Toldot (5765)

From the book Mida Tova: Articles on the Hermeneutical Principles by Rabbi Michael Avraham. Translated from Hebrew using gpt-5.4 (reasoning_effort=high, batch API).

---

With God’s help

Midah Tovah — Eve of the Sabbath of Parashat Toledot, 5766

Questions

1. Is the wording of the Torah optimal?
2. Are there no superfluous words, letters, or verses in the Torah?
3. When are there exceptions to the optimality of the biblical text?
4. Do the schools of Rabbi Ishmael and Rabbi Akiva disagree on these questions?
5. What are the criteria for optimality?

The Hermeneutical Principle

Gezerah shavah (verbal analogy).

A. Summary from Last Year

Mishnah: The meat of a clean domesticated animal cooked in the milk of a clean domesticated animal is forbidden to cook and forbidden for benefit… Rabbi Akiva says: Wild animals and birds are not prohibited by Torah law, as it is said three times, “You shall not cook a kid in its mother’s milk,” excluding wild animals, birds, and unclean domesticated animals.

Gemara: From where are these matters derived, that the prohibition applies to every clean domesticated animal? Rabbi Elazar said: The verse says, “Judah sent the kid of the goats.” Here it says “kid of the goats”; from here we infer that wherever Scripture says simply “kid,” even cattle and sheep are included. But let us derive it from that verse alone. Another verse is written: “and the skins of the kids of the goats.” Here it says “kids of the goats”; from here we infer that wherever Scripture says simply “kid,” even cattle and sheep are included. But let us derive it from that verse. These are two verses that come together, and any two verses that come together do not teach a general rule.

This is satisfactory according to the one who says that they do not teach; but according to the one who says that they do teach, what can be said? Two restrictive terms are written: “goats” and “the goats.”

Babylonian Talmud, Hullin 113a-b

In last year’s essay we dealt with the relation between the two rules: “The Torah speaks in terms of what is usual” and “The Torah speaks in human language.” From there we moved to a general discussion of the relation between Rabbi Akiva’s interpretive system and that of Rabbi Ishmael, and we noted that it is the school of Rabbi Ishmael that generally employs the principle that “the Torah speaks in human language.”¹ Some have seen this principle as the foundation of Rabbi Ishmael’s general method.

The dispute between the two schools of interpretation is not resolved in the Talmud, and there are internal contradictions among the halakhic decisors on this issue.² We brought examples of divergent talmudic discussions that apparently belong to these two schools of thought. The discussion quoted above is probably in Rabbi Akiva’s style, while a parallel discussion in Babylonian Talmud, Bava Kamma is probably in Rabbi Ishmael’s style.

We then considered the explicit talmudic differences between these two schools: Rabbi Akiva expounds by means of inclusions and exclusions, whereas Rabbi Ishmael expounds by means of generalizations and particulars, as in Babylonian Talmud, Shevuot 26a and parallels. We also discussed this basic dispute in several additional essays last year.

In the second part of the essay we dealt with the relation between peshat (plain sense) and drash (midrashic exposition) as reflected in the course of the Hullin discussion cited above. We saw that, as a matter of halakha, the midrashic consideration overrides the plain-sense consideration when the two conflict, even where there is no difficulty at all in the plain-sense reading, as in the example of “an eye for an eye” meaning monetary compensation.

We then proposed a method of globally examining midrashim. There we suggested an a priori comparison among different alternative ways of phrasing the relevant verses, and an evaluation of those alternatives in terms of “economy”: the number of words required as against the number of halakhot (Jewish legal rules) learned. Our conclusion was that considerations of midrashic economy do not provide a full and sufficient explanation for the Torah’s formulation. It also takes into account considerations of stylistic plausibility on the level of peshat.

We saw two possibilities regarding the general relation between peshat and drash:

1. Each level stands on its own and is meant to offer a complete explanation of the wording of Scripture, alongside the other explanation, with drash parallel to peshat.³
2. The two interpretive levels are not independent, and they intervene in one another.⁴

We commented on the position of those who maintain that “two verses that come together do teach,” for their view implies that the Torah is not economical. That is, sometimes it does with two verses what it could have done with one. The opposing intuition holds that when the Torah writes two verses, it is clear that it does not intend us to construct a paradigm from them, since for that purpose one verse alone would have been sufficient. Hence, those who adopt this intuition conclude that the purpose of the two verses is only to teach about the two specific cases discussed in them.

