Gate Four: Our Proposal for Explaining Maimonides’ Position

From the book The Spirit of Law by Rabbi Michael Avraham. Translated from Hebrew using gpt-5.4 (reasoning_effort=high, batch API).

---

Our Proposal for Explaining Maimonides’ Position

After presenting three proposals for explaining Maimonides’ position, we shall now offer, in this gate, a different and more radical direction for understanding his view. As we shall see, this direction joins what we have already concluded thus far regarding Maimonides: for him there is apparently no distinction between the question of source and the question of halakhic force. We noted that he apparently does not recognize the question of halakhic force as an independent and well-defined question at all. Here we shall also see clearly how this connects to the hermeneutic dispute, already mentioned above, between Maimonides and Nahmanides: can a single text bear several interpretations? We shall also clarify Maimonides’ language in this root, especially his expression “branches from the roots.” Beyond that, we shall see that this proposal resolves almost all of the difficulties in Maimonides’ position.

Chapter 1: The Concepts “d’oraita” and “d’rabbanan” on the Conceptual Plane

Introduction: “d’oraita” and “d’rabbanan” — the Standard Terminology

As we have already noted several times, it is customary to understand the concept of a halakha d’oraita (a law of Torah origin), certainly in the language of the Sages, in terms of the halakhic force of the law under discussion. A d’oraita law is a law with the highest possible halakhic force, that is, force derived from the Torah. Which laws have such force? Simply put, the distinction is between laws we were commanded directly by God, whether in a halakha le-Moshe mi-Sinai (a law transmitted orally to Moses at Sinai) or in the Torah itself, and laws whose lawgivers are the Sages (for the source of their authority, see the first unit).

As we saw in several places, and especially in the introduction to the first unit, this yields the distinction between legislation and interpretation. Laws that are the result of interpretation are d’oraita laws, since interpretation shows that they are God’s will as it appears in Scripture or in tradition. By contrast, new laws instituted by the Sages are the result of legislation, and are therefore classified as d’rabbanan laws. The force of laws that arise from interpretation comes from God’s will as expressed in the commandments of the Torah or in tradition, that is, the halakha le-Moshe mi-Sinai. These are therefore d’oraita laws. The force of laws that arise from legislation, however, comes from the authority of the Sages, whatever the ultimate source of that authority may be, and they are therefore d’rabbanan laws.

What Are Derashot? A Preliminary Inquiry

What is the status of laws learned from derashot (rabbinic interpretive derivations)? The derashot were created by the Sages, and therefore they would seem to contain a “rabbinic” element as well. To understand this, we must ask what exactly the difference is between a derashah and a rabbinic enactment.

Take, for example, the derashah discussed by Maimonides in the second root: “You shall fear the Lord your God” — this comes to include Torah scholars. Why are Torah scholars included here specifically, rather than Torah scrolls, or chairs? It would seem that behind this derashah stands a rationale to the effect that Torah scholars ought to be honored, and in that sense there is something akin to an enactment here. On the other hand, there is also reliance on something in the verse itself. That is, this is not an ordinary enactment, but something that has some relation to the text. The law of reverence for Torah scholars is not presented as a mere enactment, but as something learned or derived from the verse.

It seems that the difference lies precisely in the question of connection to the verse. If this law were derived solely from the reasoning of the Sages, who thought it proper to require reverence for Torah scholars, then it would be an enactment. But from the way this derashah proceeds, like almost all derashot generally, it appears that the textual indicator also stands in the background. Tradition tells us that the particle et is meant to include something. Only afterward do the Sages search for what it is reasonable to include here, and conclude that Torah scholars are to be included. Had the word et not appeared here, this would have been an enactment. But now that we have a source in the word et, it is a derashah.

Let us imagine what would happen if the word et did not appear in the Torah. The matter would then depend on the reasoning of the Sages. If they had a reason to require reverence for Torah scholars, they would institute an enactment; if not, no such obligation would exist. Now suppose that the Sages in fact saw no independent reason to impose reverence for Torah scholars, but the word et does appear in the verse and must be interpreted. The Sages are then compelled by the rules of derash to include something. They search for what to include, and arrive at the conclusion that Torah scholars should be included.

If their reasoning alone does not lead them to impose reverence for Torah scholars, why do they reach that conclusion here? There are two ways to understand this:

1. Torah scholars are the thing closest to the Holy One, blessed be He, whose reverence is commanded by the verse itself; hence it is reasonable to include specifically them. Here the constraint is interpretive-textual: what is closest to the content of the verse?
2. Torah scholars are the most plausible object of commanded reverence; hence it is reasonable to include specifically them. This is a rational consideration: what is most plausible?

But in both cases, the reasoning by itself would not have led us to legislate an obligation of reverence for Torah scholars. The verse compels us to do so through the tools and rules of derash. That is the essence of derashah. Its foundation is not the reasoning of the Sages; it begins with a textual compulsion. True, what follows depends on reasoning, whether interpretive or substantive, but that is only the second stage. In an enactment there is only the second stage.

If so, laws learned from derashot should seemingly be d’oraita laws, since after the interpretive process we discover that the verse itself commands us to revere Torah scholars. The basis of the obligation is not the reasoning of the Sages but the command of the verse. The reasoning of the Sages serves only to uncover what the verse is commanding us. And indeed, it is generally assumed that laws learned from derashot are d’oraita laws.

Another example: in halakha there is an equation between the laws of the slave and those of the woman, derived through the gezerah shavah (verbal analogy) based on the word “lah” and the word “lah.” This equation is not made because the Sages independently thought it proper that a slave and a woman should be similar for some reason. It is made because that is what emerges from the verses through the use of gezerah shavah. It therefore seems that this law too is d’oraita, and not a rabbinic enactment.

And What About a Halakha Le-Moshe Mi-Sinai?

What, then, of a halakha le-Moshe mi-Sinai? It too would apparently be classified as d’oraita, since it too is not founded on the reasoning of the Sages but on a divine command. Here, admittedly, the command is neither learned from the Torah nor derived from it; rather, it is transmitted through tradition from Sinai via Moses our teacher. But in the final analysis the source of the law’s force is the divine command, and therefore it is a d’oraita law.

Written Torah and Oral Torah

To complete the description of the conventional picture, we should add another common distinction, not necessarily one with direct halakhic significance, between the Written Torah and the Oral Torah. Possibly this is reflected in the Sages’ expression “a matter acknowledged by the Sadducees,” which does carry certain halakhic implications, though only indirect ones, and apparently not with regard to the force and status of such laws. It is important to note that this distinction is not equivalent to the distinction between d’oraita and d’rabbanan, since substantial parts of the Oral Torah belong to the realm of d’oraita law.

The reason is simple and follows from what we have described so far. Laws belonging to the Oral Torah are produced by the Sages, but not all of them are produced as enactments. Some are products of interpretation, and therefore we view them as an uncovering of God’s will and commandments. It is therefore obvious that their halakhic force is d’oraita.

In fact, many treat the division between Written Torah and Oral Torah as a distinction entirely internal to the category of d’oraita. Laws written in the Torah are Written Torah, whereas halakhot le-Moshe mi-Sinai and laws learned from derashot belong to the Oral Torah. On this picture, d’rabbanan laws stand outside that framework.

This description sharpens even further the claim that the d’oraita-d’rabbanan distinction and the Written Torah-Oral Torah distinction are entirely different from one another. The first deals with halakhic force; the second deals with the source of the law. And it is precisely here that we arrive at Maimonides’ position, for as we have seen, he identifies these two planes with one another.

Maimonides’ Position on the Conceptual Plane

In Maimonides’ view, independently of the question of the halakhic status of laws derived by midrash, we see very clearly a completely different definition of the concept of a d’oraita law. His terminology differs from the accepted one. Maimonides holds that a d’oraita law is a law found in the Torah. This is not necessarily a law transmitted to Moses at Sinai, nor necessarily a law that expresses God’s will as manifested in a direct or indirect command. For Maimonides, the expression “d’oraita” means that the law is found explicitly in the written Torah. We should note that this is exactly the literal meaning of the expression d’oraita: from the Torah. A d’oraita law is a law “from the Torah,” not in the sense of halakhic force, as we are accustomed to think, but in the sense of source.

To sharpen the point, let us ask: according to this conception, what is the status of a halakha le-Moshe mi-Sinai? Such a law does not appear in the Torah, and therefore it is not “from the Torah.” As for laws learned from derashot, we shall discuss them below. It is important to notice that this is not the ordinary definition. Maimonides, contrary to the common view, does identify the Written Torah-Oral Torah distinction with the d’oraita-d’rabbanan distinction. “D’oraita” means belonging to the Written Torah, and “d’rabbanan” means belonging to the Oral Torah.¹

We must notice that we have still not addressed the implications of this terminology for the question of halakhic force and status. We are still on the conceptual plane. First we shall bring several proofs from Maimonides’ wording for this claim, and then see what it means, and how it explains the full range of phenomena in Maimonides’ position that we have encountered so far. Only after that shall we turn to the question of halakhic force.

A clear and central source for this definition appears in the responsum mentioned above on the Laws of Marriage, which Nahmanides cites here on p. 54. There Maimonides writes as follows:

Even regarding a matter that is a halakha le-Moshe mi-Sinai, we call it “from the words of the Sages.” Nothing is called “from the Torah” except a matter explicitly written in the Torah, such as mixed wool and linen, diverse species, Sabbath, and forbidden sexual relations; or a matter about which the Sages said that it is from the Torah.

This is the kind of derivation we have already encountered with regard to laws generated through derashot.

From here it is clear that according to Maimonides even a halakha le-Moshe mi-Sinai is not, for example, a d’oraita law. And this is stated explicitly in Maimonides’ Commentary on the Mishnah, Kelim 17:12, where he writes:

Do not be misled by their saying, “Its measure is from the words of the Sages,” together with the rule in our possession that all prescribed measures are a halakha le-Moshe mi-Sinai. For everything that was not explained in the language of the Torah is called “from the words of the Sages,” even things that are a halakha le-Moshe mi-Sinai. When they say “from the words of the Sages,” they mean either the tradition received from the Sages, like all interpretations and laws received from Moses, or an enactment of the Sages, like all enactments and decrees. Remember this too.

It is important to note that this principle appears in exactly the same words in the responsum. That is significant, because some have tried to claim that Maimonides later retracted this innovative principle in his Mishneh Torah. But the responsum was written after the Mishneh Torah, and it refers to laws that appear there, in the Laws of Marriage 1:2. It is therefore clear that this principle remained valid for Maimonides even when he wrote his halakhic code.

We find the same thing in Maimonides’ Commentary on the Mishnah, Mikvaot 7:7:

When they say, “because it is from the Torah,” they mean the basic obligation of immersion in a pool of water. The language of the Tosefta is: “If there is an olive’s bulk from a carcass or a lentil’s bulk from a creeping thing, and there is doubt whether the requisite measure is present or not, the doubt is impure; for anything whose basic rule is from the Torah, even if its measure is from the words of the Sages, its doubt is impure.” We have already explained that this does not contradict their statement that the measures are a halakha le-Moshe mi-Sinai, because everything not found in Scripture is called “from the words of the Sages.”

This is also proven by Maimonides’ wording at the end of the first root, where he writes:

This entire category [that is, the commandments] is wholly written in the Torah; it contains nothing rabbinic, as we shall explain.

Thus the opposite of the term d’rabbanan is “written in the Torah.” That is precisely our point.

To complete the picture, let us now look at Maimonides’ formulations in the second root itself. At the very beginning of his remarks he writes:

If they do not explain this and do not speak of it, then it is rabbinic, because there is no written text indicating it.

That is, Maimonides’ definition here of d’oraita is that there is a scriptural text indicating it, while a d’rabbanan law is one for which no such text exists.

Later in the root he writes:

And all of them, without doubt, are rabbinic if one would count them among the commandments, even though the plain sense of the verse does not indicate any of these matters…

That is, in order for a given law to be counted, and not to be d’rabbanan, it must be included in the plain sense of the text. As we already mentioned, Maimonides’ theory of interpretation, to be discussed further below, implies that a single text cannot bear two interpretations. Thus here too the criterion for d’oraita is presence within the text, or being the interpretation of the text.

This is also how we should understand the explanation Maimonides proposes for the very principle of the second root:

Perhaps you think that I refrain from counting them because they are untrue. Whether the law yielded by that rule is true or untrue is not the reason. The reason is that everything a person derives is but branches from the roots that were stated to Moses at Sinai in explicit form, and those roots are the 613 commandments. Even if the one deriving them were Moses himself, it would not be proper to count them.

Here too Maimonides states that the criterion is not the truth of the law, but whether it is within the text. His claim is that laws derived by midrash are not found in the text; rather, they branch out from it like branches from roots. Therefore, in his view they count as d’rabbanan laws. We shall return to this point below as well.

Maimonides says the same in Mishneh Torah, Laws of Immersion Pools 4:2:

It is from the words of the Sages that drawn water is invalid for immersion, although the invalidation of drawn water is from the words of the Sages, and they learned it by hekesh.

Thus we see that Maimonides treats laws learned by hekesh (textual comparison) as a kind of “from the words of the Sages.” Admittedly, the expression “although” suggests that the plain meaning of the term “from the words of the Sages” does not refer specifically to midrashic laws; but that is exactly our point. In truth, the term “from the words of the Sages” describes all laws not found explicitly in the Torah, that is, what is generally called d’rabbanan laws, such as enactments and decrees. But midrashic laws are also part of that semantic category.²

Toward the end of his remarks in the second root, however, Maimonides formulates the matter somewhat differently:

For everything that they did not hear at Sinai in explicit form is from the words of the Sages.

At first glance this seems to offer a different criterion: whatever was not said at Sinai. According to this, a halakha le-Moshe mi-Sinai would have to be included among d’oraita laws, in contradiction to what we saw above in the responsum and in the Commentary on the Mishnah. But we must notice that Maimonides adds one word here: “in explicit form.” That is, not everything “heard at Sinai” is d’oraita, but only what was “heard at Sinai in explicit form.” According to what we have said so far, this means: only what all Israel heard directly from God in explicit terms, that is, what is explicitly included in the written Torah given to us at Sinai. The word “explicitly” comes to exclude the halakhot le-Moshe mi-Sinai, which are laws from the words of the Sages, since they were not heard explicitly at Sinai, but were given to Moses at Sinai as something accompanying the Written Torah while not contained within it. For this reason Maimonides is also careful in the previous passage to add that same word: “explicitly.”

This also clarifies Maimonides’ words in the responsum cited above, where he writes:

In those chapters I explained that not everything learned by hekesh, or by kal va-homer (a fortiori inference), or by gezerah shavah, or by one of the thirteen interpretive rules by which the Torah is expounded, is Torah law, unless the Sages explicitly say that it is from the Torah; and I brought proofs for this. There too I explained that even something that is a halakha le-Moshe mi-Sinai, we call “from the words of the Sages,” and nothing is from the Torah except a matter explicit in the Torah, such as mixed wool and linen, diverse species, Sabbath, and forbidden sexual relations, or a matter about which the Sages said that it is from the Torah — and those are only some three or four matters. In that book I explained everything, and when you read it, it will become clear to you, although it is written in Arabic, because most of those chapters that cite the words of the Sages are in the holy tongue.

Maimonides’ words here imply that the principle that a d’oraita law is only what appears in the written text, and that a halakha le-Moshe mi-Sinai is “from the words of the Sages,” is supposed to be explained in our root. But on the face of it, our root contains no direct discussion of that question. As we saw, it is discussed rather in the introduction to the Mishnah. So what does Maimonides mean in this responsum?³

In light of what we have said here, however, one can see that the matter is indeed explained in our root as well. Our root discusses the criterion of d’oraita, and determines that only what is found explicitly in the written text is d’oraita. From that, it follows immediately that a halakha le-Moshe mi-Sinai is not a d’oraita law but “from the words of the Sages,” even if Maimonides does not need to address that question explicitly. This is certainly stated several times in his wording in our root, as we have seen. According to our proposal, then, it is entirely clear what Maimonides meant when he said in the responsum that the matter is explained in the second root.

This is also the place to add what we already mentioned in the previous gate, in the discussion of Maimonides’ first objection to Halakhot Gedolot. There Maimonides describes the view of Halakhot Gedolot as holding that everything derived through inclusion belongs to “the aforementioned category.” Our conclusion there was that this means that a law learned through inclusion is actually contained within the verse itself. If so, that is the criterion by which laws can be treated as d’oraita and counted among the commandments. According to Maimonides himself, however, although the criterion is correct, laws learned through inclusion “are not in the aforementioned category”; that is, they are not included in the text, they do not unfold from it, but rather branch out from it. Hence they too are not counted. We see there as well that, according to Maimonides, the concept d’oraita means what is included in the text. His dispute with Halakhot Gedolot concerns the question of what is included in the text and what is not. This is a first consequence of the semantic distinction between d’oraita and d’rabbanan. As we saw, this is not yet a halakhic distinction; here Maimonides determines its significance for the count of the commandments. Below we shall see that it also has halakhic consequences, at least indirect ones.

Let us add here one more point from our interpretation of the final passage of the second root, where Maimonides says that a derashah supporting a law transmitted by tradition is needed in order to show the “wisdom of the text.” There, though for different reasons, we pointed to the possibility that his intention is that the purpose of the derashah is to show us that this law is included in the text itself. Below we shall elaborate on this.

Two Talmudic Sources for Maimonides’ Position

If so, Maimonides identifies two distinctions that almost all the other medieval authorities keep separate: the distinction between Written Torah and Oral Torah, and the distinction between d’oraita and d’rabbanan laws. For Maimonides, Written Torah means d’oraita laws, and Oral Torah means d’rabbanan laws. We shall bring here several proofs from the Talmud for this connection that Maimonides made. See also Appendix 7 at the end of the book.

A.

We begin with the textual variants cited by Henshke, in his article “Sheniyyot le-Divrei Soferim,” among parallel midrashim regarding eruv tavshilin. The Babylonian Talmud, Yoma 28b, states:

Rav, and some say Rav Ashi, said: Abraham our father fulfilled even eruv tavshilin, as it is said, “My Torahs” — one is the Written Torah and one is the Oral Torah.

The Maharsha and the Rashash already noted, and see also Rabbi Elijah Mizrahi on Rashi to Genesis 26:5, that eruv tavshilin is only a rabbinic enactment, and it is therefore unclear why it is called here “Oral Torah.”⁴ The author of Dikdukei Soferim notes there that all the manuscripts read: “My Torahs — one is words of Torah and one is words of the Sages.”

