, and I sit here before two terminals. One is connected to the computer, one connected to the person, and I talk to them in language. I ask questions, talk, exchange opinions, things like that. The moment I cannot clearly identify which is the person and which is the machine—the machine is a person. That is what is called the Turing test. Okay? Today there are programs that pass the Turing test. Because Turing did not understand that obviously by classical programming a computer would never manage to speak like a person. No chance. You can’t tell a computer with if-this, if-not—that is, all sorts of conditionals—how to speak correctly in every situation. There is no chance. Okay? But if you teach it the idea of language without formulating the rules for it—so that it learns the rules by itself exactly as a person does—then you can get computers that pass the Turing test. Okay? So that, I think, is exactly the idea of the world of derash.

What happens in the world of derash? Moses our teacher learned it from the Holy One, blessed be He. During the mourning for Moses—forty days and forty nights—remember, the Oral Torah was not written down, right? For forty days and forty nights they did not study Torah. The leading sage of the generation died; everyone was in mourning; they didn’t study Torah for forty days and forty nights. No wonder a large part of the material was forgotten. Since we didn’t have it in books and we didn’t deal with it for forty days, a large part was forgotten. But much remained. What does Othniel son of Kenaz do? He restores it through his analytical reasoning. What does that mean? He does scientific research. You say: let’s take the examples we still remember, and let’s try to derive from them the grammar rules. In essence, he creates an ulpan. Let’s try to formulate rules, because we no longer know how to speak the language naturally. We no longer know how to read the verses through the lens of derash. So let’s formulate the rules and use those rules as ulpan students. We’ll use those rules to produce derash in those places that were lost to us, that we no longer remember. That’s basically what he did. Therefore I think the rules of derash were created over the generations. The scholars are right: they were not given to Moses at Sinai. But that does not mean they are not a law given to Moses at Sinai. What was given at Sinai is the language of derash. The conceptualization and formalization of the language of derash—that is the work of human beings. The system of grammar rules governing the language of derash—that is done by human beings. Therefore there can indeed be disputes about this. One proposes one formalization, another proposes another. The language is shared, but when you try to distill the rules, much depends on what you think—which system of rules seems more plausible to you as an approximation of the examples I know. And that is how disputes came about. Rabbi Akiva interprets this way, Rabbi Ishmael that way.

More than that: why does the system of the rules of derash become thicker over the generations? There are more and more rules or principles of derash over the generations. We said seven, thirteen, thirty-two, and there will be more. How does that happen? The answer is: because as history advances, we move farther from the source, Mount Sinai. We lose the natural intuition of a native speaker, of a child—how one speaks the language. We need the help of rules, to learn in an ulpan how to speak this language. Each generation establishes an ulpan for the next generation. And the next generation has to create more rules, because it understands less naturally how one speaks. So it creates more rules and more rules and sub-rules. And if so, then the rule says this, and all sorts of epicycles and deferents, like in Ptolemaic cosmology. In Ptolemaic cosmology—they tried, Ptolemy’s system—they tried to describe all the paths of the stars as circles, because the circle seemed to them, in Pythagorean terms, the perfect shape. Therefore the stars must move in circles. But what do you do when they don’t move in circles? Ellipses—we know today they’re ellipses. But then they said, “A circle doesn’t fit.” So what do we do? Add little circles. There’s a circle, and here another two small circles—you see, this is already starting to look like an ellipse. And another two small circles. And that still doesn’t fit, because a circle plus a circle should be something like this, smoother. So they added another little circle here. They kept adding little circles and little circles to somehow organize the motion of the stars in the language of circles. Until Kepler or Copernicus came—the modern cosmology, the start of the modern age. Forget it—it’s ellipses. Suddenly everything was solved. You didn’t need all the crazy complications they had with circles inside circles inside circles, minus a circle, plus a little circle, to fit everything into the language of circles. No—it’s ellipses, and everything is fine; everything works out exactly. You need nothing.

