Q&A: Hermeneutical Principles
Hermeneutical Principles
Question
Hello Rabbi Michael, may he live and be well,
Since we met last night, it occurred to me afterward that you are an excellent address for a question that came up for me recently and that I haven’t found an answer to:
In Bava Kamma 25a it is explained that when a kal va-chomer cannot be refuted, everyone still says “it is enough” (dayyo). For example: if we did not have a scriptural source that semen conveys impurity in a zav, and we were to learn it by a kal va-chomer — “if something pure in one pure case is impure in an impure case, then something impure in a pure case should certainly be impure in an impure case” — we would say dayyo: that what is derived by the inference can be only like the source case, so it would impart impurity only by contact, like a man with a seminal emission, and not by carrying, like the bodily fluids of a zav.
My question is why, in a gezerah shavah (and according to Rashi in Zevahim 91b and Hullin 120a, also in hekkesh and binyan av), there is an opinion that says “derive from it, but establish it in its own context” (for example, a reed mat in the case of corpse-impurity for seven days, Bava Kamma 25a), and it does not derive from the dayyo said regarding kal va-chomer that in the other hermeneutical principles as well one should not be more stringent with what is learned than with that from which it is learned.
I would be glad for your answer,
Answer
Hello Rabbi Y.,
One should divide the hermeneutical principles into two types: logical principles and textual principles. The textual principles begin from some phenomenon in the text (two similar words — gezerah shavah; a shift from singular to plural — general and particular; and the like). The logical ones (the two forms of binyan av and kal va-chomer, and apparently also two verses that contradict one another) are not anchored in any textual phenomenon, but in substantive logic.
And although even in interpretations based on the textual principles there is ultimately reasoning that relates to the content as well (since one still has to decide what is being included, or in what respect the two things are being compared), there is still a major difference between the types. A clear indication of the difference is that we never find a refutation of a derashah based on a textual principle. (There is apparently one exceptional example in Hullin regarding a general-and-particular derashah, but in the second book of our logic series I showed that this is not so.) The reason is that if there is a textual phenomenon that compels interpretation, you must interpret it. And if you remove the interpretation because of a refutation, you are still left with the question of what to do with the textual phenomenon. Every such textual occurrence is a kind of question (“Why is there a shift from singular to plural?” and the like) that demands resolution. That is why, in a gezerah shavah, we require the term to be available for interpretation, and ordinary similarity between two words is not enough — because two similar words are not themselves a question. It may be that the plain meaning required that wording, and there is no need to interpret it. When you check, you will see that in the principles of general and particular the result is always broader than the source case (that is the very definition of the principle, and all it comes to teach). In general-particular-general, after all, we broaden in two or three respects and do not make a total equation. Everything this principle comes to tell us is not to restrict ourselves only to the source examples, but to expand beyond them. Therefore, if there you were to say “derive from it and from it” (that is, not to broaden), you would collapse the entire set of principles of generalities and particulars altogether — including the principle of general and particular, which at first glance seems not to broaden anything. But that is not so. That principle too broadens in some respect, for otherwise there would be no need for it, and we explained this in the second book mentioned above.
And from here we return to your question. If the inference of a kal va-chomer or a binyan av is refuted, we simply do not make it, and that is that. For even if we do not make it, we are not left with any unresolved question (there simply is no kal va-chomer or binyan av, and that is all). Therefore a refutation knocks out a kal va-chomer and a binyan av, and we do not broaden (that is, we do not say “derive from it and establish it in its own context”). But with a gezerah shavah, which is a textual principle, it is impossible to dismiss the conclusion on the basis of a refutation, because then we have not answered the redundancy or availability in the text itself. Therefore, when we do not find a full analogy between the two contexts compared by the gezerah shavah, we still make the comparison even when the similarity between them is incomplete, that is, when the conclusion is broader and not fully identical to the source case.
Now, I was surprised to discover from your words that Rashi says this also regarding binyan av, which is a logical principle. I had not known this, so I checked. In the Zevahim passage this is not mentioned in Rashi. Only hekkesh and gezerah shavah appear there, both of which are textual principles. And it is no accident that binyan av does not appear. Moreover, there the Gemara itself is not dealing with binyan av. However, in Hullin Rashi does indeed write this also about binyan av, and there he has no choice because that is what the Gemara itself does there (it is a derashah of binyan av, from one verse and from two verses — the common denominator). In other words, this is really explicit in the Gemara and not only in Rashi. And indeed, that is very surprising to me.
Offhand it seems to me that according to the tanna who holds this way, that difference is just an incidental variation and does not constitute a refutation of the binyan av (unlike an extra stringency in what is learned beyond the source case in a kal va-chomer). Why should it be restricted דווקא to grain, wine, and oil? Something like this is discussed with regard to a “minor refutation” in Hullin 115b. And according to the other tanna (“derive from it and from it”), apparently even such an extension is not made (like “dayyo” in a kal va-chomer).