Henshke explains the development of the received reading as an accidental conflation of two different derashot, producing this otherwise implausible expression, and concludes at the end of chapter 1: “Words of Torah fertilize one another and grow rich from one another.” It is unclear whether this is said ironically, but in any case it is not obvious what enrichment he found in such a meaningless mistake. At first glance, this looks like a grafting together of unlike kinds.⁵

Even if the historical account of the corruption is correct, so long as we have found no substantial meaning in the text now before us, it is hard to see such a blind process as an enrichment of Torah. Moreover, the medieval authorities who adopted this reading still need explanation. What is the meaning of their position? Did they not distinguish between d’rabbanan laws and the Oral Torah?

The answer is that apparently they did not distinguish between these two divisions. This is an interesting source for our claim here regarding Maimonides, who identifies the term d’oraita law with the expression Written Torah, and d’rabbanan law with Oral Torah. That is exactly the point.

B.

We should also mention the sugya in the Babylonian Talmud, Bava Kamma 25a. In the Mishnah there, Rabbi Tarfon and the Sages disagree about the rule of dayo — the principle that the result of an a fortiori inference may not exceed its source case. At the start of the sugya, the Gemara raises a difficulty against Rabbi Tarfon:

And does Rabbi Tarfon not accept dayo? But dayo is from the Torah, as it was taught: “And the Lord said to Moses: If her father had but spit in her face, would she not be ashamed seven days?” Then by an a fortiori inference, with respect to the divine Presence, it should be fourteen days. Rather, it is enough that what comes from the inference be like the case from which it is learned.

The Gemara assumes that if there is a verse from which the rule of dayo emerges, then Rabbi Tarfon cannot disagree with it. That is reasonable, so long as Rabbi Tarfon does not interpret the verse otherwise. But the terminology the Gemara uses is, at first glance, problematic: why does the fact that dayo is d’oraita imply that Rabbi Tarfon cannot disagree? Are there not many d’oraita laws over which the Sages did in fact disagree?

At first glance one might infer from here that dispute is the indicator of whether a law is d’oraita; indeed, we mentioned above that some understood Maimonides that way. But as we saw, that is not correct as an explanation of Maimonides’ view. For example, according to Maimonides, although no dispute arose over halakhot le-Moshe mi-Sinai, they are nevertheless classed as “from the words of the Sages.” We therefore need to say that the assumption in this sugya is the reverse: it is not true that the criterion for a d’oraita law is the absence of dispute, but it is true that no disputes arose regarding d’oraita laws. Therefore, if a law is d’oraita, it is clear that we have a tradition about it, and therefore it is clear that no dispute arose about it. But there can certainly be laws about which no dispute arose that nevertheless are not d’oraita laws, such as halakhot le-Moshe mi-Sinai.

In any event, this sugya is indeed evidence for Maimonides’ position, that the term d’oraita is to be understood literally as “from the Torah,” and not as denoting halakhic force. According to the view of the other medieval authorities, there is no connection at all between these questions, and the concept d’oraita expresses the force of a law, not its source or agreement about it. If so, the assertion that dayo is d’oraita would not undermine Rabbi Tarfon at all.

C.

The Mishnah, Orlah 3:9, states: “Kilayim outside the Land is from the words of the Sages.” In the Babylonian Talmud, Kiddushin 39a, this expression is interpreted literally, that is, as genuinely rabbinic law. But in the parallel Jerusalem Talmud it appears that, at least according to Rabbi Yohanan, the phrase refers to a derashah of the Sages. That also follows from his statement in the Babylonian Talmud there: “They administer lashes for kilayim as a matter of Torah law.”

D.

So too we find in Mishnah Sanhedrin 11:3: “If one says there are five frontlets, thereby transgressing the words of the Sages, he is liable, for it is only the interpretation of the Sages.” And indeed, the Sages’ criterion for liability of a rebellious elder is: “He is liable only when he denies and rules concerning a matter whose root is in the Torah and whose interpretation is from the words of the Sages.” See Babylonian Talmud, Sanhedrin 87a and 88b, and also the Jerusalem Talmud. And in the sugya on 87a, the opposite of “the words of the Sages” appears as “the fine points of the Sages”:

The Sages taught: A rebellious elder is liable only over a matter whose deliberate violation incurs karet and whose unwitting violation incurs a sin-offering — these are the words of Rabbi Meir. Rabbi Yehudah says: over a matter whose root is in the words of Torah and whose interpretation is from the words of the Sages. Rabbi Shimon says: even over a single detail among the fine points of the Sages.

We see that “the words of the Sages” is a term for derashot. Enactments and decrees are here called “the fine points of the Sages.”

E.

Another source is Tosefta Mikvaot 5:3, where we read: “Anything whose root is in the words of Torah and whose measure is from the words of the Sages…” There the reference is probably to a halakha le-Moshe mi-Sinai. Above we saw this kind of formulation also in the Mishnah and in Maimonides’ Commentary on the Mishnah to tractate Kelim.

Chapter 2: Applications to Different Halakhic Categories

Introduction

In the previous chapter we examined the conceptual shift Maimonides makes in his understanding of the terms d’oraita and d’rabbanan. Maimonides understands this distinction on the conceptual plane, not on the plane of halakhic force. Admittedly, we saw that this has an implication for the count of the commandments — only d’oraita commandments are included there — but the count itself has no direct halakhic consequence. See the end of the introduction to the book. We have already noted, however, that this Maimonidean division probably does have halakhic consequences, and that will be discussed in the next chapter. Here we shall examine its meaning, and how this distinction is applied to the various kinds of laws.

Expanding Maimonides’ Classification of Laws — the Scope of the Two Roots

Let us return now to the classification of laws as Maimonides presents it in his introduction to the Mishnah and at the beginning of the root. See the full citation at the beginning of the book. There Maimonides classifies laws into five types:

1. Laws transmitted from Sinai that have an anchor in the text, whether by a supporting derashah or otherwise.
2. Laws transmitted from Sinai that have no such anchor in the text — that is, halakhot le-Moshe mi-Sinai.
3. Laws learned through a creative derashah, but not transmitted from Sinai — that is, midrashic laws.
4. Decrees.
5. Enactments.

The first root deals with the last two types. Our root deals with type 3, and also determines a position regarding the first two types. The conclusion, both from what emerges in the second root and from what is written in the introductions, is that types 2 and 3 are laws “from the words of the Sages,” whereas type 1 consists of d’oraita laws.

It is important to note that there is one further type of law that does not appear here at all: laws explicitly written in the Torah. They too were of course transmitted to Moses at Sinai. In Maimonides’ language in our root, as we saw above, they were “transmitted explicitly.” This is also what Maimonides means in his responsum to Rabbi Pinhas the Judge on the Laws of Marriage, when he divides d’oraita laws into two types and writes:

Nothing is “from the Torah” except a matter explicitly written in the Torah, such as mixed wool and linen, diverse species, Sabbath, and forbidden sexual relations, or a matter about which the Sages said that it is from the Torah…

At first glance, the division between these five types of law and the rest — that is, those explicitly written in the Torah — is simply the distinction between the Oral Torah and the Written Torah. It would appear that all five parts listed above belong to the Oral Torah, and only them does Maimonides count in his introduction and in the second root. But as we already noted, what emerges from our discussion here is that in Maimonides’ view the distinction between Written Torah and Oral Torah is exactly the distinction between what is found in the Torah and what is not found in it, in the sense relevant to d’oraita and d’rabbanan.⁶ Below we shall see that the first type above does not belong to the Oral Torah. Rather, as we shall explain, it is found in the Torah explicitly, and is therefore included in the Written Torah and counted as d’oraita law. According to this, the categories are indeed identical for Maimonides, and the explanation of this will be given below.

In fact, there is yet another type of law to which Maimonides does not refer, and which we too shall not discuss here. These are laws for which there is no command at all, neither in verses nor by derashot, and which were also not instituted as rabbinic enactments. Such laws derive from sevara, that is, rational judgment. We shall discuss these laws below, and their status will be examined in more detail in Appendix 4 at the end of the book.

After all this, it seems that, from the perspective of classifying laws — again, not necessarily with regard to their halakhic force, which will be discussed below — we have found three principal types:

1. D’rabbanan laws. These are types 4 and 5 above. In the first unit we argued that this category probably refers to halakhic force as well, though there are disagreements about this.
2. D’oraita laws. These include laws that appear explicitly in the Torah, that is, what was “transmitted at Sinai explicitly,” and also the laws of type 1 here: laws transmitted to Moses at Sinai — not transmitted explicitly, and therefore not included in the plain sense of the Written Torah — but which nevertheless have a midrashic or other anchor in the text. These laws belong to the Written Torah, and according to all views they possess d’oraita halakhic force.
3. Laws that for simplicity we shall call divrei sofrim, that is, “the words of the Sages.”⁷ These are the laws of types 2 and 3 above. Maimonides defines them as laws from the words of the Sages, or as d’rabbanan. These laws belong to the Oral Torah, since they are not found in the Torah itself.

Defining Halakha Le-Moshe Mi-Sinai and Midrashic Laws

The conclusion is that, according to Maimonides, what characterizes both halakha le-Moshe mi-Sinai and laws learned from derashot is that they are not found in the text, at least not “explicitly.”

With regard to halakha le-Moshe mi-Sinai, type 2, the application is the simplest.⁸ Such a law is defined by Maimonides in his introduction to the Commentary on the Mishnah as a law given to Moses at Sinai but lacking any anchor in the text, whether through derashah or reasoning. He does not mean that there can never be any rational way of understanding a halakha le-Moshe mi-Sinai. He means that its foundation is not interpretive reasoning, that is, there is no way of deriving it from the text, whether by reason or through any of the interpretive rules. It is an independent source, outside the Written Torah. Consequently, according to the criterion we have proposed for the term d’oraita, it follows that such a law is d’rabbanan. Above we already saw that this is stated explicitly in the responsum and in the Commentary on the Mishnah: these laws are “from the words of the Sages.” Now we understand that this is consistent with Maimonides’ general definition of the terms d’oraita and d’rabbanan.

According to this, we may also understand why, with respect to some of the interpretive rules, it is said that they are not to be applied to halakhot le-Moshe mi-Sinai. For example, there is a tannaitic dispute whether one may formulate a kal va-homer from such a law. Rashi, on Babylonian Talmud, Pesahim 81a, s.v. “umi,” explains that a kal va-homer may be derived only from something written in the Torah. And Rashi on Babylonian Talmud, Shabbat 132a, s.v. “Akiva,” adds that the Oral Torah was not given to be expounded by the interpretive rules.⁹ On our approach, the explanation is that the interpretive rules are tools given to us in order to operate on the biblical text, that is, on the Written Torah. Therefore, whatever is not included in the Written Torah is not a proper subject for the tools of derash.

There are, however, certain interpretive rules about which the Sages disagreed as to whether they may be applied to halakhot le-Moshe mi-Sinai.¹⁰ It may be that those disputes depend on whether the rule in question is rational in character, in which case it can be applied even to a halakha le-Moshe mi-Sinai, or whether it is an interpretive tool for Scripture, that is, a textual rule, in which case it should not be applied to a halakha le-Moshe mi-Sinai.

Thus Maimonides’ classification into d’oraita and d’rabbanan is equivalent to the ordinary classification into Written Torah and Oral Torah. A halakha le-Moshe mi-Sinai certainly belongs to the Oral Torah and not to the Written Torah, and therefore in Maimonides’ terminology it is called a d’rabbanan law. It is not found in the written text, and therefore it is not Written Torah; hence the rules of derash do not apply to it, at least those rules that function as tools for biblical interpretation.

There are several consequences to this point, namely that a halakha le-Moshe mi-Sinai is not found in the text. Below we shall briefly discuss the law of doubt with respect to such a halakha. See also Encyclopedia Talmudit, entry “Halakha le-Moshe mi-Sinai,” from the paragraph following note 121 onward, where several halakhic consequences are listed that distinguish halakhot le-Moshe mi-Sinai from other laws.

To complete the picture of Maimonides’ terminology, two main points remain to be clarified:

1. Why are midrashic laws that do not have an anchor in the text not regarded as being found in the text itself? What is the meaning of derash if it does not uncover contents already present in the text? This question concerns Maimonides’ theory of interpretation.
2. If midrash is not a tool that interprets what is found in the text itself, as emerges from point 1, then the opposite question arises: why are laws transmitted by tradition but possessing a midrashic anchor in the text considered to be truly present in the Written Torah? Either derash interprets and uncovers what is in the text, or it does not interpret the text.

These were the two main explanatory questions that Nahmanides raised against Maimonides, and with them we shall now proceed. It should be stressed, for the sake of clarity, that these questions still do not touch directly on the issue of halakhic force. They could be raised even if Maimonides were discussing only the question of source, as the Tashbetz understood him.

And Again: Branching Out — “Branches from the Roots”

The first question is why midrashic laws that have no anchor in the text are not considered to be found in the text itself. This is the foundational point in Maimonides’ whole argument, and therefore we must return here to his theory of interpretation. In chapter 4 of the previous gate, section 10, we saw that Maimonides expressed a revolutionary position regarding biblical hermeneutics: in his view, every verse has only one legitimate interpretation, namely its peshat, its plain sense. Derash is not the plain sense of the text, and therefore cannot count, for him, as a legitimate interpretation. Nahmanides attacks Maimonides precisely on this point, saying that on this “falling mountain” Maimonides hung his entire edifice. Why really is this the root of everything? We can now understand it very clearly: if indeed only the plain sense is the interpretation of the verses, then derash, like any other interpretation, cannot count as their interpretation. If so, derash does not interpret the verse, and therefore it is not a tool for extracting what is contained in it. It follows clearly that the laws learned by derash are not found in the text, that is, they are d’rabbanan laws.

At first glance, then, it follows that according to Maimonides derash is not the correct interpretation of a verse, and therefore the law it yields is not d’oraita. And indeed, shortly afterward Maimonides addresses this difficulty, which naturally arises from his approach:

Perhaps you think that I refrain from counting them because they are untrue. Whether the law yielded by that rule is true or untrue is not the reason. The reason is that everything a person derives is but branches from the roots that were stated to Moses at Sinai in explicit form, and those roots are the 613 commandments. Even if Moses himself were the one deriving them, it would not be proper to count them.

He explains that derash is a correct interpretation of the verses. But how can that be reconciled with the claim that there is only one correct interpretation — the plain sense? Maimonides describes this in the metaphor of “branches from the roots.”

What does the expression “branches from the roots” mean in general interpretive theory, and in Torah interpretation in particular? In the first unit we already mentioned this expression and cited it as an example of a special kind of relation: neither unfolding nor asmakhta (a scriptural support text), but branching. Laws that branch from the text are like branches emerging from roots. Maimonides’ intention here is that the branch is connected to the root, but it is an expansion and branching-out of it, and therefore not identical with it. Laws learned by the methods of derash are not found in the text, but rather constitute its expansion, like a branch from a root. Their relation to the text is one of branching, not of unfolding. A relation of unfolding means that the result is only a particular application of the source, the root, and is therefore already contained in it. Branching means that the law produced in this way is a certain expansion of what is in the source, and therefore is not itself present there.

We encountered this mechanism in the previous gate as well, in chapter 2, where we dealt with Maimonides’ objections to Halakhot Gedolot, especially the first objection. Let us recall that we saw a similar phenomenon in the first unit, where we argued that such a relation, a kind of branch from a root, which we called there “branching,” also exists between rabbinic laws and the d’oraita prohibition of “do not turn aside.” There too we cited the example, brought by Maimonides himself with respect to our root, from chapter 14 of the Laws of Mourning, where the commandments to gladden bride and groom and to bury the dead branch out from the commandment “You shall love your fellow as yourself.”

The general conclusion from all these sources is that, according to Maimonides, neither laws that merely unfold from a commandment nor laws that branch out from it are included in the count of the commandments. Laws that unfold are not included because they are already counted within the commandments that contain them. Laws that branch out are likewise not included, even though they do have an independent status, since they are genuinely independent branches. The reason is that these are not d’oraita laws, because they are not found in the text itself and do not belong to the Written Torah. Only commandments found in the text are included in the count of the commandments.

Thus, in the first two roots, Maimonides determines that laws which are d’rabbanan are not to be included in the count of the commandments, because only commandments found in the Torah itself are included there. The first root excludes d’rabbanan laws in the sense of halakhic force, that is, in the usual sense: enactments and decrees. The second root deals with d’rabbanan commandments in the sense of source, not necessarily in the sense of force.

In light of our discussion here, we can say that the reason underlying the first two roots in Maimonides is the failure to count laws that are not located within the Written Torah, that is, because they belong to the Oral Torah. Maimonides calls laws of these types d’rabbanan, or “from the words of the Sages.” Types 2 and 3 are discussed in the second root and are “from the words of the Sages”; types 4 and 5 are discussed in the first root and are d’rabbanan.

In the previous unit we already explained the mechanism of branching from several angles. Here we shall continue and deepen the explanation, and by doing so we shall also arrive at the differences between these two roots.

The Relationship Between the Results of Derash and the Verses

In light of what we have seen so far, according to Maimonides a d’oraita law is a law found in the text, that is, a law belonging to the Written Torah. If so, the two types of law, 2 and 3, that Maimonides classifies as divrei sofrim are laws not found in the text. With regard to halakha le-Moshe mi-Sinai, we saw a straightforward application of this criterion: it is not written in the Torah and is therefore classified as divrei sofrim. But what about type 3, namely midrashic laws?

Above we saw that Maimonides claims that such laws are not included in the verses, since they cannot count as interpretation of them. For Maimonides only the plain sense is the unique interpretation of verses. But how can we truly understand the claim that such laws too are not included in the written text? In another formulation, we may ask, as Nahmanides framed the issue: how can tools given to us by God at Sinai as keys to the text, applied to the text itself, which was also given at Sinai by God, produce laws that are not included in the text? When God wrote the verse in question, did He not also intend these laws? If so, are these laws included in the text or not?

Several writers have already wrestled with this issue and formulated it in different ways. We shall bring here the views of two of them, and then propose a fuller picture.

Rav Pinchuk’s Description

Rabbi Moshe Pinchuk, in his article on the nature of the thirteen interpretive rules,¹¹ suggests two ways of understanding the thirteen middot:

1. The thirteen middot are interpretive tools by means of which we interpret and understand the intention of the words. They are the tools of an archaeologist, by which he uncovers what lies below the surface.

In our terms, this is a way of viewing the methods of derash as keys to decipher what is present in the text.¹²

2. The thirteen middot are not interpretive tools. They do not reveal what already exists deep within the text. Rather, they are creative tools: the Written Torah is the raw material, and the thirteen middot are the implements by which laws are fashioned out of that raw material. They are the chisel and hammer with which the sculptor creates the sculpture.¹³

Rabbi Pinchuk explains that the metaphors of archaeologist and sculptor are not accidental. At first glance, both perform a similar action: the archaeologist extracts from the ground an object of definite form by removing the soil that surrounded and covered it. The sculptor likewise creates a form from wood or stone by removing what is superfluous and leaving what is required. Superficially this is the same action, but in reality it is radically different. The sculptor creates; the archaeologist reveals. In both cases the result was in some sense already inside the material, but in the sculptor’s case that is only a formal fiction. In reality the form was entirely created by the sculptor. The material was present in the raw material, but the form is wholly his work. These are the two approaches he offers for understanding the function of the interpretive rules.