When we are within a system and don’t understand how it really works, we create more rules and more rules, and add another sub-rule and another rule and another, in order to get as close as possible to natural speech. But someone who speaks the natural speech doesn’t need all those rules. He knows how to speak. Now, the farther we move from the source, from Mount Sinai, our natural control of the speech becomes weaker and weaker, and therefore we create more rules and more rules and more rules. These rules are our own handiwork, but that does not mean derash is not a law given to Moses at Sinai. The scholars are right: the rules were created over the generations. The medieval authorities are right: it is a law given to Moses at Sinai. They are both right. Why? Because it is a dynamic law given to Moses at Sinai. What was given to Moses at Sinai is that there is another language for reading the Torah, the language of derash, besides the language of peshat. Now, the grammar of this language—the rules—we are trying to conceptualize out of the language that was transmitted by tradition from Sinai. So of course over the generations it develops and thickens and branches out—branches in the sense of branches—with more and more branches, and so on. Disputes: these interpret this way and those that way. But in the end, fundamentally, it is a law given to Moses at Sinai—the language. The conceptualizations are ours.

Yes, that continues what we just learned, about the analogy with language. We know that language is—you might call it degenerating, or changing. As you say, little by little we’re sort of becoming detached from it. Then the regularities we distill are regularities that don’t correspond one hundred percent; the whole idea of language is that the rule is approximate. Right. Now my question is: if that is really so, and then a person comes and says, “I used this regularity and I got A, B, C, D,” I tell him: listen, that’s probably in the right area, but it isn’t exact. Where does derash get the power to obligate? Because I don’t have anything better. “A judge has only what his eyes see.” I don’t have anything better. You know what—since you asked, let’s read the Talmud in Pesachim. There’s a fascinating passage in Pesachim, I think 22; let me find it. Ah, that’s almost the same sugya I spoke about this morning in Kiddushin, and I didn’t notice. The same sugya of Rabbi Abbahu. Okay, this is Pesachim 22b.

The Talmud there discusses—not important now—the goring ox that was condemned, and so on. The Talmud says as follows—I’m starting here, I won’t go into the sugya itself. “And the other one: he does not interpret ‘et.’” There is a dispute among tannaim: one interprets the word “et,” one does not. “As it was taught…” Fine, I’ll read it to you in my own words, okay? Listen. “And the other one: he does not interpret ‘et.’” There is a dispute among tannaim whether one interprets “et” or not. “As it was taught,” a famous baraita: Shimon HaAmsuni—and some say Nechemiah HaAmsuni—used to interpret every occurrence of ‘et’ in the Torah. Every place where ‘et’ was written, he would come to include something. Once he reached “You shall fear the Lord your God,” he withdrew. He did not find what to interpret in “You shall fear the Lord your God,” so he withdrew. What does “withdrew” mean? He decided that “et” does not come to include, and he gave up the previous derashot he had made as well. His students said to him: Rabbi, all the ‘et’s that you interpreted—what will become of them? What, you throw out all the derashot because here you didn’t find one? He said to them: just as I received reward for interpretation, so I receive reward for withdrawing. I throw it all out, I give up everything. Until Rabbi Akiva came and interpreted: “You shall fear the Lord your God”—to include Torah scholars. He found a derash even for this verse.

What lies behind this story? It’s really fascinating. Why did Shimon HaAmsuni have a problem specifically with “You shall fear the Lord your God” more than with all the other occurrences of “et”? Multiplying divinities. How can it be that one must fear something additional in the same way one fears the Holy One, blessed be He? That is idolatry by association. Such a thing cannot be. He says: it is impossible to include anything from this verse—out of the question. If from this one cannot include, then apparently the rule that one includes from “et” is not correct, because otherwise it should have worked here too. This is Popper: there is a counterexample that refutes the theory. So I give up everything—“just as I received reward for interpretation, so I receive reward for withdrawal.”