B. The Optimality of the Torah’s Wording

Introduction

As we saw in the summary above, the fundamental dispute between Rabbi Ishmael and Rabbi Akiva concerns the question whether the Torah speaks in human language. The method of Rabbi Ishmael’s school, which does not expound superfluous words because sometimes the Torah speaks in human language, seems at first glance problematic. We are accustomed to the assumption that the Torah is sparing in its words, and therefore every word has a role that cannot be dispensed with. In talmudic discussions we frequently find give-and-take in which each side to a dispute must account for every biblical word, and sometimes every letter, that seems prima facie superfluous according to its position.

Is the wording of the Torah really the most economical possible? Does the Torah express itself in forms that are not “optimal”? Why does it do so? What considerations guide the Torah’s formulation in such cases?

This question carries considerable weight, because it also stands at the center of debates over the divine origin of the Torah. Those who believe in the Torah’s divine origin tend to assume that every word in it is calculated and that every word has a function. By contrast, those who deny the Torah’s divine origin, in their various forms, naturally regard the Torah as a human document, and therefore as incapable of perfection. One would indeed expect a document of human origin to contain superfluous words and non-optimal formulations.

Midah Tovah has set itself the goal of confronting difficult questions through systematic, non-apologetic inquiry and with intellectual honesty. Commitment to tradition is indeed our guiding light, but we are not willing to allow it to distort the empirical results that emerge from research. The basic foundation of such an approach is therefore a straightforward examination of the sources and of our Sages’ treatment of the issue. Interpretations will naturally depend on one’s assumptions and beliefs, but the facts must be examined courageously and honestly.

In the essay for Parashat Ki-Tetze, 5765, we examined the optimality of the Torah’s wording.⁵ This week as well we wish to address this question by examining several considerations of formulation that seem, prima facie, “non-optimal.” In doing so, we rely on determinations made by the Sages or by the early authorities. Of course, each type requires separate study, but presenting them as a whole is of great importance in guiding broader research on this topic, research that should also include focused investigation of each subtype in its own right.

We shall begin with a survey of several different types, and then summarize and offer some initial reflections on the general question.

A Matter Derivable by A Fortiori Inference, Yet Scripture Took the Trouble to Write It

One of the clearest cases of apparent verbal redundancy in Scripture appears in a claim that recurs several times in rabbinic literature in a number of different contexts. There are cases in which the Sages identify halakhot that can be derived by kal va-homer (a fortiori inference), and yet they nevertheless appear explicitly in Scripture itself. This is, prima facie, a clear case of redundancy, and the Sages account for it with the well-known formula: “A matter that could be derived by kal va-homer, Scripture nevertheless took the trouble to write explicitly.”

From a computerized search that we conducted, it emerges that this phenomenon appears only in the Babylonian Talmud, fifteen times, and in Yalkut Shimoni, eleven times. In the Jerusalem Talmud and in other midrashim there is no such formulation at all. In almost every case we found, this claim is bound up with a dispute. Usually the discussion begins with some halakhic disagreement, and in the course of the analysis each side must give an account of the meaning of every biblical word. When one of them reaches an impasse and cannot offer an explanation, he produces this formula as a last resort: “A matter that could be derived by kal va-homer, Scripture nevertheless took the trouble to write explicitly.”

In several places the Gemara emphasizes that this is only a last resort, used when there is no other option. See Babylonian Talmud, Hullin 108b; Pesahim 77b; and Kiddushin 4b: “Where a verse can be expounded, we expound it.” In other words, this formula is used only when there is no way out, when we have no means of expounding the biblical word under discussion.