We may note that a similar distinction can be applied to Maimonides’ own formulation at the beginning of the root, where he writes:

We have already explained, in the opening of our composition, the Commentary on the Mishnah, that most of the laws of the Torah are extracted through the thirteen interpretive rules by which the Torah is expounded…

The wording that laws are “extracted” from the verse can be understood in two ways:

1. They extract from the verse what had always been there.
2. They now draw from it something new.

Rabbi Pinchuk points there to one implication of this distinction in understanding the sugya in Babylonian Talmud, Shabbat 87a:

It was taught: Moses did three things on his own initiative, and the Holy One, blessed be He, agreed with him: he added one day on his own initiative, he separated from the woman, and he broke the tablets.

He added one day on his own initiative. What did he expound? “Today and tomorrow” — today is like tomorrow: just as tomorrow includes its night, so today includes its night. But the night of the present day had already passed. Learn from this that two days are required in addition to today. And from where do we know that the Holy One agreed with him? Because the Divine Presence did not rest until Sabbath morning.

He separated from the woman. What did he expound? He reasoned by an a fortiori inference on his own: If Israel, with whom the Divine Presence spoke only once, at a fixed time, and the Torah said to them, “Be ready… do not approach a woman,” then I, with whom the Divine Presence speaks at any moment and sets me no fixed time, all the more so. And from where do we know that the Holy One agreed with him? Because it is written: “Go tell them: Return to your tents,” and immediately afterward: “But you stand here with Me.” And some say: “Mouth to mouth I speak with him.”

He broke the tablets. What did he expound? He said: If with regard to the Passover offering, which is only one of the 613 commandments, the Torah says, “No foreigner shall eat of it,” then with regard to the whole Torah here, while Israel are apostates, all the more so. And from where do we know that the Holy One agreed with him? Because it is said, “which you broke,” and Resh Lakish said: “May your strength be praised for having broken them.”

Tosafot there ask the following question, s.v. “u-mah Yisrael”:

If so, what does it mean that he added the day on his own initiative? It is from the Torah, for a kal va-homer was given to be expounded. One may answer that this was not a full kal va-homer, because there it was different on account of the Ten Commandments.

Tosafot take for granted that when a person expounds a kal va-homer, he is deciphering what is written in the verse. Therefore they ask: why does the baraita describe these additions as things Moses did on his own initiative? The kal va-homer “was given to be expounded,” that is, it is an interpretive rule given at Sinai, and applying it to the verses yields laws hidden within them. We should note that even in their answer, which we need not enter here, Tosafot do not retract that assumption.

But if we assume that the kal va-homer is a creative rather than a deciphering tool, then Tosafot’s question does not arise at all. Moses added something on his own initiative, and supported it by a kal va-homer. Even if Moses’ kal va-homer was a full one, the kal va-homer, as one of the interpretive rules by which the Torah is expounded, is not a deciphering tool but a creative one. It does not reveal what is written in the Torah itself; it adds to it and expands it. Therefore it is perfectly appropriate to call this “on his own initiative.”¹⁴

It is also worth bringing part of Rabbi Pinchuk’s continuation there. He goes on to cite a dispute between Amoraim in Babylonian Talmud, Gittin 60b:

Rabbi Elazar said: The Torah is mostly in writing and only partly oral, as it is said: “I write for him the majority of My Torah; they are regarded as a strange thing.” Rabbi Yohanan said: It is mostly oral and only partly in writing, as it is said: “For according to these words.” But according to the other view too, is it not written, “I write for him the majority of My Torah”? That verse is an expression of astonishment: “Shall I write for him the majority of My Torah? They would be regarded as a strange thing!” And according to the other view too, is it not written, “For according to these words”? That is because they are difficult to learn.

The dispute concerns the definition of the concept Oral Torah. According to Rabbi Elazar, most of the Torah is Written Torah; according to Rabbi Yohanan, most of the Torah is oral. Later authorities have raised several questions here.¹⁵ In essence, the dispute seems rather strange. How can one even measure “most” and “least”? And if one can, then what is the dispute? Let us take the books and see. Moreover, is it not obvious that most of the Torah is Oral Torah? The Five Books do not equal in scope all the Talmuds, Toseftot, baraitot, and so forth.

Rashi there explains as follows:

Mostly in writing — most of the Torah depends on midrash, because it is written in order to be expounded by general and particular statements, verbal analogy, and the other interpretive rules by which the Torah is expounded.

And only partly oral — because only a small portion has no textual hint from which it can be learned, but was told to Moses orally.

According to Rashi, the Amoraic dispute concerns the status of laws that arise from the interpretive rules: are such laws considered Written Torah or Oral Torah? According to Rabbi Elazar, only halakhot le-Moshe mi-Sinai count as Oral Torah; according to Rabbi Yohanan, midrashic laws do as well, and therefore in his view the Oral Torah constitutes most of halakha.¹⁶

Plainly, the dispute seems to turn precisely on the question whether the interpretive rules reveal or create. If they reveal, then midrashic laws belong to the Written Torah. If they create, then midrashic laws belong to the Oral Torah. In this regard the Gemara in Hagigah 3b says:

Scripture says “planted”: just as a planting bears fruit and multiplies, so too the words of Torah bear fruit and multiply.

Midrash expands the Written Torah, and the “branches that emerge from the roots” are an expansion of the written text, that is, of the root.

Admittedly, the discussion there concerns the distinction between Written Torah and Oral Torah, not the distinction between d’oraita and d’rabbanan. But as we saw above, Maimonides identifies these two questions, and therefore the conclusions of that sugya can certainly be applied here. According to Maimonides, the interpretive rules create and do not reveal; therefore their conclusions are not included in the Written Torah and are accordingly called d’rabbanan laws. According to Nahmanides, by contrast, the rules decipher, and therefore midrashic laws are indeed included in the Written Torah and in d’oraita law.

Perhaps this also provides a source for Maimonides’ statement in the responsum that almost all midrashic laws are the result of creative derashot, whereas derashot that support preexisting laws are only “some three or four,” and no more. Possibly his source is this sugya in Gittin, from which it appears that most of the Torah is oral and only a minority is written.

According to this, however, Maimonides would not interpret the Gemara there as Rashi does. In his view, the Gemara contrasts creative derashot, which belong to the Oral Torah, with supportive derashot, which belong to the Written Torah. According to this, a law of type 1 — a law transmitted to us by tradition and supported by a derashah — is considered written in the verses themselves, and such laws are therefore classified by Maimonides as d’oraita. Below we shall explain why it is truly considered written in the Torah.

Nahmanides himself apparently understood the dispute with Maimonides in this way, and therefore he writes, on pp. 52-53:

Perhaps he, meaning Maimonides, holds that a law derived through one of the interpretive rules, although true, has no hint in the verse, and Scripture was written only for its plain sense, not in order to derive that law from it; therefore it is called rabbinic. And when we hear them say that it is from the Torah, or that it is the very substance of Torah, we shall count it, because by tradition it is known that the inclusion or the verbal analogy in that verse was written in order to derive from it that learned law. But with regard to those cases where they did not mention that they are from the Torah, the truth of that law remains doubtful: perhaps it is entirely from the words of the Sages, innovated by a court; or perhaps they learned it from Moses our teacher, and it is called “from the words of the Sages” only because it has no hint in the written Torah and is not included in “Moses commanded us Torah,” meaning the 613 commandments. And we have not found such a view among the Sages, for with them all the interpretive rules are as though explicit in the Torah, and they derive them on their own.

To conclude, Rabbi Pinchuk’s account still does not really answer the question of how exactly we are to understand the logical role of derashot and the interpretive rules. Are they merely asmakhtot? That is not a plausible reading of Maimonides. Maimonides explicitly distinguishes between asmakhtot and midrashic laws, both in his introduction to the Mishnah, where asmakhtot belong to types 4 and 5 while derashot belong to types 1 and 3, and in the second root, where he objects to Halakhot Gedolot separately for counting asmakhtot and separately for counting midrashic laws.

Moreover, it is not clear how Rabbi Pinchuk would explain the fact that when there are derashot supporting laws received by tradition, they reveal and decipher rather than create. Why does the function and logical status of the inferential process change in these cases? Why not say there too that the midrash offers a grounding for the law transmitted to us, but that the grounding is an expansion of the verses rather than a deciphering of them? According to Rabbi Pinchuk’s explanation, one would have expected to say that a supportive derashah reveals that the anchor of the transmitted law is an expansion of the verses, not an analysis and deciphering of them.

Such a result would classify type 1 laws as d’rabbanan, since they are not found in the text itself. And indeed, both halakhot le-Moshe mi-Sinai, type 2, and midrashic laws, type 3, are classified by Maimonides as d’rabbanan laws. But type 1 laws he classifies as d’oraita. Rabbi Pinchuk’s account does not answer this question.

Nahmanides, in the passage quoted above, does touch on this point. He explains that according to Maimonides the tradition reveals to us that in such cases the derashot derive a law that actually exists in the text. The text was written for the sake of deriving that law, unlike the case of ordinary type 3 midrashic laws. But the reason for this is not clear. How can one explain logically that a midrash, which is an expansive tool, can also reveal what is latent in the verse? Either it is a revealing tool, in which case every use of it should yield a d’oraita law, or it is not, in which case the law yielded by it should never be d’oraita. Nor can one say that the existence of tradition alone makes the law d’oraita, because a law transmitted by tradition, that is, a halakha le-Moshe mi-Sinai, is also considered by Maimonides a d’rabbanan law.

In fact, to understand Maimonides we need to look for an intermediate category between asmakhtot and tools of logical deciphering and analysis, one that will describe the logical status of the interpretive rules as expansive tools. Such a category must also explain why, when there is a tradition, the rules of derash function as deciphering tools, so that the result is included in the Written Torah and counts as d’oraita; whereas when there is no tradition, those very same tools function expansively, and the laws produced by them are d’rabbanan. Halbertal continues precisely this logical discussion, and we now turn to his proposals.

Halbertal’s Description¹⁷

In his article, Halbertal proposes an interpretation of Maimonides’ description of midrashic laws as “branches from the roots” that can partly answer the question of the logical character of the interpretive rules. It should already be noted, however, that in the background of Halbertal’s discussion stands the assumption that Maimonides’ distinction between d’oraita and d’rabbanan commandments is connected with the distinction between commandments over which disputes did or did not arise. In the second part of his article he draws a number of conclusions from this about Maimonides’ theory of interpretation, in contrast to that of Nahmanides. But above we saw that this assumption is untenable, from several explicit statements of Maimonides himself, and from reason as well. A halakha le-Moshe mi-Sinai is a law transmitted by tradition, and therefore no dispute arose over it, yet Maimonides classifies it as d’rabbanan. Conversely, there are d’oraita laws explicitly written in the verses, that is, reached through plain-sense interpretation and not through derash, over which disputes did arise. We therefore do not adopt Halbertal’s conclusions as a whole, though we shall make some use of the way he interprets the expression “branches from the roots.”

First, we should note that in a footnote there, note 20, Halbertal adopts the intermediate approach: the status of midrashic laws is indeed d’rabbanan, but the derashot that yield them are not asmakhtot. The question he tries to answer is how those two assumptions can be reconciled. He attempts to do so in chapter 3 of his article, where he proposes an explanation of the metaphor “branches from the roots” and, within it, an interpretation of the function of the interpretive rules as expansive tools.

Halbertal distinguishes between two types of interpretation:[^211]

1. The first type is interpretation meant to clarify or disclose the meaning of the terms of the text,¹⁸ and the rules that guide it are rules of discovery and definition.
2. The second type of interpretation does not clarify the text, but draws additional conclusions from it. The rules of this second type of interpretation resemble rules of inference.

Halbertal’s distinction is very similar to that of Rabbi Pinchuk. The first type is interpretation as a tool of deciphering, whereas the second is expansion and inference of additional conclusions from what is said in the text itself.¹⁹

Halbertal argues that Maimonides means that the methods of derash belong to the second type, the inferential type. To illustrate this, he gives the following example:[^214] suppose there is a law stating, “Vehicles shall not enter a public garden.” An interpreter may face a definitional problem, such as determining the scope of the term “vehicle.” Does it include bicycles? Motorized bicycles? Motorcycles? Something else? When an interpreter argues that the term includes motorized bicycles, he is trying to clarify one of the terms appearing in the law, namely “vehicle.” That is interpretation of the first type.

The second type of interpretation appears when another interpreter tries to infer from this law an additional conclusion, for example: “Vehicles are also forbidden to enter a schoolyard.” There is no doubt that the term “public garden” does not include schoolyards. It is therefore clear that this interpreter is not asking about the meaning or scope of the term “public garden,” as in the previous case. He is engaged in an inference, whose conclusion is a further prohibition not included in the first one, though it emerges from it by analogy, or by a fortiori reasoning, and the like. For example: if the legislator forbids vehicles from entering a public garden out of concern for the lives of the children playing there, then all the more so there is cause for concern about the lives of schoolchildren who play in the schoolyard every day and are present there in large numbers. That is a kal va-homer. Similarly, one might construct an analogy between a schoolyard and a public garden that yields the same conclusion.²⁰

Halbertal points out that everywhere Maimonides speaks about derashot based on the thirteen interpretive rules, he always uses the verb “to extract,” in Arabic tastakhraj, to describe the mode of operation of derash. We saw one example above at the beginning of the second root, and he adduces many others there. He also notes that the language of branches emerging from roots, or principles, recurs several times in these contexts.

In his introduction to the Mishnah, Maimonides explains that disputes sometimes arise in derashot because this is the nature of “inferential rules by way of proofs.” In his Treatise on Logic, Maimonides defines the technical term heqesh, or qiyas, as reasoning by analogy. This is also how the term is understood in the Muslim legal tradition from which Maimonides was influenced. Qiyas is the method for dealing with problems for which no solution is found in the existing law, and therefore one must extend the law by analogy. All this strongly supports the claim that Maimonides understood the interpretive rules as methods of inference and expansion, not as methods of interpretation.

This helps us understand how it can be that an interpretation that is correct and reached by a proper process still does not count as extracting principles already present in the interpreted text. If we say that “motorized bicycle” is a correct interpretation of the term “vehicle,” then the conclusion is undoubtedly embedded in the language of the law itself. But with regard to the inference that vehicles may not enter a schoolyard, which certainly does not interpret terms in the text, it is entirely possible to say that it is a correct and fitting inference, a proper use of the rules of analogy, while still maintaining that the conclusion is not embedded in the language of the law itself. That is expansion.

Let us conclude this section with the same remark we made regarding Rabbi Pinchuk. Halbertal’s proposal gives no explanation at all for Maimonides’ principle that type 1 laws are counted as d’oraita. If derashot are indeed expansive inferences, then it is unclear why the existence of a Sinaitic tradition concerning them makes them laws found in the text itself, especially since a halakha le-Moshe mi-Sinai as such is not found in the text and is therefore, according to Maimonides, also a law “from the words of the Sages.”

Broadening and Generalizing the Picture

In this section we shall propose a broader and more general account than those of Halbertal and Rabbi Pinchuk, one that will allow us to explain the full range of Maimonides’ remarks on these issues.

To understand the principle behind midrashic inference, we must distinguish between three standard forms of inference in logic: deduction, induction, and analogy.²¹

Deduction is reasoning from the general to the particular. For example:

• Major premise: all objects that move on wheels are vehicles.
• Minor premise: bicycles move on wheels.
• Conclusion: bicycles are vehicles.

Induction is reasoning from the particular to the general. For example:

• Premise A: bicycles, which are objects that move on wheels, are vehicles.
• Premise B: motorcycles, which also are objects that move on wheels, are vehicles.
• Conclusion: all objects that move on wheels are vehicles.

Analogy is reasoning from one particular to another, and likewise from one general case to another:

• Premise A: bicycles are vehicles.
• Conclusion: motorcycles are also vehicles, because both move on wheels.

The first form of reasoning, deduction, is necessary. Its conclusion follows necessarily from its premises. This necessity comes from the fact that deduction does not innovate anything. Its conclusion was already hidden within its premises. Whoever knows the two premises implicitly already knows the conclusion. Sometimes a great deal of ingenuity is required to extract that knowledge and formulate it explicitly, as in geometry or other mathematical domains, but in principle nothing new is being added beyond what was already known in the premises. Deductive inference thus analyzes conclusions, but does not add anything beyond what we already know. It only clarifies what is known and what follows from it. Let us call this an analytic inference.

By contrast, analogical and inductive inferences give us new information beyond what was contained in their premises. For precisely that reason they are not necessary inferences. Their conclusions contain some element of speculation, some step beyond what was previously known with certainty, and therefore the conclusion is not necessary. These are synthetic forms of inference, as opposed to analytic ones, in the terminology of the first book.

It is important to note that even synthetic inference is not pure speculation. An analogical or inductive inference possesses some degree of reliability, though unlike deduction its reliability is not complete. Pure speculation is wholly unreliable, and bears no connection to what was known before it was made. As an example of extreme speculation, think of a person with no economic knowledge who decides to invest his money in some way without any inquiry or knowledge about the state of the market and prevailing conditions.²² By contrast, someone familiar with economics chooses investment channels on the basis of various economic considerations. These are not logically certain considerations, but they certainly cannot be dismissed as mere speculation. It is not just a shot in the dark.

Let us now examine the significance of these three forms of reasoning for our discussion. Deduction is always interpretation of the first type. It clarifies what is known, since its conclusion is included in its premises. If someone offered a deductive interpretation of the law cited above, then the conclusion would be only a disclosure of what was latent in the law itself. If the deductive argument described there is correct within the legal system in question, then the conclusion that bicycles are forbidden in a public garden is an interpretive conclusion that was latent in the law itself. This is the kind of inference we earlier described as deciphering or disclosure.

By contrast, induction and analogy are modes of reasoning that expand what is known to us in the premises. If it could be shown that within the legal system in question bicycles are vehicles, and I were to infer from that that motorcycles too are vehicles, and therefore forbidden from entering a public garden, then the conclusion would go beyond what is included in the language of the law itself, that is, in the premises. This is a kind of expansion.

But even expansion, whether analogical or inductive, can occur in two forms. One interpreter may come and argue that although the conclusion does not follow from the wording of the law, it is in fact included in it, since ultimately we have discovered that the term “vehicle” in this law includes motorcycles as well. That is interpretation aimed at clarifying the concepts in the law, though it does so in non-deductive ways. Another interpreter, by contrast, may argue, as in Halbertal’s own example, that vehicles are also forbidden in schoolyards, because there is an analogy or a kal va-homer from public gardens. Here he is not interpreting a term in the law, but deriving conclusions that expand the law. This is therefore interpretation of the second type.