How can that be? But the principles of derash are a law given to Moses at Sinai. Inclusion is one of Rabbi Akiva’s principles of derash. Shimon HaAmsuni was in Rabbi Akiva’s study hall, so he interprets “et” as coming to include. Now, if the principles of derash are a law given to Moses at Sinai, what do you mean, you withdraw? Stay with “this requires further study” on this verse. You have an interpretive principle saying that “et” comes to include. The fact that you didn’t find one verse how to interpret—fine, that requires further study, I understand. But suddenly you throw away all the other derashot too? There is an explicit law from Sinai that says “et” comes to include. Like Ben-Gurion—you know Ben-Gurion used to speak without “et.” Why? “I took the ball”; he wouldn’t say “I took et the ball.” That’s how he spoke. His direct object appeared differently than in our speech. And in the Talmud we see that that is the correct way to read the Torah, because when “et” appears, it comes to include. If they don’t want to include anything, they should write it without “et.”

The claim is that in effect… from here we see that the rule that “et” comes to include is not a law given to Moses at Sinai, because otherwise Shimon HaAmsuni could not have abandoned it. Rather, what happened? Within the framework of conceptualizing and reconstructing the language of derash, Shimon HaAmsuni had a proposal. I see that everywhere there is “et,” if we assume that “et” comes to include, it explains a great many midrashic phenomena. Okay? So apparently there is such a grammatical rule: “et” comes to include. Suddenly he says, wait, wait—but here is an example where that cannot work. Then apparently that rule is not correct. He gave up that rule. Okay? From here we see that the rules of derash, unlike the world of derash itself, which is a law given to Moses at Sinai, the rules themselves are the result of trial and error by the Sages. They are trying to test it: if it works, I’ll assume that is probably the rule; if not, then probably not. So you see that the Sages create these rules. They are not rules received by tradition; otherwise how could he give up all the derashot he made until now?

Now what is beautiful here—in this story in Pesachim 22b—is that then Rabbi Akiva comes and says to him: “You shall fear the Lord your God”—to include Torah scholars. What does that mean? Rabbi Akiva says to him: look, if a theory—that’s the criticism of Popper—if a theory works in very many cases, then one counterexample does not make you throw it away. It does not throw it away. You know, that’s the criticism of Popper. Popper says that a scientific theory must be falsifiable. What does that mean? You have a theory: all ravens are black. That is a scientific theory. Why? Because it can be falsified. If you bring me one raven that is not black, we have falsified the theory. To prove the theory you can’t—you can never know whether you’ve seen all ravens, right? But to falsify a theory, you can. Popper basically argues that a scientific theory is one that is falsifiable—it can be subjected to a falsifying test.

One of the critiques of Popper is that this is a childish way of looking at things. A scientific theory is not falsified like that. If you had a theory that all ravens are black and it stood up in thousands of tests—in Australia and everywhere and among all kinds of ravens, suppose, okay? And now you find one raven that is pink. You would assume either that this is a genetic defect, or that it is not exactly a raven but something else, or who knows what—you would not give up the theory that all ravens are black. Think: we have a theory that all bodies with mass are attracted to the earth, right? Suppose I put this in the air and it just stayed suspended in the air. Suppose, just for discussion—a magician does such things sometimes, right? Would we give up the law of gravity? Obviously not. The law is so strong that clearly it is true, and there is some case here that requires explanation—it requires further study, who knows. Move on to the next Tosafot, as the joke says about Rabbi Akiva Eiger. Someone asked his rabbi a question, a very difficult question. He said to him: look at Rabbi Akiva Eiger on such-and-such a page. He looked there—it had nothing to do with the question. He asks a difficulty on Tosafot, “this requires further study,” and that’s it. What did he mean? He goes back to his rabbi and says: tell me, this has nothing to do with the question I asked you. He says: notice that Rabbi Akiva Eiger asks there two questions on two consecutive Tosafot, and in both he remains with “this requires further study.” That means after the question on the first Tosafot, he stayed with “this requires further study,” and he moved on to the next Tosafot—he didn’t get stuck there.