An interesting example appears in Babylonian Talmud, Bava Metzia 94b:

A paid bailee and a renter swear concerning an animal that was broken, captured, or died, and they pay for loss and theft. Theft is understandable, for it is written: “If the stolen item is stolen from him, he shall pay its owner.” But from where do we know loss? For it was taught: “If it is stolen, stolen” — I have only theft; from where do I know loss? Scripture says: “If it is stolen, stolen,” in any event. This works well according to the one who says that we do not say “the Torah speaks in human language”; but according to the one who says that we do say “the Torah speaks in human language,” what can be said? In the Land of Israel they said: by kal va-homer. If theft, which is closer to unavoidable accident, requires payment, then loss, which is closer to negligence, all the more so. And the other one says: A matter that could be derived by kal va-homer, Scripture nevertheless took the trouble to write explicitly.

The Gemara discusses the obligation of a paid bailee and a renter to pay for theft and loss of a deposit entrusted to them. It adduces an exposition: “stolen, stolen” — in any event, that is, to include loss. The Gemara then raises a difficulty according to the one who says “the Torah speaks in human language.” On that view, the repetition need not be expounded, since it is not superfluous; it is simply part of the Torah’s ordinary manner of expression, like that of ordinary people. The Gemara answers that according to this view the obligation to pay for loss can be learned by kal va-homer from the obligation to pay for theft. It then turns back and asks, according to those who hold that the Torah does not speak in human language, why should the superfluous word be expounded if the halakha in question can be learned by kal va-homer? To this it answers with our principle: “A matter that could be derived by kal va-homer, Scripture nevertheless took the trouble to write explicitly.”

One should note that here this explanation is placed specifically in the mouth of the one who holds that the Torah uses optimal wording and does not employ redundancies in the way common to ordinary human speech. It is precisely the holder of that position who explains that the Torah took the trouble to write superfluous matters that could have been learned by kal va-homer.

Prima facie, this connection runs against the intuitive thesis we proposed at the outset. Rabbi Akiva, who refuses to treat the Torah like a human text — and as we noted in the past, it was not for nothing that he expounded even the crowns of the letters, and derived inclusions from every letter and every connective word, such as every occurrence of the particle et in the Torah — specifically he holds that a halakha that can be derived by kal va-homer is nevertheless written explicitly in the Torah.

Thus, already here we can see that sweeping conclusions about the optimality or non-optimality of the Torah’s wording are not so simple.

“If You Wish, Say: Scripture; and if You Wish, Say: Reason”

A similar phenomenon of apparent redundancy in rabbinic literature appears with respect to reasoned argument. Although we do not find in the Talmud a formula like “A matter derivable by reason, Scripture took the trouble to write,” we do find an alternative formula that, in most cases and for most views, is equivalent to it: “If you wish, say Scripture; and if you wish, say reason.” That is, there is some halakha that can be derived in two alternative ways: by reason or from a verse. It follows that the verse is superfluous; it was written even though the halakha could have been learned from reason.

In the search we conducted in order to locate these cases, we found, to our surprise, results very similar to the previous case: this expression appears only in the Babylonian Talmud, thirty times, and in Yalkut Shimoni, twenty-six times. We found no such expression in the Jerusalem Talmud, nor in any other midrash.⁶

In these cases the Talmud itself, and the commentators after it, sometimes try to identify what the verse teaches beyond what reason already yields. One well-known example is the principle “the mouth that prohibited is the mouth that permitted,” in Babylonian Talmud, Ketubot 22a; see that discussion and the commentators there.

On the other hand, the Talmud quite often asks: “Why do I need a verse? It follows from reason.” This means that there is an assumption that if the halakha can be learned by reason, it should not be written in a verse, and therefore the verse must be interpreted otherwise. Thus here too it seems that this is a last resort, used when no possible interpretation or alternative exposition can be found for the verse.

It is interesting to note that even this formula, “Why do I need a verse?” — not only in contrast to reason, but generally — appears only in the Babylonian Talmud, thirty-four times, and in Yalkut Shimoni, seventeen times. We did not find it in the Jerusalem Talmud or in the other midrashim.

The significance of this last finding is the opposite of what we might have expected. specifically the Babylonian Talmud and Yalkut Shimoni, which as we have seen are willing to interpret Scripture in a non-optimal way, are specifically those that tend to seek an optimal interpretation, and refuse to accept the existence of a verse teaching a halakha that could also be learned from reason.