Thus, when we examine and classify the modes of inference that serve us, we cannot necessarily infer from that alone which type of interpretation is involved. A deductive argument is analytic by nature, and therefore inherently functions as interpretation of the first type, that is, deciphering and disclosure. But synthetic tools, that is, expansive and non-analytic ones, can serve in both types of interpretation, the deciphering kind and the expansive kind.

As an example, take Maimonides’ discussion in the Laws of Marriage. We saw there, from several sources, that according to Maimonides betrothal by money is fully effective in the same way as betrothal by document and by intercourse. Its doubtful cases are treated stringently, a woman betrothed this way may eat terumah, one who has relations with her is liable to death, one who rebels against the ruling of the Great Court on such a matter is liable to stoning, and so forth. We suggested there that this can be explained in two ways:

1. Maimonides uses the expression “from the words of the Sages” only regarding the source of the law, not its halakhic force, as the Tashbetz understood him. Therefore its halakhic force is that of a d’oraita law.
2. Maimonides uses that expression with regard to halakhic force as well, as Nahmanides understood him. But if the midrashic consideration merely adds a detail within an enumerated commandment, then that detail has the status of a d’oraita law. For example, if there is a Torah commandment of betrothal, as Maimonides proves in positive commandment 213 from betrothal by intercourse, and we now find two further methods of betrothal, by money and by document, through different interpretive rules, then this move only clarifies the term “betrothal” that appears in the interpreted law, that is, in the d’oraita commandment. This is interpretation of the first type. But if the move creates a new law rather than merely clarifying a detail of an existing law, then we have expansion. The status of such a law will be d’rabbanan. This remains true even if the methods of interpretation are not deductive in logical character.

Thus Maimonides’ treatment of betrothal by money as having d’oraita status is not a proof for the Tashbetz against Nahmanides. It may be true only in a case like betrothal, which is already an enumerated commandment, whether by intercourse or by document. Only in such a situation does the result of derash have d’oraita status. But when an entirely new law is produced through derash, its force will be that of a d’rabbanan law.

This illustrates the importance of distinguishing between two ways in which a midrashic analogy may function. The interpretive rules are analogical, but such analogy can operate in two ways: either as clarification of terms, that is, interpretation of the first type, or as inference of expansion.

A second example is the very derashah Maimonides brings in his first objection to Halakhot Gedolot: “You shall fear the Lord your God” — to include Torah scholars. This is certainly not deduction. Yet one may still understand this inference as interpreting the meaning of the term “fear of God” in the verse, and concluding that it includes His representatives in our world, namely Torah scholars. But one might also understand it as an expansion of the commandment to fear God, that is, as establishing a new law that one must revere anyone connected to Torah, and not only the giver of Torah. On this second understanding, reverence for the Sages is not an expression of reverence for God, but rather an analogical expansion of it. The obligation to revere Torah scholars does not depend on the obligation to fear God, but is merely learned from it by analogy.

With respect to the examples Maimonides himself brings to challenge Halakhot Gedolot — “Honor your father and your mother” — to include one’s father’s wife, that is, a stepmother, and one’s mother’s husband, that is, a stepfather, this has a straightforward halakhic consequence for the question whether the obligation to honor the stepparent continues after the death of the biological parent. If the obligation derives from the obligation to honor the actual parent, so that honoring the father’s wife is itself an expression of honoring the father, then once the father dies there would no longer be such an obligation. But if this obligation is derived by expansion from the obligation to honor the father, there is room, though not a necessity, to say that the obligation remains even after the father’s death.²³

The conclusion is that there is analogical interpretation that expands a given law, that is, the enumerated commandment or the verse explicitly interpreted in the Torah, and there is interpretation that only clarifies the meaning of the given law. All the interpretive rules are forms of analogical interpretation. They do not involve logical deduction in the strict sense.²⁴ Yet there remains latitude in how to characterize their operation: are they deciphering and revealing, analytic, or expansive and creative, synthetic?

How are we to distinguish between these two ways in which the methods of derash operate? What criterion determines when a derashah by means of the thirteen rules is deciphering and revealing, and when it is creative? It seems that the main dispute between Maimonides and Nahmanides in our root lies precisely at this point. Nahmanides holds that any use of the thirteen rules yields a d’oraita law. The reason is that he understands every such inference as revealing. For if God gave us the Written Torah, and also the interpretive rules, then when He wrote the verse He also intended the conclusion exposed by derash. Therefore derash yields d’oraita law.

Maimonides agrees with Nahmanides that derash reveals God’s will, and that God had in mind in advance the conclusions we infer through the interpretive rules. But Maimonides’ claim is that this still does not make the resulting law one that is explicitly written in the Torah. Therefore it is not d’oraita. It is certainly a law that God knows and wants us to fulfill, but He did not write it in the Torah. Midrash expands the verse rather than deciphering it. Hence it belongs to the Oral Torah, not the Written Torah. As we saw, Maimonides calls this a law “from the words of the Sages.”

For illustration, consider again a halakha le-Moshe mi-Sinai. There too it is obvious that God wants us to fulfill it, since He is the one who gave it to Moses our teacher. Why then does Maimonides not classify it as d’oraita? Because in the end it is not explicitly written in the Torah. D’oraita laws are not, as most medieval authorities thought, laws rooted in God’s will; they are laws found in the Written Torah. At bottom there is here a basic dispute about the concept d’oraita. Nahmanides understands it conventionally: a d’oraita law is an expression of God’s will, whereas a d’rabbanan law expresses the will or reasoning of the Sages. But as we have seen, Maimonides identifies the distinction between d’oraita and d’rabbanan with the distinction between Written Torah and Oral Torah. We can now finally explain why type 1 laws are counted as d’oraita.

The Classification of Type 1 Laws — Laws Received by Tradition and Anchored in the Text

As we already mentioned, Nahmanides wonders why Maimonides classifies as d’rabbanan laws either laws received by tradition that lack any anchor in the text, or laws derived by derash for which we have no tradition. But he continues and asks: on Maimonides’ own premises, if each of these two sources, tradition and derash, yields only d’rabbanan law, how does their combination produce d’oraita law? How can the combination of two mechanisms generate something stronger than either one by itself?

We saw above that synthetic inference, whether analogical or inductive, can function both as revealing derash and as expansive derash. According to Nahmanides, all such inferences reveal, namely they reveal God’s will, not necessarily the verse itself; for Nahmanides, d’oraita is not defined by presence in the text but by God’s will. But according to Maimonides, as we saw, there are midrashic inferences of both kinds. The question, then, is how we know when a midrashic inference functions one way and when the other.

Several possible criteria might be used. One might, for example, look at the substantive context and ask whether there is a connection between the law produced by the derash and the law explicitly written in that very verse, in its plain sense. But there are many problematic cases, such as the example we saw of “You shall fear the Lord your God” — to include Torah scholars — where, substantively, the midrash can be read in either direction.

We saw above another possible criterion: when the derash introduces a detail within an existing commandment, it receives d’oraita status, because then it is clarifying a term or another component included in a commandment already enumerated as d’oraita. The author of Lev Sameach develops this at length in his comments on the second root. But this criterion too is not unequivocal, since in different contexts a given innovation can be interpreted in more than one way: as included within some enumerated commandment, or not.

It therefore seems that in cases where these criteria do not clearly decide the matter, Sinaitic tradition comes and tells us that the derash is a deciphering one and not an expansive one. The absence of such a tradition indicates that it is an expansive derash, and therefore the status of the law produced by it is that of a d’rabbanan law.

We thus learn that a supportive derashah is always a deciphering derashah, whereas an expansive derashah is usually creative, unless one of the other criteria is present to show that it is deciphering. As we already noted above, this connection is not logically necessary. There could in principle be a creative supportive derashah that is either deciphering or expansive. But here we see that Maimonides apparently identifies these concepts.

The logic is quite simple. If the law produced by such a derashah were merely an expansion, then there would be no need for tradition at all. Why would God transmit to Moses a law that could already be derived through the thirteen interpretive rules? Maimonides apparently infers from this that when there is a supportive derashah, the tradition is there to teach us that the derash is deciphering rather than expansive, and therefore the law it yields is d’oraita.

If so, in type 1 laws tradition reveals to us that the derash is deciphering, meaning that the law it yields is explicitly written in the text and therefore counts as d’oraita. This is another reason, beyond those we raised above, for the need for derashot that support preexisting laws. According to Maimonides, type 1 laws are produced by deciphering derashot. In that sense they parallel all midrashic laws as Nahmanides understands them, namely as deciphering derashot. The indication that such a tradition exists is a statement of the Sages in the Talmud that the law in question is the very substance of Torah or that it counts as d’oraita. When we find such a source, we know that they had a tradition concerning that law, and therefore it is treated as explicitly written in the verse, making it a d’oraita law.

This seems to be Nahmanides’ intent in the passage quoted above in which he explains his disagreement with Maimonides on this point. Let us cite the relevant line again, from the end of p. 52:

And when we hear them say that it is from the Torah, or that it is the very substance of Torah, we shall count it, because by tradition it is known that the inclusion or the verbal analogy in that verse was written in order to derive from it that learned law.

Nahmanides says here that the existence of tradition reveals that we should regard the verse as written and intended for the sake of that derash, and therefore as effectively written explicitly in the verse.

But this argument still requires further explanation. Where the deciphering derash explains some concept appearing in the verse or in the enumerated commandment, it is clear that one can treat it as though it were explicit in the text. The derash determines the meaning of that word, and therefore the law is indeed written in that very word in Scripture itself. But when the derash does not explain a concrete term in the enumerated commandment, it is not logically clear how the law can be seen as explicitly written in the verse. After all, derash is an analogy; if it is not included in the language of the text as a concept interpreted through it, how can it be treated as written in the verse?

Take again the example: “You shall fear the Lord your God” — to include Torah scholars. We saw that this can be understood in two ways. Either reverence for Torah scholars is itself part of reverence for God, or it is learned by analogy from the obligation to fear God. If it is itself reverence for God, then it is indeed written explicitly in the verse, for the verse is not commanding reverence for God alone, but reverence for everything connected to Torah and to God. Reverence for God is then only one example hinting at the broader principle. If so, reverence for Torah scholars is indeed explicitly written in the text. But if we understand it as an independent duty branching out from the commandment to fear God, then it is an expansive derash and yields a d’rabbanan law.

This is where tradition enters. If we had a tradition that this is a d’oraita law, then it would become clear to us that the verse must be read as a command to revere everything connected to God, and not merely as a command to fear God Himself, who is only one example of the general principle latent in the verse. In that case, reverence for Torah scholars too would count as explicitly written in the verse. The tradition reveals this mode of reading, and thereby turns the law into d’oraita. But according to Maimonides we have no such tradition regarding this commandment, and therefore we must read the derash here as expansive; the commandment to revere Torah scholars is thus d’rabbanan.

It should be noted that ordinarily we do not engage in such interpretation. Halakha rules that we do not derive law from the rationale of a verse. See a detailed discussion of this in the essay on the fifth root. That means that when we see a command in Scripture, we do not interpret it on the basis of its rationale, but on the basis of what the verse says. Therefore, when we see a command to fear God, we do not interpret it as a command to revere everything connected with holiness, because that would be interpretation based on the rationale of the verse. The role of tradition is to tell us that in certain cases the rationale of the verse must be taken into account. In cases where there is such a tradition, the verse is read as though it commanded the rationale and not only the concrete formulation that appears in it. That is why tradition is needed even for laws that have a derash grounding them.

The Relationship Between Analogy and Induction

This point brings us back to the theoretical framework sketched above, namely the three forms of inference.

When we examine analogical inference, it seems at first glance to conceal an inductive inference behind it. For example, from the fact that bicycles are vehicles we infer that motorcycles too are vehicles, because both move on wheels. But ordinarily such an inference presupposes nothing beyond the fact that both move on wheels. If so, in reality there is here an inference valid for every object that moves on wheels, such as roller skates or a scooter. We can now go one step further: in fact this is not really analogy but induction. From our observation of bicycles we infer a general conclusion about all objects that move on wheels. From that general conclusion we then infer the particular conclusion contained in it, namely that a motorcycle, which also moves on wheels, is a vehicle. Thus every analogy from one particular to another is really the unfolding of an induction: first we perform an induction from the source case to the general rule, and then infer deductively from that general rule to the target case.

Take another example. From seeing one raven that is black we infer that another raven is probably black as well. Again, what we really do is infer, tacitly, a general conclusion about all ravens, since nothing unique is known to us about the second raven, and then unfold that conclusion deductively to the second raven. Thus every analogy is based on a hidden induction.

Let us now return to the halakhic discussion about reverence for Torah scholars. From the obligation to fear God we infer by analogy that one should also revere Torah scholars. One can understand this as an inference that begins with a generalization: one must revere anything connected with holiness. From that generalization we then return to the specific conclusion about Torah scholars: if Torah scholars are connected with holiness, and anything connected with holiness must be revered, then Torah scholars must be revered. If every analogy rests on a hidden induction, then the Torah’s command to fear God can indeed be interpreted as a general command to revere everything connected with holiness. But if we understand the command this way, then the command to revere Torah scholars is derived from the general command in an analytic, deductive manner. In that case this is an interpretive-midrashic consideration, and the derash here is deciphering rather than expansive.

By contrast, one can also understand analogy as a direct inference from one particular to another, without the mediation of an inductive generalization in the background. One might even think in the opposite direction: every induction is built up from a collection of analogies. For example, the generalization that all ravens are black is based on a series of analogies to each raven separately. The sum of these analogies creates the generalization. On that picture, the analogy from fear of God to reverence for Torah scholars is likewise direct and is not mediated by an inductive generalization to all things possessing holiness. If we understand midrashic analogy in that way, then it is not plausible to say that the command to revere Torah scholars is included within the command to fear God. On this conception it is clearly an expansive derash and not a deciphering one.

Thus the question whether a given analogical midrash is deciphering or expansive depends on whether we are dealing with direct analogy or with analogy mediated by induction. We can now see that the dispute between Maimonides and Nahmanides turns exactly on this point. Maimonides holds that midrashic expansion takes the form of direct analogy, and therefore it is not deciphering but expansion. Nahmanides, by contrast, understands midrashic analogy as mediated by an inductive generalization, and therefore the midrashic conclusion was implicitly included in the original command.

On this basis, we can continue to understand type 1 laws, those for which we have both tradition and a supportive derashah, as explicitly written in the Torah. If we take the tradition as teaching us about the nature of the midrashic inference grounding the law in question, then tradition tells us that the midrashic analogy is based on hidden induction and is not a direct analogy. Therefore we conclude that the law was included in the command stated in the verse, and hence it is a law found in the Torah and therefore classified by Maimonides as d’oraita.

This is a logical explanation of how analogical inference can reveal something that was hidden in the premises, and not only expand them. Such an explanation is relevant even to midrashic analogy that does not involve explaining a term appearing in the Torah’s command. Even a midrashic inference that yields a new law can be understood in this way, as based on an interpretation of the command in the Torah, as though what binds us were the rationale and not merely the concrete formulation of the command itself.

Let us add here what we saw regarding the final passage of the second root, where Maimonides says that the derashah supporting a law transmitted by tradition is needed in order to display the “wisdom of the text.” We noted there the possibility that his meaning is that the derashah is intended to show us that this law is included in the text itself. Here we may add that this also seems to emerge from Maimonides’ own wording in the introduction to the Commentary on the Mishnah, p. 32 in the Mossad Harav Kook edition, near note 4, and see also the note itself. There it appears that the wisdom of the text is the wisdom of how to decipher the text and see the various laws emerging from it.

This also follows from reason. In creative derashot it is hard to see the wisdom of the text. After all, if the law is a creation, not something that had to satisfy any criterion in the verse, what wisdom of the text is involved? Clearly the wisdom is the wisdom of the text itself, in that it contains multiple layers of meaning simultaneously.²⁵

This explanation of Maimonides’ reference to the “wisdom of the text” also supports the general criterion we have defined here in his view: a d’oraita law is only a law found in the text itself. It therefore follows that derash, in itself, is not found in the text and is therefore not classified as d’oraita. But if the derash is merely supportive, and there is another source for the law that came through tradition, then the derash belongs to the “wisdom of the text”; that is, the law is found in the text, for the logical reason we have seen here, namely that the midrashic analogy is deciphering and not expansive, and therefore it too will be classified as d’oraita.

Let us return for a moment to Rabbi Pinchuk’s metaphor. He compares the two approaches to the thirteen interpretive rules to the work of a sculptor and an archaeologist. Both expose material from an environment that covers it, but there is a great difference between them: the sculptor fashions the form himself, even if he does so by removing part of the material and leaving the remainder. In such a case the material indeed existed beforehand, but the form was made by him out of nothing. It is a new creation. The archaeologist, by contrast, reveals an object whose form already existed. He created nothing, neither the material nor the form.

For our purposes, derash as revealing parallels the work of the archaeologist. Derash as expansive is like giving form to material found in the Torah itself. It is a new creation, even though the material had always existed. When we have a tradition regarding a law, it reveals to us that the use of derashot did not create a new form but disclosed one that already existed.²⁶

To summarize, let us illustrate the distinction just made through the laws governing the modes of betrothal. As noted above, there are three modes of kiddushin: money, document, and intercourse. In positive commandment 213, Maimonides states that only betrothal by intercourse is explicitly written in the text, and therefore it is d’oraita.

By contrast, betrothal by document is learned from the gezerah shavah of “she goes out” and “she becomes” with respect to divorce. Yet, as explained in the responsum, it has an additional source in the text, midrashic and perhaps even a plain-sense hint, as explained in the sugya in Babylonian Talmud, Kiddushin 9a. Thus betrothal by document is a law yielded by derash, and apparently we have no tradition about it, but in place of tradition there is a hint in the plain sense of the verse, which apparently serves the same role as tradition. If so, this law is included in the text, because that hint teaches us that the derash is deciphering rather than expansive. What tradition usually does, the plain-sense hint does here. As we have explained, tradition is not an essential addition in itself, but only an indication of the deciphering nature of the midrashic analogy. Therefore this law too is classified as d’oraita, as a law of type 1, even though we have no tradition about it.

Unlike those two laws, betrothal by money is learned from a derashah: either the gezerah shavah of “taking” and “taking” from the acquisition of a field, or from “there is no money” — this master has no money, but another master does. This is the derashah Maimonides himself cites there. We have already noted that the second derashah is not easy to classify: is it an inclusion, an exclusion, or something else? In any case, it appears to belong to the list of interpretive rules Maimonides is discussing. Maimonides therefore concludes that betrothal by money is a midrashic law, or perhaps a midrashic detail within a d’oraita law, as we noted above, and it has no anchor in the text. Its status is therefore that of divrei sofrim, namely type 3.