And that’s the answer to Popper. This is Thomas Kuhn—the paradigms, philosophy of science—Thomas Kuhn, exactly. Kuhn basically says against Popper: there has to be some critical threshold of problems before you throw away a theory. One isolated example does not discard a theory if the theory is strong and well established. You do not discard a theory because of one example. That is what Rabbi Akiva says to Shimon HaAmsuni. He says: “et” comes to include. It is not a law given to Moses at Sinai, this rule—I agree. It is our conceptualization. But look, it works in so many places that if you have one counterexample, I do not throw away such a strong theory because of that. So let us find a solution. Right, it is not ideal. “You shall fear the Lord your God”—to include Torah scholars. On the face of it, I would not have said that, because this is still idolatry by association. What, the Holy One, blessed be He, is to be feared, and you can’t include someone who will be like the Holy One? But the fact is: if there really is a rule that “et” comes to include, and I have guidance from the Torah to include, even though seemingly you can’t include anything alongside the Holy One, then I choose what is least implausible. I could have included doves, trash cans, clouds—I don’t know. No, the least implausible is Torah scholars. Okay? So Rabbi Akiva says: fine, no choice—you have to include something.

By the way, from the Talmud here we see that Shimon HaAmsuni did not accept this from him. Because the Talmud says “shall we say this is a tannaitic dispute?” Right? About the two tannaim who differ at the beginning whether to interpret “et,” the Talmud says this is like Shimon HaAmsuni and Rabbi Akiva. In other words, the Talmud understood that after Rabbi Akiva proposed his suggestion, Shimon HaAmsuni remained in dispute with him; he did not accept it. He really thinks you do not interpret “et”; he backed away from it. Rabbi Akiva thinks you do. The tannaitic dispute here is like the dispute between Shimon HaAmsuni and Rabbi Akiva. But for our purposes, what do we see here? We see a wonderful demonstration of the process I’m talking about, besides the Talmud in Temurah, which I think also illustrates it well. But here we see that the Sages try to formulate principles of derash, and they test them against examples to see: does it work or doesn’t it? They are even willing to throw away principles of derash.

Rabbi Akiva works with inclusion and exclusion; Rabbi Ishmael with general and particular rules. That isn’t because these are necessarily different traditions. There were indeed two different traditions there. The question is how it began. The Talmud in Shevuot says that Rabbi Akiva received from Pinchas ben Yair and Rabbi Ishmael received from Nechunya ben Hakanah—or vice versa, I don’t remember exactly. So these are indeed traditions that passed through the generations. But how did it begin? What, did both descend from Sinai? One descended and the other did not, if at all? My claim is that neither descended from Sinai. Rather, these are two ways of trying to formulate the grammar rules of the midrashic world. He proposed one system; the other proposed another. There is no absolute right or wrong here. These are two languages, both only approximations. You think this is more plausible; you think that is more plausible. That is why two study-houses emerged here with two forms of derash. This is an expression of the fact that the system of the rules of derash is the result of conceptualization by the Sages—or formalization by the Sages. It did not descend from Sinai as a closed list. Yes, the baraita of Rabbi Ishmael with the thirteen principles did not descend from Sinai. It was a conceptualization proposed by Rabbi Ishmael. Rabbi Akiva proposed another conceptualization. That is also why it keeps developing over the generations, because as we move farther from the source—what I said before—we no longer speak it as a mother tongue. We no longer have the feel for what is right and what is not right. With Moses our teacher everything was feel. He would immediately tell you the derash in any verse; he didn’t know what “principles of derash” even meant. He would simply tell you how it ought to be read midrashically. As we move farther away, we lose this ability, and to compensate for it we create more and more conceptual rules. Fine? Therefore the number of rules grows with the generations.

By the way, this is true of all rules of Jewish law. The rules of Jewish law keep developing over the generations until we arrive at hundreds and thousands of rules in books of rules. But these rules never really existed in that form, and no one ever truly used them. They are attempts by later sages to simulate what lay behind what the earlier sages did. Exactly like the Netziv in the introduction to Ha’emek, what I mentioned. You are trying to understand how those sages worked, when they themselves could not have told you: we worked with this rule. Rather, this is an attempt to understand, or to fit into rules, their intuitive mode of thought. That is basically what we are doing here. Therefore the development of the world of derash—what I want to claim—is that it is a dynamic law given to Moses at Sinai. There is a law given to Moses at Sinai that undergoes conceptualization. The naïve conception is that a law given to Moses at Sinai is passed on as-is, like a hollow pipe. Right? I pass it to you, you pass it to the next, the next to the third—there can be no disputes, nothing can happen, it passes through a hollow pipe. That is simply not plausible. Things do not get passed on that way. Think of the children’s game of broken telephone. You know it? One says a word to the next, the next says the same word to the third, and then the last says the word out loud, and it never comes out the same as the first one said. There are always little mistakes in the middle. Things are not transmitted as-is. What gets transmitted is some general mode of thought. Gradually more and more rules develop because we need an ulpan—we no longer speak it as a mother tongue.