However, according to this, one must ask why the Gemara sees this as a consistent position of that same tanna, who must hold everywhere “derive from it and establish it in its own context.” Seemingly, in each place we should have to examine the matter on its own merits: if the difference constitutes a refutation of the inference, then we should not broaden; and if not, then we may broaden. From this it appears that necessarily in all the places where they disagree about this, the difference is only of the category of “minor” (that is, a distinction that does not constitute a refutation; for otherwise, on the logical plane, one could not make a binyan av at all, as explained above). And yet there is still an opinion that holds “derive from it and from it” rather than “establish it in its own context”; that is actually the innovative view that requires explanation. In other words, the novelty is that according to this position (that of “derive from it and from it”), one nevertheless does not broaden the derived case beyond the source case even when this does not constitute a refutation — a substantive difference (something like the reasoning that we apply even a minor refutation). And the opinion that says “derive from it and establish it in its own context” disagrees, holding that when the difference is merely incidental, we do not pay attention to it.
But according to this, there would be room for that position to say the same thing even in a kal va-chomer: that all the refutations and the dayyo that we find in kal va-chomer apply only where the refutation is not merely minor (rather, where the kal va-chomer is actually refuted, in which case the situation is different). And then we would have no need to resort to the distinction between textual and logical principles. It seems to me that in hekkesh or gezerah shavah we are dealing even with cases where the extension would indeed constitute a refutation, and everything said here applies only to binyan av. And it still requires further analysis.
Discussion on Answer
You are right. “Derive from it and from it” also teaches something and could in principle explain the availability. But still, the difference between logical and textual principles is fundamental here. In logical principles such a situation cannot arise (because you cannot derive more than what exists in the source case; even one who rejects dayyo does not expand without limit), but in textual principles it is at least possible, even if not necessary (and that is why there is a dispute whether we say “derive from it and from it” or “derive from it”). I’ll try to explain further.
In a logical principle, the inference rests on substantive similarity between the two contexts. Therefore there we make the analogy or the kal va-chomer only for things that are similar (“dayyo,” and so on), and there cannot be a refutation of the derivation and yet we still derive. But in a textual principle, the trigger for the derashah is not substantive similarity but the verse. Therefore there is no reason to stop at what is similar. So according to the view of “derive from it,” we derive for all cases that are relevant in the derived context. The similarity to the source context is not essential; what matters is the logic of the derived context.
And from another angle (as a continuation of my previous remarks): in a textual principle, the need to insist on similarity between the derived case and the source case is less than in a logical principle. In a logical principle, you cannot derive beyond what is similar. On what basis would that law apply? In a textual principle, one can in principle derive beyond what exists in the source case, but one needs a reason for it. The difference does not create a refutation of the principle (because the comparison does not rest on substantive similarity, as above), unlike derivation through a logical principle. Therefore, the view of “derive from it and from it” holds that even in textual principles you have only what you found in the source case. But the view of “derive from it” holds that if, from the standpoint of the derived context, there is logic to extending beyond what exists in the source case, nothing prevents that.
I’ll only note that all this is said about gezerah shavah and hekkesh. In the principles of general and particular there is no question at all, because their whole essence is instruction to expand beyond the examples. The structure (general and particular, particular and general, general and particular and general) only determines how far to expand (how many sides of similarity, as in the passage in Eruvin 27). So there it is obvious that one cannot insist on extending only to examples identical to those mentioned (because that could be done with binyan av, and we would not have needed the textual structure of general and particular or general and particular and general).
Good morning Rabbi Michael, may he live and be well,
After a break of a few days, I went back to look at the Gemara in Hullin 120a, and it became clear to me that in the dispute between Rabbi Eliezer and Rabbi Yehoshua, which the Gemara links to “derive from it and from it” or “derive from it and establish it in its own context,” the derived case (terumah) is narrower than the source case (bikkurim). So there is no question here of why we do not say dayyo. And one may say that it is precisely about this that Rashi wrote that in binyan av too they disagreed whether we say “derive from it,” etc. (and see Tosafot Menahot 15a, s.v. “lechem,” cited in the margin, that there is no hekkesh of terumah to bikkurim here, but rather terumah learns from bikkurim by “what do we find,” that is, by binyan av). But in a case where the derived case is broader than the source case, perhaps even in binyan av, as in kal va-chomer, we would say that what is learned can be only like that from which it is learned.
All the best,
Yes, that really is a strange case. “Derive from it and from it” broadens more than “derive from it.” Especially since according to some opinions there, regarding the other fruits it is only rabbinic.
According to Rashi, one cannot derive from a source itself using all thirteen hermeneutical principles by which the Torah is expounded, except for kal va-chomer (Sukkah 31a).
The rule “it is enough for what is derived by inference to be like the source case” is a limitation on how kal va-chomer is expounded.
Hello Rabbi Michael,
What you wrote about the difference between textual and logical hermeneutical principles is good and convincing, but I did not understand the answer to my question: why is it that according to everyone, where a kal va-chomer is not refuted, we still say dayyo and do not extend more than in the source case (that is, we do not say “derive from it and establish it in its own context”), whereas in an available gezerah shavah, according to the view that says “derive from it and establish it in its own context,” we do extend beyond the source case? After all, the question raised by the availability is answered by comparing the derived case to the source case with respect to the law written in the source case and not written at all in the derived case, even not in its narrower scope. So why should we not say in gezerah shavah as well, “it is enough for what is derived by inference to be like the source case,” and “if you grasp too much, you grasp nothing”?