In this way, in effect, we close the circle on the discussion of the modes of betrothal. In the end it turns out that all three have sources of different kinds, and therefore their halakhic classification is not entirely identical. As for their halakhic standing and force, see below.

In this fourth gate we shall clarify the classification Maimonides proposes for the different kinds of laws by viewing the dispute between Maimonides and Nahmanides in a broader philosophical context.

A Prohibition Derived by Inference: Halakhic Parallels to Type 1 Laws

Above we cited Maimonides’ words in the introduction to the Book of the Commandments, at the end of the fourteenth root, regarding a prohibition derived by inference. We saw there that according to Maimonides a law produced by means of the thirteen interpretive rules cannot serve as an explicit prohibitory command, and therefore one may not punish someone who transgresses such a law.²⁷ But when Scripture states an explicit punishment for that transgression, then a prohibition derived by inference suffices for us to punish the offender. What is the logic of this distinction? Either a law learned by derash counts as a prohibition written in the Torah or it does not.

In light of our discussion here, Maimonides’ intention appears to be exactly parallel to what we saw regarding type 1 laws. There we saw that if there is a tradition regarding a certain law, the tradition teaches us that the derash yielding that law is a deciphering and analytic derash, that is, a revealing one, not an expansive one. In that case the law counts as written in Scripture itself.

In the previous section we saw a very similar example. Betrothal by document is learned by means of a gezerah shavah. Yet Maimonides classifies it as d’oraita, that is, as a law explicitly written in the text, because of the indirect proof he brings from the laws of the betrothed maiden. A plain-sense hint functions as a substitute for Sinaitic tradition. What is the logic of that? Apparently, here too we have proof that the derash is deciphering rather than creative, because the existence, already at the d’oraita level, of a betrothed maiden who is not married by intercourse shows that betrothal is possible other than by intercourse. Admittedly, this hint is not phrased as a commandment, and therefore by itself it cannot serve as an independent source for the law of betrothal by document. But it shows us that the derash did not create a new law here, but revealed one already present in the text.

Exactly the same logic underlies Maimonides’ innovation here regarding a prohibition derived by inference. If the Torah explicitly writes a punishment, then admittedly there is still no verse phrased as a command or prohibition, but there is a clear plain-sense indication that such a law exists at the d’oraita level. Therefore, once we find a derashah supporting the law in question, that hint reveals that the derash is revealing rather than creative. Consequently, the law learned from the derash is found in the text, and we indeed have a d’oraita commandment. Just as such a law may be counted in the count of the commandments, so too it may serve as a prohibition that grounds punishment.

These are two halakhic parallels in which we encounter the same logic that we saw in type 1 laws. In all these cases there is ancillary evidence that the law learned by derash was present in the text. Admittedly, without explicit prohibitory wording, such as “take care,” “lest,” or “do not,” this cannot be treated as an independent commandment. But that is only a linguistic deficiency. The derash fills in that deficiency.

Note 3: Maimonides on Gezerah Shavah as Establishing a Prohibition

In Mishnah Yoma 8:1 it is explained that one who eats and drinks in a single lapse of awareness is liable to bring only one sin-offering, whereas one who eats and performs labor on Yom Kippur in a single lapse of awareness is liable to bring two sin-offerings. Maimonides, in Mishneh Torah, Laws of Inadvertent Transgressions 6:5, brings the rule that one who eats and drinks is liable to one sin-offering, but he does not bring the second rule.

Perhaps this may be explained in light of Babylonian Talmud, Keritot 3b, which states that if there are two prohibitions learned by gezerah shavah, for example slaughtering and offering up, they count as two prohibitions. Maimonides does not codify this as law, and perhaps he holds that there is only one prohibition here.

Indeed, in two places we find that Abaye holds that what is learned by gezerah shavah is considered one prohibition with the source from which it is learned:

1. Babylonian Talmud, Sanhedrin 54b, in Rashi, s.v. “ela mena lan”; and Maimonides rules likewise in Mishneh Torah, Laws of Inadvertent Transgressions 5:3. In Ayin Mishpat there the reference is made to Rabbi Akiva, though according to our approach it may also fit Rabbi Ishmael. See also Hazon Ish, Keritot, siman 37, sec. 5.
2. Babylonian Talmud, Sanhedrin 34b, regarding slaughtering and sprinkling, and see Rashi there. Admittedly, in the case of slaughtering and sprinkling these are two matters both learned by hekesh from offering up, and not one from the other, so that case could be rejected. But Ran, in his novellae there, writes that the same would hold for sprinkling and offering up. See also the Maharsha there.

With Maimonides, however, one may ask whether the fact that he does not count two prohibitions is because they are regarded as one, or because what is learned by gezerah shavah is not d’oraita at all, and therefore only the written prohibition exists, not the derived one. The practical difference concerns source 2 above, where the issue is two gezerot shavot and not a verse and a gezerah shavah.

From Maimonides’ omission of the rule concerning one who ate and performed labor, it would seem that he held that the prohibition corresponding to the negative commandment of affliction is derived by gezerah shavah from labor, and therefore he does not count them as two. On the other hand, in the count of the commandments Maimonides does count those two prohibitions separately, which would seem to prove that they are not learned from one another, nor learned by gezerah shavah.

But above we saw that when an explicit punishment is written, even a prohibition established by gezerah shavah can be counted among the commandments. It may nevertheless be that for purposes of incurring two distinct punishments it still counts as one, since in the end it emerges through gezerah shavah from labor. If so, there is a prohibition here, but no separate punishment, and that is a consequence of the fact that the prohibition comes by derash.

From here we learn that Maimonides treats all the interpretive rules as cases of a prohibition derived by inference. We also learn that when there is a prohibition derived by inference together with an explicit punishment, that suffices.²⁸

The Sorites Paradox: What Is the Difference Between Being “Very Much Included” in the Torah and Being “Included Enough” in the Torah?

Until now we have treated the question from the logical side, distinguishing between deduction, which reveals what is latent in existing knowledge, and synthetic modes of inference, which expand what we knew until now. We saw that synthetic modes can sometimes also function as forms of disclosure.

The questions raised here can also be examined conceptually. The problem before us is this: how can Maimonides say that the halakhic conclusions yielded by the interpretive rules given at Sinai, when applied to the Bible given at Sinai, produce d’rabbanan laws? Put differently: how can these laws fail to be implicitly included in the Written Torah, and yet still emerge from it? This is a question about the concept of branching that has repeatedly appeared in our discussion.

Let us now reflect on the concept “included,” which we used in presenting the problem in the last paragraph. What does it mean for a proposition to be “included” in a text? Is every proposition either included or not included? Maimonides seems to say that a halakhic proposition that branches out from a creative derashah is included in the text, but not completely. At first glance, such an approach seems to violate simple logic, which tells us clearly that a proposition is either included in a text or it is not. But that is a mistake, and we shall try to explain it briefly here.²⁹

There is a well-known phenomenon according to which many everyday concepts do not conform to these logical rules. It was already presented by the Greeks, originally with respect to the concept “bald.” In contemporary philosophy it is usually called the Sorites paradox.³⁰ Let us formulate the common version:

• Premise A: one grain of gravel is not a heap.
• Premise B: if a collection of gravel is not a heap, then adding one stone will not change its status; it is still not a heap.
• Premise C: a collection of one million stones is a heap.

It is easy to see that although each of these three premises, taken separately, sounds highly plausible, all three together are incompatible. They cannot all be accepted at once. So what has gone wrong? Which of them is mistaken, and why?

Before answering, let us note that almost all everyday concepts are vulnerable to the same sort of attack. Take the concept “afternoon,” a formulation that emerged from a conversation with my son:

• Premise A: 12:00 is not afternoon.
• Premise B: if a given moment is not afternoon, adding one second will not change that.
• Premise C: 16:00 is afternoon.

Again, each of these claims sounds plausible on its own, but the three together are not compatible.

The same holds for the question what counts as “red,” where the steps of progression are tiny wavelength increments; or what counts as an “island,” as against mainland or a rock in the sea; or what counts as a “table,” where there is a continuous metamorphosis, whether functional or formal, from chair to table; and so on.

The root of all these perplexities is one and the same: everyday concepts do not have sharply bounded meanings. The concept “heap” is not binary. A collection of stones is not simply either a heap or not a heap. It may be not a heap, or somewhat a heap, or very much a heap, or wholly a heap. More generally, the truth-values relevant to everyday concepts are not just the pair true and false, but a continuum of values.³¹

We can now return to the question of midrashic laws. We asked how Maimonides can treat them as included in the Torah, thereby distinguishing them from rabbinic enactments, that is, can treat them as a kind of interpretation rather than legislation, and at the same time say that they are not included in the Torah itself, thereby distinguishing them from d’oraita laws, that is, treat them as legislation rather than interpretation.

Our answer is that the concept of being “included,” or “present,” in the Torah is not sharp. It is an everyday concept, and therefore has an inherent vagueness. A law can be entirely included in the Torah, very much included in the Torah, sufficiently included in the Torah, only slightly included in the Torah — which is probably the Ritva’s concept of asmakhta in Rosh HaShanah 16a, which we already encountered in the first unit — or not included in the Torah at all. Laws truly included in the Torah, that is, written in the verses, are d’oraita laws. Laws very much included in the Torah, that is, received by tradition and possessing a midrashic anchor in the verses, and therefore hidden in the Torah itself, are also d’oraita laws. Laws sufficiently included in the Torah, that is, laws with some connection to the biblical text, unlike enactments and decrees, are divrei sofrim, in the sense we defined above; Maimonides classifies them as a separate unit, even though he does not use exactly that label. And laws not included in the Torah at all, or included only slightly, that is, having merely an asmakhta, are d’rabbanan laws.

It should be noted that halakhot le-Moshe mi-Sinai are exceptional, since they are not included in the Torah at all, yet they occupy an intermediate status: they are divrei sofrim, in our terminology, and not d’rabbanan laws. This is apparently an expression of their force, not of their classification. This point is connected to the question we shall discuss below: whether, according to Maimonides, there exists an independent plane of halakhic force, separate from the discussion of the sources of halakha.

Here we encounter again the concept of branching that we have already met several times. In the first unit we distinguished between three concepts lying on the same axis: unfolding, branching, and asmakhta. Here we see that saying a law unfolds from the text means that it is truly included in the text, for example, though not necessarily, in the sense that it follows by deduction. When we say that a law branches out from Scripture, we mean that it is sufficiently included in it, but not wholly included as in unfolding. A third category of relationship to Scripture is when a law is merely supported by the text, that is, a d’rabbanan law with an asmakhta in the text. Such terminology indicates that the law is only slightly included in Scripture, if at all. As noted, there are also laws not included at all, and laws explicitly written in Scripture. But in the previous unit we dealt only with laws that have an incomplete relation to the text, that is, laws which in Maimonides’ definition belong to the Oral Torah. Among these we found only three types.

In summary, we can say in this language that Maimonides’ innovation operates on two planes:

1. He defines the concept d’oraita as what is found in, or included in, the written text.
2. He introduces a conceptual innovation according to which the term “included” takes on a range of values, and therefore there are degrees of inclusion in the text.

Returning to the Theory of Interpretation

Let us add one further remark. We noted above that Maimonides and Nahmanides disagree over whether a single text can bear several interpretations. Nahmanides sees no problem in that; Maimonides holds that such a situation is impossible. In light of what we have said here, there is room to view this dispute somewhat differently. Maimonides explains the concept of interpretation in a very flexible way, recognizing several degrees of interpretation. The interpretation of a text, in the strict sense, is only what is explicitly written in it, namely the plain sense. There are, however, many additional quasi-interpretations that bear some sort of relation, of varying strength, to the text; but these are lower degrees of interpretation, or of connection to the text. Nahmanides, by contrast, holds that several interpretations of a single text are possible, but he does not explicitly distinguish among different levels of interpretation.

In light of what we have said here, it is reasonable to say that the interpretive theories of Maimonides and Nahmanides are not really so far apart. It may be that both agree there can be several different interpretations of a single text, but that they stand at different levels of connection to the text. The difference is that Nahmanides sees no difference in status among them, since all are interpretations, and all were included in God’s intention when He wrote the text. Therefore they all have the halakhic status of d’oraita. Nahmanides is here assuming, implicitly, that d’oraita means what follows directly from God’s will and intention. That is why even halakhot le-Moshe mi-Sinai, which do not appear in the Torah at all, are d’oraita. Maimonides, however, disagrees. For him the concept d’oraita does not express God’s will, as opposed to the will of the Sages, but rather inclusion or presence in the written text. Therefore not every such interpretation counts as truly included in the text. True, they all express God’s will in one way or another, but the degree of their inclusion in the text varies. Therefore some count as d’oraita laws, while others are divrei sofrim or d’rabbanan. The logic behind the criterion of degree of inclusion in the text will be clarified below.

Implication: the Uniqueness of the Thirteen Interpretive Rules and of Inclusion — Returning to Maimonides’ Title

This point brings us back to the title of Maimonides’ second root, where he defines the discussion as concerning laws produced by the thirteen interpretive rules and by inclusion. At the beginning of our discussion we suggested several possibilities for understanding exactly what he means. The simplest explanation, also given in Lehem Mishneh on the Laws of Marriage 4:6, is that Maimonides means just that: the thirteen interpretive rules and inclusion. Against that interpretation we asked: why does Maimonides single out precisely those methods of derivation? What distinguishes them from other methods of derash? Why are laws derived by other methods not included in the discussion of this root, and in which direction does that cut? Are they certainly d’rabbanan or certainly d’oraita?

In fact, a very similar question could already be asked of Rabbi Ishmael himself, who includes in his baraita only the thirteen rules. Why does he not include all the other rules as well? The obvious answer is that the other rules belong to the school of Rabbi Akiva, and Rabbi Ishmael indeed disputes them and does not use them. And in fact the main additional rule, inclusion, is what characterizes Rabbi Akiva’s method of halakhic derash, as Babylonian Talmud, Shevuot 26a, states: Rabbi Akiva expounds by inclusions and exclusions, whereas Rabbi Ishmael expounds by generals and particulars.

But we saw that this cannot be said, at least not with respect to some derivational methods. It is quite clear that Rabbi Ishmael also accepts the validity of additional interpretive rules, and nevertheless he does not list them in his baraita. We mentioned that the author of Sefer Keritut already raises this difficulty with respect to the baraita of thirty-two rules, and offers several explanations. The main one is that with the other rules, at least those belonging to halakhic derash, the laws produced by them are considered explicitly written in the text. Rabbi Ishmael included in his baraita only derivations whose results are not explicitly present in the text.

We explained this by saying that those other methods of halakhic derash are based on mufneh words in the verses, that is, words that have no plain-sense role. In such a case, the midrashic interpretation is itself the plain sense of the verse; it is precisely the meaning of the otherwise redundant word. The fact that the word is mufneh means that it has no other meaning besides the derash. If so, the law produced by such a derivational method is explicit in the text, because the midrash reveals that it is the legitimate interpretation, indeed the only one for Maimonides, of that word.

In light of our explanation of Maimonides’ view, this picture emerges naturally and clearly. First, Maimonides includes the method of inclusion because he rules in accordance with Rabbi Akiva and not only with the baraita of Rabbi Ishmael. See our discussion in the second gate on the historical development of the methods of derash. Second, as we have seen, Maimonides’ whole discussion turns on the question whether the midrashic law is found in the text or not. Maimonides’ central innovation is that these laws do have some connection to the text, but are not fully present in it; rather, they branch out from it. They are present enough in it, and are therefore divrei sofrim. According to this, it is obvious that if there are other methods of derash whose results count as truly present in the text, even if the text chose to indicate them only indirectly, through derash, but the verse has no simpler plain sense, then they are not relevant to the discussion of this root. They are certainly d’oraita laws, not divrei sofrim.

Thus the picture described so far, together with the assumption of the author of Sefer Keritut, naturally yields the conclusion that the subject of discussion in this root is only the fourteen rules listed in the title of the root. Laws yielded by the other rules are laws found in the text, and therefore their status is d’oraita. We need only add that the rule of inclusion, which Rabbi Ishmael does not accept at all in the form used by the school of Rabbi Akiva, though Rabbi Ishmael certainly has his own use of inclusion, is probably treated by Rabbi Akiva as something not explicitly written in the text, for the same reason Rabbi Ishmael does not accept it. Even if the letter or word, such as et in the derashot about reverence for Torah scholars or honoring an older brother, cited by Maimonides in this root, is mufneh, there is still no truly superfluous word here, and therefore this is not treated as plain-sense interpretation. If there were laws derived from inclusions involving a fully mufneh word that is not just a single letter or a connective particle such as et, then perhaps Rabbi Ishmael and Rabbi Akiva alike, and Maimonides following them, would regard those too as d’oraita laws.

It should be noted that these considerations provide a basis for resolving a number of difficulties in Maimonides’ view, both from various sugyot in the Talmuds and from internal tensions in his writings, especially concerning different rulings in the Mishneh Torah. There will be places where the Talmud, or Maimonides himself, treats midrashic laws as d’oraita. But this will not contradict the principle of the second root, because the derivations in question are not by way of the thirteen interpretive rules or inclusion. According to this, when Maimonides writes in the responsum that there are only “some three or four places” where a law received by tradition has a midrashic anchor and is therefore treated as d’oraita, he means only a midrashic anchor via the thirteen rules or inclusion, not any derash whatsoever. In other kinds of derash there are additional cases that yield d’oraita laws, and for that no tradition is needed at all. Even without a tradition accompanying the derash, the laws produced by those methods are d’oraita.³²

Halakhic Implication

Until now, however, we have not dealt at all with the question of halakhic force, but only with the meaning of the concepts d’oraita and d’rabbanan in relation to the source of the laws under discussion. In the next chapter we shall turn to the halakhic force of the different types of law. We have already seen that this question does not necessarily depend on source. The Tashbetz and those who follow him hold that, despite Maimonides’ conceptual division concerning the sources of the various laws, the status of all these types of law is that of d’oraita law.

There is, however, one rare and interesting example that reflects the meta-halakhic distinction between d’oraita and d’rabbanan with respect to the source of laws, and also displays one halakhic consequence of that distinction. It does not address the question whether the status of midrashic laws is that of d’oraita laws or not, but rather a halakhic consequence of the meta-halakhic fact that such laws are not found in the text itself.

Note 4: An Example of a Halakhic Consequence of the Meta-Halakhic Plane — Priestly Garments³³

In the portion of Pekudei there are several verses describing the making of some of the priestly garments, Exodus 39:27-29:

They made the tunics of fine linen, the work of a weaver, for Aaron and his sons; and the turban of fine linen, and the splendid caps of fine linen, and the linen breeches of twisted fine linen; and the sash of twisted fine linen, blue, purple, and crimson, the work of an embroiderer, as the Lord commanded Moses.