There’s a nice example of this brought by the philosopher of science Michael Polanyi. When he talks about this very phenomenon, he says: what is an internship in science? You finish a doctorate, okay? Now you enter research. You’ve studied, you know the material, everything is fine. You still don’t know how to do research. To know that, you have to be someone’s apprentice, someone’s intern—some more experienced researcher who teaches you how the business really works. It is a kind of art. It is not just knowledge. It is understanding where to attack and where not to attack, what is worth checking and what is not worth checking, which technique is worthwhile using. Very often it is not something that emerges from the material you studied. You need to get a feel in your fingers. It is a kind of art. Exactly like speaking a language. So you are equipped with all the rules, supposedly, but you still cannot speak. It is something you need—what Hazal call usage. You work beside the expert, or the craftsman if you like, and you try to understand or imitate him, to understand how it works. In the end you won’t necessarily work exactly like him—you’ll do it differently—but you received something from him that you could not receive in lessons and could not receive in books. It is a kind of craft, like an internship in any field—doctor, lawyer, whatever. You do an internship because beyond the material itself, the formal knowledge itself, you need to understand how the business works on the ground. Okay.

And I once was at a conference—either the Israel Democracy Institute or Van Leer, I don’t remember—there was a meeting of rabbis and jurists. This was in the cheerful days of Aharon Barak, with all the fierce arguments over judicial activism and everything, and there were very sharp disagreements there. And there was with us Professor Englard—no, not Englard, sorry—from Hebrew University, a religious man; I forgot his name, it slipped my mind. Someone—he was a Supreme Court justice, then later returned to the university and taught there. The last positivist in Israel; I forgot his name. What does “the last positivist” mean? There are no positivist jurists anymore; that’s a childish conception. What is positivism? Someone who thinks it’s all deduction from rules—I explained that before. German law, unlike British law. It’s childish; you can’t work by deduction from rules. It doesn’t work. Englard—exactly, Englard. So he was sitting with us there too, and he defended himself. The rabbis were attacking strongly, and he said: listen, this is all professional; there is nothing connected to a judge’s values, ways of thinking; it’s all professional work—this kind of silly positivism. People were tearing their hair out: how can such an idiot not understand? And he was not an idiot. How can he not understand such a basic thing? After all, you see judges rule according to their outlooks, according to their values—you can’t escape that. It’s not even a criticism; a person is the landscape of his birthplace—you can’t be a computer. Okay? Don’t deny it; it’s a simple fact.

And at a certain point I said: wait a second, but he isn’t stupid. There must be something—he’s not talking total nonsense. Let’s at least try to think what the logic behind his words might be. Then I said to them there: look, when you look at us—the rabbis, that is—you say the same thing about us. We do whatever we want. If the rabbi is of this sort, he’ll find a leniency; if he’s of that sort, he’ll find a stringency—Haredi, Hasidic, Zionist, whatever, not important at the moment. You know what he will say in the relevant issues, right? Now your feeling—and there we are insiders—in the legal world I’m not an expert, but in the rabbinic-halakhic world I am an expert. There I play the game, I’m inside it. And I know that this criticism is incorrect—incorrect even though it looks that way. It is true that there is a correlation between what a person says and his outlook. That is true. But it is not true that people cannot talk to each other, exchange views, change views. And it is not true that you simply do whatever you want. Many times you will work with the rules and understand how the business works, and it won’t come out aligned with what you think. And if you are honest, you will give up your own preference. If you are not honest, you won’t. If you are honest, you will give it up, because you understand that here it is not right, even though that is what you think. It is not as open-ended a game as it looks. I’m not claiming it’s mathematics, yes? But it isn’t as open as it seems. Why? Because there are certain disciplinary rules. Nobody knows how to formulate them, but you know: this is done, and this is not done. In other words, there are things you do and things you don’t do. Something developed there in the course of the process. It is not just a collection of formal rules and simple knowledge.