The explanation is that the Babylonian Talmud and Yalkut Shimoni, which search for and assume the existence of an optimal interpretation, do not always find one. As a last resort, that is, when no such interpretation appears available, they propose the non-optimal solutions we have seen.

Clean Language

In several midrashim the Sages explained that Scripture altered its wording and added letters and words in order to employ clean language.⁷ For example, in Leviticus Rabbah 26, section beginning “Say to,” and parallels:[^9]

“Say to the priests, the sons of Aaron.” Rabbi Tanhum son of Rabbi Hanilai opened with the verse: “The words of the Lord are pure words.” The words of the Lord are pure words; the words of flesh and blood are not pure words… “Pure words” — Rabbi Yudan in the name of Rabbi Yohanan, Rabbi Berekhiah in the name of Rabbi Elazar, and Rabbi Yaakov of Kefar Hanin, and some say in the name of Rabbi Yehoshua ben Levi, said: We find that the Holy One, blessed be He, bent eight letters and did not utter an offensive expression from His mouth, as it is said: “of the clean animals and of the animals that are not clean.” Elsewhere He bent two or three words in the Torah so as not to utter a word of impurity from His mouth, as it is said: “of every animal that is not clean.” It does not say “that is impure,” but rather “that is not clean.” Rabbi Yudan ben Menasheh said: Even when He came to open with the identifying signs of an unclean animal, He did not begin except with what is clean. With regard to the camel, it is not written here “because it does not have split hoof,” but rather “because it chews the cud”; and likewise with the hyrax, and likewise with the hare, and likewise with the pig.

Prima facie, here too there is an addition of words that is not needed for its own sake. However, it would seem that in this case the matter teaches us nothing about the optimality of the biblical wording. The Torah altered its language here in order to use clean language. In other words, these are not superfluous words, and there is a substantive reason for their presence in the text. True, they do not teach us any halakha, but considerations of purity of expression are apparently also factors that affect the optimality of a text’s formulation. Until now we assumed that optimality means only the minimum number of words that expresses the desired contents. But perhaps optimality must also take other criteria into account: the minimum number of words in such a way that there is no defect at all in the text — neither the absence of any needed halakha nor an unclean form of expression.

We may add yet another aspect. First, it should be noted that we do not find additional sources in which the Torah added words in order to employ clean language. Moreover, many have already asked: the Torah itself elsewhere uses the language of impurity, so it is not clear why specifically here it adds words in order to avoid that.

It therefore seems that our initial proposal for understanding this midrash is not correct. The midrash does not teach us that the Torah always chooses a clean formulation. Rather, the point of the midrash is to explain that specifically here the Torah departed from its usual wording in order to teach us that we ourselves should speak cleanly. For that purpose, a single instance of altered wording is sufficient. Therefore, elsewhere the Torah does not add words in order to purify its expressions. The word “impure” is not in itself problematic, and in this midrash it serves only as a device for teaching us the lesson of how to speak.

And so we find in several of the parallel sources that the midrash does not suffice with the claim that the Torah used clean language, but also adds the lesson to be learned. For example, in Otzar ha-Midrashim, edited by Eisenstein, section beginning “A person should always,” p. 274:

A person should never utter an offensive expression from his mouth, for Scripture bent eight letters and did not utter an offensive expression, as it is said: “and of the animals that are not clean” (Genesis 7:8).

If so, the conclusion is the exact opposite. The Torah never adds a superfluous word. Therefore, in all the other places it did not change the formulation from “impure” to “not clean,” because the optimality of wording matters to it. Only in one place does the Torah change the formulation, and even that not for no reason. It does so in order to teach us a needed lesson. If so, the wording is completely optimal.

We should note that in parallel midrashim a discussion arises about additional places in Scripture, and some even cite further sources in which Scripture altered its language. See, for example, Yalkut Shimoni, Parashat Noah, remez 55, section beginning “Male and female shall they be.” Prima facie, these sources contradict our claim, since they indicate a broader phenomenon in which Scripture indeed bends its language in various places in order to employ clean speech, and not only in one place in order to teach us to do so. The difficulties remain in force, of course — why was this not done everywhere? — and perhaps there is here a disagreement between the various passages.