Mentioned here are the tunic, the turban, the caps, the breeches, and the sash. These are the white garments of the High Priest, which he wears on Yom Kippur, and also the garments of an ordinary priest. Three of them — the tunic, the breeches, and the sash — are identical for the High Priest and the ordinary priest. On the head, however, there is a difference: the High Priest wears the turban, while ordinary priests wear caps.

All the garments described in these verses are made of fine linen, and the Torah takes the trouble to repeat that requirement separately for each garment. The sugya in Babylonian Talmud, Yoma 71b, is troubled by this point: why did the Torah not simply say once, in general, that everything must be made of fine linen? The Gemara concludes that the repetition of the word “fine linen” comes to teach several laws concerning the priestly garments:

The Sages taught: In garments concerning which “fine linen” is stated, the thread must be sixfold. If “twined” is stated, it is eightfold. The robe is twelvefold. The curtain is twenty-fourfold. The breastplate and the ephod are twenty-eightfold.

From where do we know that their thread must be sixfold? Because Scripture says: “They made the tunic of fine linen… and the turban of fine linen, and the splendid caps of fine linen, and the linen breeches of twisted fine linen.” Five scriptural occurrences are written there: one for itself, to teach that they must be of linen; one to teach that their thread must be sixfold; one to teach that they must be twined; one for the other garments regarding which “fine linen” is not stated; and one to make these rules indispensable…

The Gemara thus determines that the repetition teaches several laws:

1. They must be made of fine linen.
2. The flax thread from which they are made must be spun from six strands.
3. The threads must be twisted together.
4. The other garments, in which the word “fine linen” is not stated, are also subject to these rules.
5. These rules are indispensable.³⁴

Now, in the parallel sugya in Babylonian Talmud, Zevahim 18b, there is a different derashah deriving five laws from garments concerning which the word “linen” is written:

The Sages taught: “Linen” teaches that they must be of flax; “linen” teaches that they must be new; “linen” teaches that they must be twisted; “linen” teaches that their thread must be sixfold; “linen” teaches that one may not wear ordinary garments together with them.

Abaye said to Rav Yosef: Granted, that they must be of flax — this teaches us that flax, yes, but not anything else. But if “linen” teaches that they must be new, does that mean new, yes, worn, no? But was it not taught that worn garments are valid?

He said to him: By your reasoning, if “linen” teaches that their thread must be sixfold, does not “linen” imply each strand separately? Rather, this is what it means: garments regarding which “linen” is written must be of flax, new, twisted, and their thread sixfold; some of these rules are for the optimal manner of observance, and some are indispensable.

The Gemara learns here that such garments must satisfy the following conditions:

1. They must be made of flax.
2. They must be new, as the flax is in its original condition in the stalk where it grew.
3. They must be twisted.
4. They must be sixfold.
5. One may not wear ordinary garments together with them.

In the sugya in Zevahim, Rashi explains that the meaning of the word “linen” is flax. That is, this is learned from the meaning of the word itself. As for the other requirements, it does not seem that they arise from the plain meaning of the word. Some even contradict its plain meaning. Indeed, Abaye objects to the requirement that the garments be new on the basis of a baraita stating that, after the fact, worn garments are valid. Rav Yosef replies that by the same token it is also difficult that the requirement of sixfold spinning contradicts the meaning of the verse, since “linen” implies something single rather than multiple. From the Gemara’s wording it is not clear why that is relevant to Abaye’s objection. Does Rav Yosef have any better explanation for how one derives from this verse a law that contradicts its plain wording? Why does he connect this problem to Abaye’s objection based on the baraita?

Rashi addresses this in his commentary and explains that Abaye’s real objection was from the plain sense of the verse. Abaye’s claim is that the verse does not imply that the garment must be new, and the baraita is only brought as proof, since it shows that this rule is not indispensable. Rav Yosef answers that the requirement of sixfold spinning also does not fit the plain meaning of the verse, and on that point Abaye too certainly agrees that this is the law, and that it is indeed indispensable. Its source is the sugya in Yoma cited above, as Rashi explicitly notes, and as we indeed saw there: even garments regarding which the word “fine linen” is not stated, but only the word “linen,” require sixfold spinning, and these rules are indispensable. This point sheds light on the relation between plain sense and derash; see Appendix 6 at the end of the book.

The Gemara concludes that among the rules stated about these garments, some are indispensable and some apply only ideally. Rashi goes on to determine the criterion distinguishing the two sorts of rules: those that are not indispensable are the rules yielded by derash, namely that the garments should be new. The indispensable rules are those that arise from the meaning of the verse itself, namely that they must be made of flax, spun sixfold, and twisted.

According to Rashi, only rules that arise from the meaning of the verse itself are indispensable. It is nevertheless hard to infer from here that in every case of a rule explicitly stated in Scripture we would not need the principle that Scripture repeats or varies the matter in order to make it indispensable. The reason these rules are indispensable is apparently the derashah in the sugya in Yoma, which teaches that they are indispensable. By contrast, rules yielded by derash are not indispensable. According to Rashi, then, the two categories are:

A. Indispensable, because they arise from the meaning of the verse:
1. They must be made of flax.
2. They must be spun from six strands.
3. They must be twisted.

B. Not indispensable, because they arise from derash:
1. They should be new.

As for wearing ordinary garments together with them, it is not clear from Rashi where that belongs: does it arise from meaning or from derash, and is it therefore indispensable or not?[^^230] And with regard to the sugya in Yoma, Rashi understands that all the derivations there arise from the meaning of the verse and not from derash, and are therefore indispensable. Only the rule that they be new does not appear there, and that indeed is not indispensable.

This requires some explanation. In the sugya in Yoma, the Gemara learns these laws through derash. Moreover, the very fact that they are indispensable is learned through derash and not from the meaning of the verse. If there were a general rule that laws learned from the meaning of the verse are indispensable, then why would we need a derashah from the word “fine linen” to teach it? Nor does this fit the well-known principle in the realm of sacrificial law that every rule, even one written explicitly in the verse, is not indispensable unless Scripture repeated or altered the verse specifically to make it so.

It therefore seems that Rashi does not mean that the rule requiring Scriptural variation in order to make something indispensable applies only to laws learned by derash. Rather, his meaning seems to be this: the sugya in Yoma contains a derashah teaching that the rules appearing there are indispensable. But it is not clear which of them are indispensable. All of them? And what about the rules for garments concerning which the word “linen” is said rather than “fine linen”? Are the three rules learned there also indispensable in those cases?

Rashi apparently determines that the rules appearing in that verse are indispensable because the derivation making them indispensable also appears in that verse. By contrast, the rule learned only from the word “linen,” namely that the garments must be new, is not found in that verse, since it is not learned from the word “fine linen” but from the word “linen.” Therefore the rule that it is indispensable does not apply there. The reason is not that it is learned by derash, but that it is learned by derash in another place.³⁵

Maimonides, however, in Mishneh Torah, chapter 8, halakha 14, seems not to read the sugya in Zevahim as Rashi does. He writes there:

Wherever the Torah says “fine linen” or “twisted fine linen,” the thread must be sixfold; and wherever it says “linen,” a single strand is valid, though the best way to fulfill the commandment is that it be sixfold…

Maimonides rules there that in places where “fine linen” or “twisted fine linen” is written, the law that the thread be spun from six strands is indispensable, whereas in places where only “linen” is written, this is only an ideal requirement. That is, Maimonides distinguishes between places where “linen” is said and places where “fine linen” is said. He does not seem to divide, as Rashi does, the rules learned from the word “linen” into two categories.

Maimonides’ commentators discuss this disagreement and understand that he read the sugya in Zevahim differently from Rashi. The Kesef Mishneh on that halakha suggests that Maimonides simply divides the laws derived in the sugya in Zevahim differently: what is indispensable is that there must be one kind of thread, apparently meaning one material, namely flax, without blue, purple, or crimson. The rule of sixfold spinning is not indispensable. The Kesef Mishneh himself notes that this is very strained in the language of the Gemara. It is also unclear how this would relate to the distinction between meaning and derash, and the whole exchange between Abaye and Rav Yosef, which takes for granted that sixfold spinning is indeed indispensable, becomes obscure.

Rabbi Yosef Korkus, on that passage, offers another explanation of the sugya in Zevahim. According to him, Rav Yosef answers Abaye that in truth the rule of spinning is not indispensable after the fact, just like the rule that the garments be new. And all this is only where the Torah says “linen”; where it does not say “linen,” the rule is indispensable, as in the sugya in Yoma. In that reading, “some are indispensable” refers to cases where “fine linen” is written, and “some are not indispensable” refers to cases where “linen” is written. According to this, wherever only “linen” is written, none of the rules are indispensable, unlike Rashi, who divides them into two categories. Even so, the flow of the Gemara remains unclear, especially the connection to Abaye’s question and the relation between meaning and derash.³⁶

Despite these interpretations, the more plausible explanation of Maimonides’ view seems to be that he in fact understood the sugya in Zevahim as Rashi did, but considered it to contradict the sugya in Yoma, and therefore ruled in accordance with Yoma. A proof of this is that Maimonides does not codify at all the rule that one may not wear ordinary garments together with them, even though that is one of the rules learned only in the sugya in Zevahim. By contrast, in chapter 8, halakha 4, he does codify the other rule learned only in the sugya in Zevahim, namely that the garments should be new, and the commentators there also infer from his wording that he rules that this is only an ideal requirement and not indispensable. But he finds the source of that law in another verse, not in the word “linen” as it appears in the sugya in Zevahim:

The priestly garments should ideally be new, attractive, and triple-folded, in the manner of the garments of dignitaries, as it is said: “for glory and for splendor”…

From this emerges a general picture according to which Maimonides did not codify the sugya in Zevahim at all. The laws of twisting and of sixfold spinning derive, for him, from the sugya in Yoma, which, as noted, he does codify. We must now ask, however, how Maimonides arrived at the result that in places where the Torah writes only “linen” rather than “fine linen,” the law of sixfold spinning is not indispensable. In the sugya in Yoma it says explicitly that the rules learned there apply even to garments regarding which “fine linen” is not stated, that is, garments described only as “linen,” and it also emerges there that these rules are indispensable. Moreover, it is not clear why Maimonides concludes that there is any contradiction at all between the sugyot.

As we have seen throughout the book, Maimonides’ view is that derash is not an interpretation of the text but an expansion of it. A law learned through derash cannot count as an interpretation of the text, because for him only the plain sense is the one valid interpretation. According to Maimonides, the law yielded by derash is not uncovered from the text but built upon it, like a branch emerging from a root. We can now try to apply this principled rule to Maimonides’ understanding of the priestly garments.

As noted, in the discussion in tractate Yoma we extend the law to the other garments, in which the word “fine linen” is not written, from one of the superfluous occurrences of the word “fine linen” in our passage. According to our explanation here of Maimonides’ view, this extension does not uncover a halakha (Jewish law) that was already present in the verses, but constructs a new halakha that expands what is present in the verses.³⁷

Immediately afterward, we derive from another superfluous word that all the laws found here are indispensable to validity. To what does this rule apply? In the simple sense, to all the laws found in the verse from which it is derived. The derivation teaches us that all the laws present in that verse are indispensable. But if we derive that everything stated in that verse is indispensable, this will apply only to the laws present in the verse. What, then, about the laws concerning the garments with respect to which the word “fine linen” is not written? Are they too present in these verses? According to Maimonides, it seems that they definitely are not. The laws regarding such garments are learned through inclusion, and as we have seen, exposition, and especially inclusion, constitutes an expansion of the text and not its decipherment. If so, the laws concerning garments for which the word “fine linen” is not written are not present in these verses. Therefore, the derivation that newly teaches that the laws in these verses are indispensable does not apply to the laws concerning the “linen” garments, but only to the laws concerning the “fine linen” garments.

At this point Maimonides’ view becomes entirely clear. For the garments in which the word “fine linen” is written, all the laws learned with respect to them fall under the principle that they are indispensable. But for the garments in which Scripture says “linen” and not “fine linen,” those same laws do indeed apply by virtue of the inclusion, yet they are not indispensable.

This is also the contradiction that Maimonides finds between the two discussions. From the discussion in Yoma, according to his view, it follows that these laws are indispensable with respect to the white garments, but not with respect to the other garments. By contrast, from the discussion in Zevahim it follows that some of them are indispensable even for the other garments. Maimonides understood the discussion there like Rashi, since that is the straightforward reading, as later authorities wrote.

As stated, this is a unique example in which Maimonides’ interpretive theory yields a halakhic consequence. The fact that laws learned through inclusion are not present in the verses from which they are learned leads us to conclude that an inclusion that applies to what is present in those verses will not apply to them.

This point brings us back once again to Rashi’s position. As we concluded above, according to Rashi the distinction is not between laws learned from plain meaning and laws learned from exposition. Rashi’s point is that laws learned from elsewhere, and not from our verses, are not subject to the rule learned from these verses that they are indispensable. But laws learned through inclusion from our verses certainly are indispensable, such as spinning from six strands in garments for which the word “fine linen” is not written. Thus Rashi understands inclusion as revealing rather than expanding, for in his view the derivation that a given law is indispensable applies even to laws that emerge from scriptural inclusion. For him, laws that emerge from inclusion are considered present in the text itself, as Nahmanides held.

We will conclude this note with a brief look at the hermeneutic measure of inclusion. We saw that at the beginning of the second principle Maimonides rules that laws derived through exposition are not to be counted, and he specifies: the thirteen hermeneutic rules and inclusion. The other interpretive rules apparently are not included in the discussion of the second principle. Our conclusion here is that Maimonides includes inclusion within the system of hermeneutic rules, because he sees inclusion as a rule that expands what is stated in the verse, and not as a rule that reveals it. This stands in contrast to the other methods of exposition that are not included in the baraita of Rabbi Ishmael.³⁸

As is known, in the laws of sacrificial matters one often does not derive from what has itself already been derived. See the discussion in Babylonian Talmud, Zevahim 50a and the surrounding passage. That is, one does not apply one of the thirteen hermeneutic rules to the result of a prior use of one of those rules, though there are quite a few exceptional cases there as well.

In practice, the rule we have reached here is something like the principle that one does not derive from what has itself been derived. We saw that one cannot apply an inclusion to the result of another inclusion. In essence, this is the prohibition against deriving from what is already derived in sacrificial law, applied to inclusions. One should remember that in the discussion in Zevahim the principle of deriving from what is derived is stated only with respect to a small subset of the thirteen hermeneutic rules, whereas here we see something similar even with regard to inclusion. This phenomenon may be connected to what we saw above, that inclusion has a character similar to that of the thirteen hermeneutic rules, since, exactly like them, it is not regarded as explicitly present in the verse.

For that reason it is also reasonable to assume that the rule that one does not derive from what is already derived applies to such a rule. Had this been a revealing rule, there would seem to be no reason why we could not apply one revealing rule to the result of another. All of these layers are present in the text, one inside another. According to our approach, the rule that one does not derive from what is already derived is said only of rules that do not reveal what is present in the text, but rather expand the text. It imposes a limitation on the amount of expansion. How far are we permitted to extend beyond what is in the text itself?³⁹

We have learned from here something about the nature of the rule that one does not derive from what is already derived, and also about the nature of inclusion. Since inclusion, like the thirteen hermeneutic rules, expands the text, at least according to Maimonides, and since the rule that one does not derive from what is already derived is stated only with respect to expanding rules, it is clear that it applies both to the thirteen hermeneutic rules and to inclusion. The same is true of the principle in Maimonides’ second principle, which, as we saw, is also stated only with respect to the fourteen expansive modes of exposition: the thirteen hermeneutic rules and inclusion.⁴⁰

The conclusion of this note is that if indeed the laws learned from expositions or inclusion are not present in the text, then if there is an exposition that introduces something about what is present in a given verse, it cannot be applied to what emerges from that verse by virtue of exposition. For example, if we were to learn by inclusion some law about everything present in the verse “You shall fear the Lord your God,” we could not apply it to reverence for Torah scholars. The reason is that this obligation emerges from exposition, from inclusion through the word “et,” and therefore, according to Maimonides, it is not present in the text itself. As stated, this is a general principle, and in this note we saw one specific example of it.

Legislation and Interpretation

We now return to the question of the relationship between legislation and interpretation. First, we should say that this question does not directly concern the plane of legal force, except indirectly. It belongs directly to the discussion on the conceptual plane, with which we are now occupied.

According to our approach, the laws that emerge from expositions are a kind of activity of the sages that cannot be defined either as legislation or as interpretation. This is an independent category, a third type. We explained why it is not interpretation: it does not reveal something latent in the text. But it is important to understand that it is not legislation either. We explained this above, at the beginning of chapter 1 of this gate, and now we can sharpen the point further.

In legislation, the basic motivation is need, together with the sages’ analysis of reality. By contrast, in expositions the basic motivation is some word or formulation in the Torah. For example, “You shall fear the Lord your God” is a verse from which, according to the methods of exposition, something must be included from the word “et.” The sages then try to determine what should be included here. Their reasoning tells them that it is plausible to include Torah scholars. But even if they saw no real need to institute reverence for Torah scholars, they would still be compelled to do so, because the Torah here contains a word that demands exposition. In other words, the basic motivation of exposition is the Torah, not need, and not the solution of some problem. Even if there is no problem at all, exposition must still be undertaken in those places where the hermeneutic methods require it.

There is room to discuss, with regard to expositions in general, whether the rules truly always dictate the need for exposition. But there are certain cases in which this is clearly so. For example, a free term from which one constructs a gezerah shavah (verbal analogy): there we have no option of ignoring it, and therefore, having no choice, we must expound it, even if there is no need or external consideration demanding the innovation of any law at all.

This is the basic difference between enactments and laws derived through exposition, even according to Maimonides, who holds that laws derived through exposition are rabbinic laws. We will now illustrate the point with one example from the book Halakhot Pesukot, by one of the Geonim. See also Appendix C at the end of the book.

From the discussion of the author of Halakhot Pesukot concerning betrothal through disqualified witnesses, it emerges that if the witnesses’ disqualification is rabbinic, then the sages nullify the betrothal, because “whoever betroths, betroths subject to the view of the sages.” But if the witnesses are disqualified by virtue of exposition, for example witnesses who are relatives through the mother’s family, then the betrothal remains valid. See Neubauer, p. 11, who infers this and changes the text on that basis. See also below, in Appendix C.