Someone looking from the outside doesn’t know this. He reads the legal rules and says: wait, you derive X from this, he derives Y from this, you do whatever you want. Not true. Because beyond the rules there are modes of thought, ways of handling things, how one compares one matter to another, how one works. I said to them: try for a moment to put yourselves in the shoes of jurists. Probably similar things happen there too. When I look from outside, I say: there are rules, there is this—you do this, you do that—you’re doing whatever you want. But that is not true. He tells you: listen, this is all professional. Now he exaggerated a bit, but what he is trying to tell you is: there are legal modes of thought, and one person can change his mind—it’s not just that I do whatever I want. True, there are biases and there is affinity to my outlooks; there are correlations with my outlooks. But it is not true that my outlooks say everything and I do whatever I want. Not true. There is right and wrong even in the legal world. And that’s what I think he was trying to tell us. In that sense, what he says is indeed correct.

Now one can take this more extremely or less extremely. He was a positivist, so of course he took it too far. But there is something real here. And this is what Thomas Kuhn says in the philosophy of science as well: in science too, you have a theory. Fine, there is a counterexample—you found one pink raven. Okay, still: all ravens are black. All ravens are black except for some problematic examples. Still, the rule is a correct rule. And the same in language. There is a rule that those letters at the start of a word are pronounced one way, but sometimes there is no dagesh because there is an open syllable at the end of the previous word. So there is no dagesh at the beginning. Fine—again a sub-rule that comes to organize the exceptions. Okay? But still there are such rules, and the rules do work. With all the exceptions, you still need usage—you need to understand how to use the rules, because the rules by themselves do not do the job. Not even in science, certainly not in law and speech and language, and in softer fields. Okay? Therefore yes—perhaps only in mathematics does it work entirely by rules. Even there one can argue about it philosophically.

What I want to say—what Polanyi basically says—is: think about the example of Stradivarius, who made perfect violins. He was an extraordinary craftsman. I don’t understand the field, but that’s what those who do understand say. He made perfect violins, simply an amazing artisan. Okay? Now today, with all the insane and precise equipment we have, we still do not manage to reproduce what Stradivarius did. We can’t make violins of that quality. Why is that? Polanyi offered the following explanation. Think of Stradivarius’s apprentice, who worked with him in his workshop. Okay? Stradivarius taught him the principles. Which wood is better, what thickness is preferable, what shape—the principles can be written in a book; one can learn them. But you need to understand with a bit of a sense of smell, as it were. If the wood is like this, make a slightly different volume or shape; if the wood is different, do this. There is something you cannot put into a set of rules. Now if you work with Stradivarius, you acquire the feeling of what is right and what is not right. You won’t be exactly like him—you won’t be. But you’ll be close. And someone who works only with the rules from the book will be very far off.

Now what happens the farther you are from Stradivarius? You no longer speak the natural language. You no longer understand naturally how it works. You rely more on rules. You create more rules. If it’s this kind of wood, do this; if it’s that kind of wood, do that. These are rules that Stradivarius himself would not have known how to tell you that he worked by. We do the conceptualization. We try to look at the different violins of Stradivarius and extract some principles from them because we do not have the natural feel of how to do it. So we try to reconstruct it and do it as close as we can to him. The farther we are from him, the more rules we need to help us imitate what he does. One who is close to him needs only a few basic rules, but he gets the feel in his fingers of what is right to do and what is not. This is exactly what is happening here. Therefore in the transmission of Torah as well—or the Oral Torah—yes, derashot—it works the same way. The farther we move from the source, the less intuitive feeling we have for how to do it correctly. Therefore more and more rules are created. And the rules become more and more detailed as we move farther from the source.