We should also note that the discussion in Babylonian Talmud, Bava Batra 123a is one example of this. There too the Gemara raises a difficulty concerning the Torah’s wording about Leah:

For it is written: “And Leah’s eyes were soft.” What does “soft” mean? If you say literally soft, is that possible? If Scripture did not speak disparagingly even of an unclean animal, as it is written: “of the clean animals and of the animals that are not clean,” would Scripture speak disparagingly of the righteous?…

Again, prima facie, there is here an assumption that the Torah does not speak disparagingly at all, seemingly contrary to our proposal. But perhaps the situation here is different, because disparagement of a righteous woman is indeed inappropriate, and therefore in situations involving genuine disparagement the Torah truly does alter the wording in order to correct it. By contrast, the word “impure” is not disparaging in itself, and therefore the Torah does use it elsewhere. The need to alter it in the case of “not clean” exists only in order to teach us to adopt clean language.

From here there may be a possible way out for all the passages we have discussed: wherever the assumption is that the Torah does indeed alter its wording in order to use clean language, the case involves real disparagement, where such speech is genuinely inappropriate. As stated, our remarks above apply only to the exposition that explains the use of the words “not clean” in place of “impure,” where no real disparagement is involved.

Scribal Ornamentation

Another example of the fixing of the Torah’s wording without adding content is “scribal ornamentation.” Thus we learned in Babylonian Talmud, Nedarim 37b:[^10]

Rabbi Yitzhak said: The scribal reading tradition, scribal ornamentation, words that are read but not written, and words that are written but not read, are a law given to Moses at Sinai. The scribal reading tradition: “earth,” “heavens,” “Egypt.” Scribal ornamentation: “after you have passed,” “after she shall go,” “after you are gathered,” “princes went before, after the musicians,” “Your righteousness is like the mountains of God.”

The Gemara states that there are places in the Torah whose wording was fixed on the basis of considerations of beauty of language. The medieval authorities on that passage disagreed about what precisely this means. The anonymous commentator printed there in place of Rashi holds that this refers to an inversion of word order: “after you shall pass” instead of “you shall pass afterward.” Such a consideration is plausible, because no extra letter or superfluous word is added; only the order is set. Even so, it should still be noted that apparently the other order was better suited to the content — for otherwise what change is there here? — and the Torah nevertheless chose to alter it for the sake of beauty of language. That is, even according to this explanation, there is here some relinquishment of a certain optimality in terms of meaning, in favor of a secondary consideration unrelated to content.

According to the Rosh in his commentary there, the Torah specifically omitted letters in order to beautify the language. Again, this is indeed more economical, but it still represents a departure from fixing the wording solely according to considerations of meaning. According to the Ran in his novellae there, the Torah actually adds words or letters in order to ornament the language: “after you shall pass” instead of simply “you shall pass.”

The Torah Speaks in Human Language

The rule from the school of Rabbi Ishmael, according to which “the Torah speaks in human language,” usually concerns repetitions. There is something akin here to scribal ornamentation for the sake of beauty of language, and by means of this principle Rabbi Ishmael’s school explains the presence of a superfluous word in Scripture. In fact, they go further and maintain that we should not derive an exposition from this word. That is, not only is this word not laden with plain-sense meaning and written only out of linguistic habit, but for that very reason no expositions are derived from it. It is not mufneh — that is, textually superfluous and therefore available for exposition.

We have already remarked several times that mufneh creates a link between considerations of peshat and considerations of drash. When there is a mufneh word, that is, a word that is superfluous from the standpoint of peshat — and at times the discussion even concerns superfluity from the standpoint of drash — the proposed solution is to expound it, mainly by gezerah shavah, which according to at least some views requires at least one word that is mufneh.

The principles surveyed above partially undermine the ground beneath the principle of mufneh. Superfluous words can be explained by various side principles, even when they are not needed in terms of meaning.