We see from his words that laws emerging from exposition are rabbinic laws, and that issue will be discussed in the appendix. But here we touch on the distinction he makes between enactments and laws derived through exposition. Precisely because he adopts Maimonides’ view that laws derived through exposition are rabbinic laws, he can help us understand the difference between these two categories in Maimonides’ thought. This distinction in Halakhot Pesukot is based on the reasoning we have developed here. The man who betroths does so on condition that this is acceptable to the sages, that is, with the sages’ approval.⁴¹ But expositions are not “the sages’ approval.” Rather, they are an act rooted in the Torah and not in the sages’ own will. Therefore, one who betroths by means of witnesses whose disqualification stems from exposition is not acting against the sages’ will, and therefore his betrothal remains valid.

Chapter 3: The Concepts “de-oraita” (of Torah origin) and “de-rabbanan” (rabbinic) on the Plane of Halakhic Validity

Introduction

Until now we have dealt with the terminological and semantic classification of the various kinds of laws. We saw what de-oraita and de-rabbanan laws are, and what laws classified as divrei sofrim (laws of the sages) are. But we have not yet dealt at all with the question of the legal force of all these kinds of laws. As stated, the main dispute over the interpretation of Maimonides’ words revolves around the question whether he is dealing with legal force, as Nahmanides understood him, or only with source, as the Tashbetz understood him. We have already seen that he is certainly dealing with source, and that he determines that de-oraita laws are laws whose source is the Written Torah, whereas de-rabbanan laws are laws whose source is not literally there. Above we already mentioned that according to Maimonides the question of legal force is derived from the question of source. In this section we will deal with the question of the legal force of laws classified as divrei sofrim.

Evidence that Maimonides’ Determination Has Halakhic Consequences

We begin by noting several points that indicate that Maimonides’ discussion unquestionably has halakhic consequences. First, it is hard to believe that Maimonides would invest such great effort and develop such a complex theory concerning the concepts de-oraita and de-rabbanan if all this had no practical halakhic consequence.

More than that, we saw in the discussion of warning derived by logical inference that Maimonides draws a clear halakhic conclusion from the fact that a certain law emerges from exposition: namely, that there is here no warning that allows punishment. That is a direct halakhic conclusion drawn from the determination of source.

It is important, however, to note that this conclusion has a unique character. Even if Maimonides’ intention is indeed as the Tashbetz and his school understood it, it is still very plausible that his conception would have halakhic consequences for the question of warning. The reason is that the requirement of warning is not a question of force, but a question of presence. The fact that one does not punish without warning does not necessarily stem from the prohibition being less severe, but from the fact that the offender was not warned. From this we understand that the criterion of presence in Scripture, which, as we saw, is what the concept de-oraita means in Maimonides’ view, is a highly plausible measure for the existence of warning. If a prohibition is present in Scripture, then a person knows of it and is regarded as having been warned, and so he can be punished. By contrast, if the prohibition is not present in the Torah, then even if its force is extremely severe, a person cannot be regarded as one who was warned.⁴²

If so, this halakhic consequence cannot serve as an indication against the Tashbetz’s interpretation. Moreover, as we saw, several commentators who belong to the Tashbetz school nevertheless derive halakhic consequences from Maimonides’ definitions. As we already noted at the beginning of this discussion, this does not remove them from the Tashbetz’s school, because in their view the main discussion is still about source, though the question of source has some halakhic consequences as well. An example is the category “a matter that the Sadducees acknowledge.” According to most of the early authorities, with Maimonides as the exception, this category is not unique in terms of legal force, because it is an ordinary de-oraita law. Even so, there is no doubt that this definition has several practical halakhic consequences.⁴³ In the simple sense, these consequences do not indicate greater severity or some special legal force. Rather, they derive from the very nature of the definition “a matter that the Sadducees acknowledge,” that is, a matter that is explicit in the Torah.⁴⁴

In Maimonides’ View, the Question of Legal Force Has No Sweeping Meaning

As we saw, the discussion of classifying betrothal by money as a law of divrei sofrim dealt only with the semantic layer. Maimonides did bring proofs, but they were linguistic in nature. In none of the various sources does he present any halakhic consequence of the determination that betrothal by money is a law of divrei sofrim. On the contrary, as we saw, from several very clear sources in Maimonides, betrothal by money appears to be a de-oraita law even in terms of legal force.

Yet from the discussion of Maimonides’ proofs and criteria, and from the very fact that Maimonides ostensibly “fails to be precise” in his language if he mixes the terms de-oraita and de-rabbanan, which ordinarily express legal force, and uses them only in the context of source, with no connection to force, a question arises. On that reading, there is nothing in common between betrothal by money, which he calls de-rabbanan, and ordinary rabbinic laws. So why does Maimonides, who is renowned for linguistic precision, use the same term?

We saw that Maimonides also never spells out anywhere the consequences regarding legal force, and that the entire discussion concerns only the question of source. Why does he not tell us what the legal force is of the different kinds of laws? This question is very serious, especially in light of Maimonides’ broad halakhic aim. The main purpose of the Mishneh Torah is the architectural ordering of practical halakha, not the determination of sources. Yet according to the Tashbetz, large parts of the principled discussions in the principles and introductions are not dealing with halakhic questions at all, but with historical questions lacking any halakhic significance. This is an implausible way of understanding Maimonides’ method.

From this schematic portrait of the difficulty in understanding Maimonides, the following solution seems at first to arise naturally: Maimonides simply does not recognize questions of legal force as separate from questions of source. For him, the discussion of source and the discussion of legal force are one and the same, because legal force is derived from source. This is also clearly indicated by the shifts we saw in different sources in Maimonides between questions of source and questions of force, and vice versa. More than that, we saw several proofs that Maimonides interprets the language of the sages in the Talmuds in the same way. When they speak of a law as de-oraita, which for Maimonides is an indication that it is accepted by tradition and therefore present in the text, Maimonides understands this as a statement about source, because he himself derives from it conclusions about source. But in the language of the sages it is quite clear that the intention is the question of legal force and not the question of source.

When we read the language of the sages in the Talmud, we generally assume that when they speak of a law as de-oraita or de-rabbanan, they mean questions of legal force. Maimonides agrees with that, but for him this is only a derivative of the question of source. Usually, the learner and legal decisor are interested in legal force and not in source, and therefore the common usage refers to force rather than source.

But from this description it would seem that determinations regarding the source of laws of types 2-3 project onto the question of their legal force. If so, Nahmanides would be correct that according to Maimonides these laws also have rabbinic legal status. Yet this conclusion seems problematic in Maimonides’ view, both in light of his own sources, in light of the lacunae in his writings, that is, the fact that there are many places where he does not treat laws derived by exposition as rabbinic laws, and in light of the Talmuds. Most of his rulings in the Mishneh Torah generally do not contradict what is accepted in the Talmud and among the decisors, and what is accepted as de-oraita law is indeed treated by him as de-oraita as well.

If so, we are trapped in a dilemma. On the one hand, it is quite clear that Maimonides does not distinguish between the question of source and the question of legal force. That is, legal force is derived from source, and it is all one discussion. On the other hand, with respect to source he seems to determine that types 2-3 are laws of divrei sofrim, whereas with respect to force it seems more likely that his view is that they are de-oraita laws.

The conclusion that suggests itself is even more revolutionary. It seems clear that according to Maimonides there is no general question of legal force at all. It is not correct to say that legal force is derived from source. Rather, Maimonides simply does not recognize or define legal force as a distinct question. In his thought there is no general and sweeping categorical classification of laws into two kinds of force, de-oraita and de-rabbanan. Each kind of law is judged on its own terms. There can be laws whose source is rabbinic while their force is de-oraita, and laws whose force is also rabbinic. Moreover, there can be certain consequences for which a given law is treated as de-oraita, while in other respects that same law is treated as de-rabbanan.

A Logical Explanation: Intension and Extension

If we try to distinguish categorically between these kinds of legal force, we will almost always turn to their halakhic consequences, such as stringency in cases of doubt, being overridden by human dignity, and so forth. The best-known consequence is the rule of doubt: a doubt concerning de-oraita law is judged stringently, and a doubt concerning de-rabbanan law leniently. But, as is well known, Maimonides has a unique view on this subject as well. He holds that all the rules of doubt are themselves rabbinic. The distinction between doubt concerning de-oraita, where one must be stringent, and doubt concerning de-rabbanan, where one may be lenient, exists entirely on the rabbinic plane.

If so, this is a significant indication that according to Maimonides this categorical division does not really exist. The set of consequences that we usually attach to the concepts de-oraita and de-rabbanan must each be examined separately. One cannot draw sweeping halakhic conclusions simply from the fact that a given law is de-oraita or de-rabbanan. Each such consequence has to be discussed independently.

If we examine the pair of concepts de-oraita and de-rabbanan in their accepted sense, namely with regard to legal force, and attempt to define that very sense, we once again encounter difficulties. The definitions will always be given in terms of halakhic consequences, such as stringency in cases of doubt. In truth, however, we should try to define the concepts themselves. We must ask: what do the concepts de-oraita and de-rabbanan mean in themselves, apart from their halakhic consequences? Their meaning must precede those consequences. One cannot say that a rabbinic law is a law whose doubtful case is judged leniently, because leniency in doubt is a rule that applies to rabbinic laws, not a definition of them. What, then, is the definition of the term “rabbinic law”? If we are interested in the essential plane, the one that underlies the halakhic consequences, then we cannot make do with listing those consequences, or with defining the concept by listing the laws that fall under it, that is, its extension. Rather, we must seek a definition through content, that is, through the intension of the concept.⁴⁵

As a starting point, we should note that Nahmanides and all the other commentators offer the following intensional definition: a law with de-oraita force is a law whose source lies in the will of God, expressed by Him directly, whether in writing or orally, or in what emerges from the interpretation, whether by plain meaning or by exposition, of those expressions of will.⁴⁶ These commentators do not deal with source as such at all, except as an indirect part of clarifying legal force, that is, whether the law was expressed directly by God in writing or orally.

By contrast, in Maimonides’ view it seems utterly unreasonable to offer this intensional definition. Had that been his intensional definition, why would he develop such a complex and ramified conceptual system, supported by the intricate and innovative halakhic theory we have seen thus far? What would be the significance of all the discussion of source, if not as a condition for clarifying legal force? It is therefore clear that Maimonides has a different intensional definition. What is it? In light of everything said thus far, it is fairly clear that Maimonides’ intensional definition of the concepts de-oraita and de-rabbanan is the one we have been discussing until now: whether something is present in the Torah or not. That is the definition that serves as the basis for the questions of legal force.

But, as we have seen, the question of legal force is not derived unambiguously from this intensional definition. It is quite clear that according to this criterion every halakhic consequence has to be examined on its own, and one cannot establish a sweeping rule for all halakhic consequences that divides between de-oraita and de-rabbanan. In other words, the reason that according to Maimonides one cannot clearly divide between de-oraita and de-rabbanan laws is that the intensional definition is not relevant to questions of extension.

For purposes of comparison, according to Nahmanides the intensional definition of de-oraita law is law that expresses God’s direct will. This is a criterion that naturally defines legal force, that is, halakhic severity. The more a law is God’s will, the more severe it is. Thus from such an intensional definition one can naturally, directly, and sweepingly derive the extension. But if the intensional definition is that de-oraita is what is present in the written text, then it is far less natural to derive from this questions of severity and force. Why should what is present in the text be more severe than what is not present there? Why, for example, should a halakha le-Moshe mi-Sinai (a law given orally to Moses at Sinai), which is not present in the written text but is the most direct possible expression of God’s will, be less severe than laws explicitly written in the Torah? According to Nahmanides, the criterion is severity, and therefore these two kinds of laws are roughly equal in their legal standing and severity.

From Maimonides’ intensional definition one cannot derive severity, and therefore the various consequences must each be examined separately. The laws of doubt, the possibility of being set aside for the sake of human dignity, and so forth, all have to be discussed independently.

Of course, even according to Maimonides, by virtue of the logical relation between extension and intension, there must be some way to define from the character and source of a law its legal force, meaning the different halakhic implications of this classification. But, as we shall see, the derivation will not be as simple as it is for Nahmanides and his school, because in their view the derivation is simply according to a scale of severity: de-oraita laws are more severe than de-rabbanan laws.

We learn from this that, contrary to what we had assumed until now, there is indeed some relationship between the question of source and the question of force. Source defines the intension of the concepts de-oraita and de-rabbanan, whereas legal force belongs to another plane, halakhic in character. We now move on to define that relationship more explicitly and more sharply.

Presence in the Text as Command: Two Planes of Relating to a Commandment or a Transgression

The basic question is this: why should the fact that something is present in the Torah have any significance at all for the question of legal force? Alternatively, what halakhic consequence could a criterion such as presence in Scripture have, beyond the trivial consequences we saw above, such as “a matter that the Sadducees acknowledge,” which even members of the Tashbetz’s school propose?

The answer probably lies in what we saw above in the discussion of warning derived by logical inference. The only halakhic consequence in all of Maimonides’ writings that is explicitly presented as deriving from his definition of laws derived through exposition as divrei sofrim is that laws generated through exposition cannot serve as a warning, unless the punishment is explicit in Scripture.⁴⁷ Above we noted that such a consequence is natural according to Maimonides’ own criterion, because the function that warning fulfills requires precisely prominence, or explicit presence in the text.

This very argument gives us an initial clue to the meaning of the criterion of presence, or inclusion, in the text. A commandment that is present in the text itself means that we have a command from God regarding it. When there is a command, a person is regarded as warned. In other words, the criterion of presence in the text is relevant to warning. And this itself reveals the foundation of the meaning of the criterion of presence: presence in the text constitutes a warning, or command, with respect to that commandment. This brings us very close to the position of Nahmanides and the other early authorities, who tie the concept de-oraita to a direct expression of God’s will. But here it is nevertheless different. The classification de-oraita does not depend on the expression of will, nor on the fact that a given act brings about repair or corruption, but on the objective existence of a command in the Torah.

From here we may perhaps better understand the distinction we made above between different degrees of presence in the text. It follows that there are several degrees of command. We are not simply commanded or not commanded with respect to a given law. We can be strongly commanded, somewhat less strongly commanded, or not commanded at all. And from this it follows that with respect to the question whether a given law is de-oraita or de-rabbanan, we can also answer in several degrees: a law can be fully de-oraita, very de-oraita, fairly de-oraita, only slightly de-oraita, or not de-oraita at all.

Of course, in formal definitions we draw some line separating de-oraita from de-rabbanan. Therefore only command at a very high level leads to the creation of a basic norm, that is, a de-oraita law.

In Appendix E we will discuss in greater detail the meaning of the term “command,” and its legal and normative significance. As we will see there, every mitzvah (commandment) contains some benefit because of which we were commanded to perform it, and it also contains the command, which creates the normative obligation to bring about that benefit. The same applies to a transgression. There is the essential dimension, that is, the defect or corruption inherent in the act, and there is the command, which turns refraining from that defect or corruption into a binding norm. Therefore, one who performs a commandment both repairs something and obeys God’s command, while one who commits a transgression both damages something and rebels against the command.

Why is the command necessary? The matter is similar to the enactment of some civil law. So long as no prohibition has been enacted against crossing a red light, the act is foolish, improper, and perhaps immoral, but one cannot say that it is legally forbidden. What makes it forbidden is the legislation, not the fact that it causes damage or defect. The same is true of a commandment: what makes it obligatory is not the benefit in performing it, but the fact that there is a command concerning it.

To summarize, every de-oraita commandment or transgression contains two parallel planes, both of which are relevant to halakhic discussion: the essential plane, that is, the damage or repair involved in the act under discussion, and the plane of command, which turns the act of repair, or the avoidance of damage, into a binding halakhic norm.

But the existence of a command cannot be the sole criterion for a commandment’s being de-oraita, since a halakha le-Moshe mi-Sinai also comes with a command, even though it does not appear in the Torah. If so, appearance in the Torah means that there is a command, but there can also be a command outside the Torah, orally. So what is the ultimate criterion for de-oraita force? It seems that a commandment is de-oraita only if it satisfies two conditions: we have a command for it, whether written or oral, and it has an essence, meaning that there is an obligation to avoid some defect or corruption, or to perform some act of repair. As we shall explain below, the halakhic standing of a given commandment or law is determined by these two measures: command and essence.

Application to Maimonides’ Halakhic Distinctions: Returning to Nahmanides’ Questions

Above we saw that Maimonides defines de-oraita laws as laws that are present in the text, and de-rabbanan laws as laws that are not present in the text but outside it. In addition, we concluded that the meaning of this criterion, presence in the text, concerns the question whether a halakhic obligation has been created, leaving aside moral obligation, which can be inferred from the facts even without command.

If so, de-oraita laws are laws with respect to which we have a command, and therefore they involve a halakhic norm or obligation. De-rabbanan laws are laws with respect to which we have no command, and therefore no basic halakhic norm is involved, but only a derived or dependent norm, that is, a de-rabbanan law.

By contrast, there can be laws with respect to which there is no halakhic norm at all, neither primary nor derivative, and these are laws that arise from reason, sevara, or laws that are not present in the Torah. These are indeed norms, and not mere facts. But they are norms in the sense in which moral norms are norms. They do not involve a halakhic norm, whether primary or secondary.⁴⁸

If there is a norm that does not belong to the halakhic layer, it nevertheless seems that there is still a divine will that we observe it. Some would ground this in verses such as “You shall do what is right and good,” or “You shall be holy.” In that case, there is room to include these norms as well within the framework of halakha. But they do not belong to its legal dimension.

For example, a law grounded in reason, such as the blessing recited before deriving benefit from food, is also based on God’s will that we observe it. But this is a will that is not expressed as a command. In this sense, the situation is similar to a halakha le-Moshe mi-Sinai, which in our earlier classification was type 2. That too is a law that God wants us to observe, but it does not appear in the Torah, and therefore, from our point of view, no command is attached to it. There is an expression of divine will concerning it, but not a command. This is similar to God telling us that Sabbath observance is beneficial and proper. Would such a statement make it obligatory? Certainly not. Only a command would do so. The verse that tells us this must be phrased in the language of command and not in the language of description, that is, as the disclosure of a physical or spiritual fact. See Appendix E at the end of the book for all of this.

We now continue to spell out the application to Maimonides’ halakhic classification, including the components we added to that classification. We saw that there are eight basic types:

1. Laws explicit in the written text.
2. Laws for which there is tradition together with a midrashic anchor, earlier called type 1.
3. Laws for which there is only tradition.
4. Laws that have a midrashic anchor but no tradition.
5. Laws based on reason.
6. Decrees and enactments with an asmakhta (scriptural support).
7. Enactments without an asmakhta.
8. Decrees without an asmakhta.