Now you ask yourself: who is smarter? The later generations are smarter, in mathematical intelligence, yes—we have rules and structures and we know how to make practical distinctions and conceptual analyses in the style of Rabbi Chaim, and all the application of rules and sub-rules and comparisons and all that, which the early authorities and the later classical authorities didn’t know how to do. This morning I spoke a little about Maimonides’ wonderful point. So they didn’t know how to do that. Who is more correct? The earlier authorities. Because they are closer to the source; they simply speak it—they feel that it is right. In the yeshivot they always say: if the Rosh brings proof for his words, you can argue. If the Rosh says “it seems to me,” then he is probably right. And there is something very deep in that. Meaning, if I believe he has a healthy feel, that the Rosh has good intuition, then if he says “it seems to me”—listen, it’s right. I don’t have this proof or that proof. If it’s only by force of proof, then I don’t have a feel; I have a proof that this is correct. Proofs—I know how to deal with proofs too. I can argue with you. But if he says: listen, I’m telling you this is right—I speak the language as a mother tongue; you are learning from me in an ulpan. It is simply right. Now go and find the rules why it is right.

And this is exactly the point when people talk about decline of the generations and ascent of the generations. There is no decline of the generations in intelligence. In my opinion there is an ascent of the generations in intelligence. But there is room to speak of a decline of the generations in the sense of losing the skill of using these tools, the natural sense of smell for how to do it—like Stradivarius, yes—how does one do it? And the world of derash is not made up of a rigid set of rules. We are equipped with masses of rules, but we do not know how to interpret. That doesn’t help us interpret, because we have lost the ability to converse naturally in this language. That is exactly the point.

And if I just summarize the picture that has emerged for us here, it is that we are dealing with a flexible law given to Moses at Sinai. More and more rules are created only because we lose more and more of the natural ability, the sense of right and wrong in this context. Therefore more rules are created, and therefore disputes are also created. The disputes are in the conceptualization; the disputes are not in the thing itself. The thing itself—the language—everyone speaks the same language. But in how to formulate the rules, there are differences. Now where will that show up? In those places where we do not have a natural feeling of what to do. If there is a natural feeling of what to do, everyone will do the same thing. And if we don’t have a natural feeling, what do we do? Use rules. Now if you have distilled one such set of rules, the result may differ from someone else who distilled a different set of rules. And that is the disputes between Rabbi Ishmael and Rabbi Akiva. They use rules in order to interpret, because these are places where they no longer have the natural feeling for what to do, so they rely on rules. But once you are already using your rules—“You shall fear the Lord your God”—then either you don’t interpret it, or you interpret “Torah scholars,” or you interpret it to include, I don’t know, Torah scrolls—whatever. Each one will interpret something else. So here disputes can exist—both in methods of interpretation and in the products of interpretation. And therefore too, as I said before, the number of principles keeps growing.

Think, for example, of Hillel the Elder. He had seven principles. Out of those seven, one was general and particular. That’s all. With Rabbi Ishmael there are already three principles of general and particular: general and particular, particular and general, general and particular and general. I claim—and I showed this in the book we have on general and particular from the Talmudic Logic series. On Tova? No—the Talmudic Logic series on Amazon. So the second book in the series is on general and particular. And the claim there is that Rabbi Ishmael’s three principles of general and particular are really just an elaboration of Hillel the Elder’s principle of general and particular. A lot of difficulties are solved this way—scholars get terribly tangled in this. Is a certain principle defined as general and particular, or as particular and general? Now with Hillel the Elder, even particular and general was called general and particular. All these principles were “general and particular,” because they had not gone down to that resolution. You understood that in these circumstances, one applies general and particular like this; in those circumstances, one applies general and particular like that. In a later generation, you no longer understand that intuitively. Ah—if there is a general and particular, you do this; if there is a particular and general, you do that. It’s different. It’s already two different rules. But they are a detailing of the first rule, not an addition. They are simply the continuation of the process of conceptualization that began with Hillel the Elder and continues with Rabbi Ishmael and Nechunya ben Hakanah. So with general and particular you can really see this process, and in the book we show it.