Let us recall here that in last week’s essay we noted that, according to scholars, the school of Rabbi Ishmael specifically requires a gezerah shavah to rest on a mufneh word. The conclusion is that not only Rabbi Akiva’s school, which expounds even the ornamental flourishes of the letters, insists on the optimality of the Torah’s wording. Rabbi Ishmael’s school too insists on finding meaning for every word. The rules of beauty of language, clean language, and human language do not express a different conception of the optimality of the biblical text. They are exceptions that provide reasons no less valid than reasons tied to meaning for fixing the wording. An explanation of the necessity of each of these criteria belongs to other essays.

Summary

The phenomenon we noted in the previous section is that even those who seem, prima facie, to relinquish the optimality of the biblical wording do not really do so. We saw similar phenomena in the earlier sections as well.

1. We saw that it is specifically Rabbi Akiva, who maintains the optimality of the biblical text and does not agree that the Torah speaks in human language, who states that “a matter that could be derived by kal va-homer, Scripture nevertheless took the trouble to write explicitly.” That is, he agrees that there are superfluous words.⁸
2. We saw that Rabbi Ishmael, who seems prima facie to adopt a looser view of the biblical wording, specifically insists on exposition from a mufneh word.
3. We saw that the claim “a matter that could be derived by kal va-homer, Scripture nevertheless took the trouble to write explicitly” is a last resort, used only when there is no alternative. Ideally, one always seeks an interpretation or novelty in the verse under discussion.
4. We should recall that these remarks apply only to kal va-homer, or to reason, and do not amount to a sweeping claim about the possibility of superfluous verses in general. On the contrary, the simple assumption is that there is no redundancy, and kal va-homer is an exception, and even then only rarely.
5. We saw that the Babylonian Talmud and Yalkut Shimoni, which seem more inclined to view the biblical text as non-optimal, are specifically the ones that always insist on asking why a verse is needed when there is reason. That is, they too assume that there are no superfluous words or superfluous verses. Only when they reach a dead end do they adopt the solution of “a matter that could be derived… Scripture nevertheless took the trouble to write explicitly.”
6. Throughout the Talmud we find give-and-take that deals with every apparently superfluous word or letter in the biblical wording. The simple assumption is that every word or letter contains some novelty; that is, that the text is optimal.

Question and Proposal

The question that arises here is this: if the text is indeed optimal, why is that principle relinquished in certain cases? Does the Torah insist on optimal wording, or does it not? If it does, then it should do so always. And if it does not, then all the talmudic discussions that seek an explanation for every letter and word become pointless.

This may also be asked from a different angle: does the fact that we do not find a suitable explanation mean that no such explanation exists? Might this not simply be a limitation on our part?

We therefore suggest that statements of the type “a matter that could be derived… Scripture nevertheless took the trouble to write explicitly,” or “scribal ornamentation,” and the like, may be understood as statements that are not categorical. Perhaps they are literary ways of saying that the Sages did not find an adequate interpretation for the verses in question, without thereby asserting that no such interpretation exists. According to our approach, it is possible that the text is entirely optimal, but we are unable to discover all of its meanings. When we lack another meaning, we express that fact in various ways.

It may be that this proposal fits only the statements about a matter derivable by inference, and not statements about scribal ornamentation or clean language, if those do indeed provide sufficient explanations for biblical formulations.

An important indication for our proposal will depend on the question whether the Sages draw any positive conclusion from such statements. For example, if a given expression is defined as scribal ornamentation, do we therefore refrain from deriving any other exposition from it — for instance, is it therefore not mufneh for gezerah shavah? Or if the Sages determine that a certain verse is superfluous because of kal va-homer or reason, do we then cease looking for another interpretation of it? We already mentioned that the commentators do seek an additional novelty in the verse that teaches the principle “the mouth that prohibited is the mouth that permitted,” since that principle can also be learned by reason.

If these expressions are merely literary ways of saying that we did not find the desired interpretation, then no positive conclusion should be drawn from them. If, however, it turns out that such conclusions are in fact drawn, then we have no choice but to conclude that the statements in question are indeed categorical.

Footnotes