This process—I claim this is what happened in the entire world of derash. In the whole world of derash there is development, but the development is in the conceptualization. The whole world of derash is a law given to Moses at Sinai. But it is not at all that Moses our teacher knew how to recite the thirty-two principles of Rabbi Eliezer son of Rabbi Yose the Galilean. He had never heard of them. All the more so. The Talmud says that seventeen hundred a fortiori inferences and verbal analogies were forgotten during the mourning for Moses. Why only a fortiori inferences and verbal analogies? Because in that period they spoke only of two principles. There was the principle of a fortiori reasoning and the principle of verbal analogy. All the other principles are developments and elaborations of those two. Why specifically those two? Because a fortiori is the logical principle. Fine—analogy, that is basically the normal way of thinking. In ordinary language it’s called a fortiori. And verbal analogy—that’s the textual principles. I spoke about the relation between textual principles and logical principles. There is a textual principle and a logical principle, and that basically covers the whole world of derash. Only then it wasn’t yet developed. Then there were textual principles and logical principles. With Hillel the Elder it was already more developed, but still general and particular was one principle. With Rabbi Ishmael, general and particular had already become three principles. But they were doing the same thing as Hillel the Elder; it wasn’t something else. They simply conceptualized it. Now, conceptualization of course leads to differences. Because once you formulate it that way—every place in the Torah where there is a general and particular is interpreted one way; every place with a particular and general is interpreted another way—I’m not sure with Hillel the Elder it really worked exactly like that. It’s an approximation. It could be that he said: no, it’s something similar, but there is a sense of smell. Sometimes even if a particular and general appears, I won’t interpret it that way; I’ll interpret it differently, because the intuitive sense here says that in this place one doesn’t work that way. We no longer have that intuitive sense, and we cling to the rules. Therefore sometimes we will miss because of it. Nothing to be done. This is the best we have. We have nothing better than that. The natural feel—we no longer have it. Okay?

Rabbi, then a question regarding what you are more or less saying with theory and some exception. No. In a scientific theory there is an assumption—at least according to Popper—that there is some theory which is the correct one. It’s not a collection of cases. There is a theory, and the cases are only its details. Then indeed one counterexample is enough to say: this theory is not the correct one. Maybe it is very close, but it isn’t the right one. But that itself is disputed. Not everyone agrees that there is even such a thing as one correct theory, even if we haven’t reached it. Maybe we haven’t reached it because there is none. The collection of cases is just a collection of cases. Who says there is one theory that explains everything? Then they relate to laws of science like grammar rules—some best approximation we have, but not that there really is some true set of rules such that if we discover it we will understand everything. It is not at all certain that such a thing exists. Einstein very much believed there was such a thing. He searched for the unified field theory. So far they haven’t found it. The question is whether they haven’t found it because we haven’t yet gotten there, not smart enough yet, or because they haven’t found it because there is no such thing. Not simple. In short—what does this refer to? A theory of everything? Say, a theory of everything. Something like that. String theory is something like that. But still I could say of a theory: this theory is not absolute truth. It misses things. It may be that all theories are not absolute truth. That’s what I want to say. There is no theory that is absolute truth. The question is whether there is—even if I haven’t reached it—is there some law that is absolute, that is always correct, or perhaps there is no such law at all? All the laws are our constructions. There isn’t really some true law we are constantly getting closer to. We build our own grammar because it organizes the cases we encountered in the best way available to us. But not that it is really true or false. So there is no absolute truth that a person can discover, and then one has to give up on absolute truth and make do with using the conceptualizations we have as truth in the absolute sense? Yes. The closest thing we have. Maybe—I’m not saying that’s how it is; I’m saying it’s a possibility. It’s a dispute.

Then if there really is some miss, how do I…? Then you missed. What can you do? We are human beings. The Torah was not given to ministering angels. And obviously we miss things. No question. Even laws that were ruled on and are agreed upon by all halakhic decisors—it is not certain they are correct. We may have erred. But this is what we have; I have nothing better. So that’s it—I work with that. A judge has only what his eyes see. Okay? Another excellent question along the way.