We have already dealt above with laws explicit in the text and with laws based on reason. The laws that emerge from expositions, which in the earlier classification were type 3, are laws with respect to which there is certainly a divine will that we observe them, but there is no command concerning them, because the Torah does not command them. The midrash creates them; it does not reveal them in the verses themselves. By contrast, laws with respect to which we have both tradition and a midrashic anchor, type 1, we saw are regarded as present in the text. If so, these are laws with respect to which we do have a command.⁴⁹

As for de-rabbanan laws such as decrees and enactments, it is clear that they have no hint at all in the text. An asmakhta is a kind of hint, and as we saw in the previous unit it reflects a certain degree of inclusion in the text. This marks a distinction between an enactment and an exposition, even though all of them are de-rabbanan laws.

What about halakha le-Moshe mi-Sinai? We saw that its classification is that of divrei sofrim, and we explained this by saying that it has no hint in the text, that is, it is not present in the biblical text. But if we tie the matter to the question of command, it is difficult to say that a halakha le-Moshe mi-Sinai has no command. After all, Moses received these laws directly from God at Sinai, just as he received the Torah itself. There is indeed an oral command regarding them.

In principle, two possibilities can be proposed in understanding halakha le-Moshe mi-Sinai:

1. Command exists only in the written text itself, and therefore laws that have no anchor in the text have no command. According to this approach, God’s statement to Moses was merely informational disclosure, that a given act is beneficial or harmful, and not the imposition of a normative obligation. Yet one would still need to explain how halakha le-Moshe mi-Sinai differs from a law derived through exposition. Both lack command, and both have essence, that is, repair or corruption in reality, whether physical or spiritual. One might say that there really is no difference in their characterization, but they are still two separate types. Even so, it remains unclear why God chose to transmit the tradition through two different channels, if there is no essential difference between these kinds of laws. Perhaps one could say, following our earlier discussion of the sorites paradox, that halakha le-Moshe mi-Sinai involves no command at all, whereas laws derived through exposition involve “half a command,” because they are “rather included” in the text. This may create some differences, but it is not sufficiently sharp.

2. It is more plausible to say that commands are usually conveyed only through the written text, but halakha le-Moshe mi-Sinai is an exceptional case, in which there is an orally given command. But then the question arises: why is this kind of law, for which there is a command, classified as divrei sofrim? According to this approach, it seems likely that these laws have command but no essence, and that this too is called divrei sofrim. It follows that the definition of de-oraita law is not simply law with respect to which we have a command, because halakha le-Moshe mi-Sinai also has a command. The technical definition is: what is present in the Written Torah. The essential definition is: what has both command and essence. The absence of either one will prevent the law under discussion from being classified as de-oraita.⁵⁰

It should be noted that the second approach gives us an explanation, lacking in the view of most of the early authorities who disagree with Maimonides, of why there is a separate category of laws that were not written in the Torah itself but transmitted orally. Why did God not write them in the Torah itself? That is difficult if we assume that their nature and halakhic significance are identical to those of the laws written in the Torah. According to our approach to Maimonides, however, the explanation is simple: this is a different type of law. It has only command and no essence.

Beyond that, the second approach also yields a fuller explanation of Nahmanides’ difficulty concerning the halakhic category of type 1. As we saw, halakha le-Moshe mi-Sinai is command without essence. By contrast, laws derived through exposition are essence without command. If so, when we have a law that was transmitted to Moses at Sinai and is also supported by exposition, we have both command and essence. It now becomes entirely clear why the combination of these two mechanisms yields a de-oraita law, even though each one separately yields only a de-rabbanan law. Thus laws of type 1 are considered de-oraita in the most ordinary sense possible.

Our conclusion is that de-oraita law is law that has both essence and command. As we shall see further on, these two parameters underlie the categorical map of halakha, and also underlie the halakhic distinctions we make between de-oraita and de-rabbanan laws.

Note 5: Essence and Command

We have already mentioned several times that every commandment or transgression has two layers: essence and command. Here we noted that this is the criterion for a commandment’s being de-oraita.

The source we cited for this is Rabbi Elchanan Wasserman’s essay “Repentance,” which derives it from Rabbi Moshe Chaim Luzzatto, who writes as follows in chapter 4 of his book Derekh Hashem:

The purpose, for man, of performing the commandments is clear: to fulfill his Creator’s commandments and to do His will. He fulfills His will, blessed be His Name, in this in two ways, each following from the other. First, he fulfills His will thereby that He commanded him to perform that act and he performs it. Second, by that act he is perfected in one of the degrees of perfection that is the product of that commandment, as we have mentioned. Thus His will, blessed be His Name, is fulfilled, for He wills that man be perfected and come to delight in His goodness, blessed be His Name.

The concept of command will be sharpened and elaborated in Appendix E at the end of the book. Here our purpose is to show a source from Maimonides himself for the claim that a de-oraita commandment is a commandment that contains both essence and command.⁵¹

For this we must turn to Maimonides’ ninth principle, where he discusses repeated commands in Scripture. This principle is divided into two parts. The first part rejects counting repeated commands separately. For example, commands regarding Sabbath observance appear twelve times in Scripture, yet we count only one commandment of Sabbath observance. In the second part Maimonides discusses a general prohibition and rules that all the branches that emerge from it are to be counted as a single commandment. For example,⁵² the verse “You shall not eat over the blood” commands us not to eat from an animal before its life departs, and likewise not to eat sacrificial meat before the blood has been dashed, and not to eat on the day on which someone is put to death, and not to eat before prayer,⁵³ and in addition it also serves as the warning for the stubborn and rebellious son. Thus several commandments are learned from one verse, and there is apparently no substantive connection among them, yet Maimonides rules that they all must be counted as only one commandment. See negative commandment 195, with respect to the stubborn and rebellious son.

Rabbi Yerucham Fishel Perla, in his comments on this principle, notes that at first glance the two parts of the principle contradict one another. The first part teaches that for purposes of counting the commandments, a “commandment” is only a command that has unique content, that is, unique essence. Commands that appear explicitly in Scripture are not counted because they have no unique and independent content; they repeat the content of another command. From this it would seem that what determines the count of commandments is the essence, that is, the repair or corruption involved, and not the presence of a command in Scripture. On the other hand, the second part yields the exact opposite conclusion: what determines the count is the presence of a command, and therefore commandments that are different in content, because they are all learned from one prohibitory verse, in the case of a general prohibition, are counted only as one prohibition. If so, it would seem that what determines the count is the existence of a command and not the unique content.

The simple resolution of this contradiction is, of course, that according to Maimonides both requirements are necessary. For a commandment to be counted separately in the tally of commandments, that is, for it to be considered a de-oraita commandment, two conditions are required:

1. There must be an explicit command for it in Scripture.⁵⁴
2. It must have its own unique content, meaning that it does not merely repeat the content of another commandment already counted.

The first part of the ninth principle deals with the second condition, and the second part deals with the first.

The basis of the matter is apparently the understanding that every commandment or transgression has two aspects: command and content. Every commandment that a person performs has two aims or two benefits: fulfillment of God’s command, and the specific benefit for which the commandment is intended, that is, the reason or essence of the commandment. The same is true of transgressions. Every transgression has two aspects: rebellion against the command, and the defect or damage that the transgression was intended to prevent.

From this we can understand Maimonides’ wording in several of the principles, where he prefaces his discussion by noting that the commandments under consideration appear explicitly in the Torah.⁵⁵ Without the appearance of the command in the Torah, there is no room at all to discuss counting the commandment.⁵⁶ Once the commandment appears in the Torah, the discussion begins concerning how it should appear in the tally of commandments, both from the standpoint of individuation and from the standpoint of filtering.

In our essay on the ninth principle we also discussed the significance of the principle’s opening words, where Maimonides details all the different purposes of the Torah’s commandments. At first glance it is not clear why he does this precisely there. Why not at the beginning of the Book of Commandments? We explained there that these remarks are intended to clarify that every commandment contains not only an element of obedience but also a substantive element. If the point of a commandment were only obedience, then every appearance of a commandment should have been counted as an independent commandment. For example, Sabbath observance should be counted as twelve commandments, because one who desecrates the Sabbath violates twelve verses that warn against it. Only because the meaning of the commandment depends also on its content, and not only on the command, is there room to see all twelve appearances in the Torah as repetitions of one commandment. For more detail, see our discussion there.

After explaining Maimonides’ classification within the conceptual framework we have proposed here, we must clarify several points that remained obscure. What is the difference between exposition and reason? A law that emerges from reason is a law that has essence and no command, exactly like laws derived through exposition. At first glance the two types are both laws with essence but no command. First, it may be that in fact there is no difference in type, only in the source that teaches us the law. Where reason is insufficient, exposition is required, and where reason is available no exposition is needed. We may note that in many places Maimonides uses the expressions “reason” and “derived through the hermeneutic rules” interchangeably. See, for example, the Laws of Rebels, the introduction to the Mishnah, and elsewhere.

Beyond that, in terms of the sorites paradox, it is fairly clear that a law derived through exposition is “half included” in the text, whereas reason is not included in it at all. That is, even exposition contains a weak dimension of command. According to this, the midrashic inference has the significance of command, though not in the binary sense we adopted until now, that there either is or is not a command, but rather as a weak command.

If so, what is the meaning of the claim that reason is de-oraita? See Appendix D at the end of the book. According to Maimonides this does not seem entirely correct. Such a law is certainly no more de-oraita than laws derived through exposition. The meaning of this claim, according to Maimonides, seems to be that a law learned through reason has essence, as de-oraita laws do, and in that sense it does not resemble de-rabbanan laws in the ordinary sense, namely enactments and decrees. That essence obligates us to perform the act or refrain from it, but this is a norm in the sense of a moral norm, not in the sense of a legal-halakhic norm.⁵⁷ No dimension of command is involved in it at all.⁵⁸

Summary

We have seen that in Maimonides’ view the division between de-oraita and de-rabbanan points to a division of sources, not to a division concerning legal force or the severity of the laws under discussion. We also saw that the classification of the different kinds of laws is based on two components: essence and command. Laws explicitly written in the Torah, as well as laws for which we have both exposition and tradition, are laws that contain both essence and command. Laws learned from creative exposition are laws with respect to which we have no command, because they are not hidden in the Torah’s verses but are created through expansion by the sages, though they do possess essence. A halakha le-Moshe mi-Sinai is a law with command, orally conveyed, but without essence, at least not essence that creates primary halakhic obligation. This is the full map of the kinds of laws that are connected in one way or another to the category de-oraita.

It is extremely important to note that this division is a division of kind and not a division of severity. In principle, there may be laws with respect to which we do not have full command, yet they are very severe. Conversely, there may be laws for which we did receive a command, even though they are not especially severe. Precisely for that reason Maimonides does not discuss the question of legal force and severity. Questions of severity simply are not connected to the terminology de-oraita and de-rabbanan, and therefore the question of legal force has no sweeping meaning at all. As we saw above, according to Maimonides the concept of “force” lacks intension. The halakhic consequences, such as stringency or leniency in cases of doubt, are not the result of severity and leniency, but of type. Every such halakhic consequence must be decided and explained on its own terms, not by way of some inclusive rule that depends only on the classification de-oraita or de-rabbanan.

In light of this principle, Nahmanides’ ninth difficulty, cited above, is no difficulty at all. He proves there that many laws are halakha le-Moshe mi-Sinai, yet their severity is greater than that of laws that appear explicitly in the text. According to our approach, there is no contradiction here. Their severity may indeed be greater, even though they are halakha le-Moshe mi-Sinai and are classified by Maimonides as divrei sofrim.

As a striking example, according to our approach it is entirely possible that doubts concerning de-rabbanan laws should sometimes be decided stringently and not leniently, depending on the kinds of laws involved and on the reasoning behind the rule that determines whether doubt requires stringency or leniency. Indeed, we shall see below that doubt concerning a law derived through exposition is judged stringently, while doubt concerning a halakha le-Moshe mi-Sinai is judged leniently, even though both of these kinds of law are conceptually classified as divrei sofrim. The same applies to the other consequences of the standard distinction between de-oraita and de-rabbanan. According to Maimonides, in every such case we must examine the application to each halakhic category, whether laws derived through exposition, halakhot le-Moshe mi-Sinai, laws explicit in the Torah, laws derived from reason, and so forth, not necessarily from considerations of severity and leniency, but from their essential characteristics.

In the next chapter we will try to examine several such consequences and show how our approach to Maimonides’ view is implemented, beginning, of course, with the laws of doubt.

Chapter 4: How Maimonides Explains the Halakhic Differences Between de-oraita and de-rabbanan Laws

Introduction

As stated, in this chapter we will try to see how Maimonides understands the differences between de-oraita and de-rabbanan laws. We saw that most of the early authorities understand these differences as consequences of a difference in severity. That follows from the fact that for them the concepts de-oraita and de-rabbanan reflect halakhic severity or force. But in Maimonides’ thought, as we have seen, the pair of concepts de-oraita and de-rabbanan does not bear that meaning. For him they express source and not force, and the differences between the laws are differences of type and not of severity. Consequently, in every halakhic context Maimonides must offer a substantive explanation. In this chapter we shall see that one can develop here a general doctrine that maps many of these differences by means of the distinction between command and essence, which we presented at the end of the previous chapter.

1. Laws of Doubt

As is well known, a doubt concerning de-oraita law is judged stringently, whereas a doubt concerning de-rabbanan law is judged leniently. As we have noted more than once, this is one of the central laws used to challenge Maimonides on these two principles. The reason is that this is the sharpest and most common consequence of the distinction between de-oraita and de-rabbanan, insofar as legal force and severity are concerned.

But according to our approach to Maimonides’ view, the division between de-oraita and de-rabbanan is a division of type and not of severity. If so, why is there a sweeping rule that a doubt concerning de-oraita is judged stringently and a doubt concerning de-rabbanan leniently? Does this rule also apply to laws derived through exposition and to halakha le-Moshe mi-Sinai?

Here we should recall that Maimonides holds that the rule requiring stringency in cases of doubt concerning de-oraita law is itself only rabbinic. His language in the Laws of Defilement by the Dead, chapter 9, halakha 12, is as follows:

It is well known that all such impurities and the like, where the issue is one of doubt, are of rabbinic law, and one is impure by Torah law only when he has certainly contracted impurity. But all doubtful cases, whether in matters of impurity, forbidden foods, sexual prohibitions, or Sabbaths, are of concern only by virtue of divrei sofrim, as we explained in the Laws of Forbidden Intercourse and in several other places. [Even so, a matter whose intentional violation incurs karet is forbidden in a doubtful case by Torah law, for one who does it is liable to a suspensive guilt-offering, as we explained in the Laws of Unintentional Transgressions.]

The bracketed addition in this text is the subject of great dispute among Maimonides’ commentators and among the textual witnesses. Many rejected it on textual grounds. For example, how could doubtful cases involving karet be forbidden, when Maimonides himself includes Sabbath and sexual prohibitions among the doubtful cases for which, according to his formulation here, there is concern only by virtue of divrei sofrim? Beyond that, in all the parallel passages in Maimonides’ writings, such as Laws of Forbidden Intercourse 18:17, Kilayim 10:27, and Laws of the Other Primary Sources of Impurity, end of chapter 16, there is no hint of this distinction, and the rule that all doubts are of rabbinic status is stated there in a general and sweeping way.

It therefore seems that this reading has no real basis, despite the fact that the Kesef Mishneh cites it there, and that it is found in other later authorities and was even adopted by the Frankel edition. It is fairly clear that the main motivation for fixing this addition as the correct text in Maimonides is halakhic rather than textual. Apparently, from halakhic considerations, those authorities felt it necessary that in cases of doubt where the prohibition has been established one must be stringent by Torah law, and therefore they corrected Maimonides’ text accordingly.

The reason for this difficulty is that Maimonides himself rules at the beginning of chapter 8 of the Laws of Unintentional Transgressions that in doubtful cases where the prohibition has been established one must bring a suspensive guilt-offering. But if the prohibition in a doubtful case is only rabbinic, then that offering would amount to slaughtering an unconsecrated animal in the Temple courtyard. Therefore Maimonides’ commentators wanted to explain that his intention, according to the bracketed addition, was to distinguish between a doubt in a case of established prohibition, where even Maimonides agrees that the doubtful case is forbidden by Torah law, since one brings a suspensive guilt-offering for it, and a doubt in a case where no prohibition was established, where the duty to be stringent is only rabbinic. But now the strain becomes even greater, because in any case that is not what the text says here, even according to the corrected version in brackets. Below we will see, and compare note 1, that this halakhic pressure is not problematic at all, and that one may retain the original text.

Let us return to the challenge posed to Maimonides. If indeed the duty to be stringent in cases of doubt is only rabbinic, then the challenge is already less severe. Perhaps the sages, for various reasons, wanted to sharpen the distinction between the Written Torah and the Oral Torah. Nahmanides, in his glosses on the first principle, suggests something like this to explain in Maimonides the rule that doubt concerning de-rabbanan law is judged leniently. On that basis, one might say that the sages laid down a sweeping rule that every doubtful de-rabbanan case is judged leniently and every doubtful de-oraita case stringently.

But that is apparently incorrect. We have already seen in Maimonides that in cases of doubt concerning betrothal by money one rules stringently. See, for example, Laws of Marriage 3:20. Beyond that, Nahmanides goes to great length to prove from several sugyot that even in cases of doubt concerning laws derived through exposition one rules stringently. It is hard to believe that Maimonides would disagree with this halakhically, and one does not see such disagreement clearly in almost any place in his writings. We must therefore seek an explanation that depends on the type of law and not on some general external label, as we already said one must do in understanding Maimonides.

First we must define the boundaries of the discussion. Laws that are written explicitly in the Torah are unquestionably judged stringently in cases of doubt. By contrast, decrees and enactments are unquestionably judged leniently in cases of doubt. As for laws derived through exposition, we have noted that Maimonides will apparently agree that their doubtful cases are judged stringently, even though he calls them divrei sofrim. With respect to halakha le-Moshe mi-Sinai, there are apparent contradictions in his writings, and we will discuss them below.

To explain this, we should first note what Rabbi Shlomo Zalman Auerbach wrote in his commentary on Shev Shema’tata, first section. Rabbi Shlomo Zalman assumes that Maimonides’ view is as Nahmanides understands it in the first principle, namely that every rabbinic transgression is in fact a violation of the de-oraita prohibition “You shall not turn aside.” If so, the obvious question arises: why is doubt concerning these prohibitions not judged stringently? We discussed this in the first unit. Rabbi Shlomo Zalman explains that de-rabbanan laws involve only command, by virtue of “You shall not turn aside,” but no essence in the object itself. Fowl cooked in milk is not forbidden in itself, at least not by Torah law. In our discussion in the first unit we saw that on the rabbinic plane there is an additional prohibition with its own distinct legal identity, namely the prohibition of fowl in milk. It is not merely a prohibition of disobedience. If so, when we face a doubtful case, what we have is only a doubtful command. But, argues Rabbi Shlomo Zalman, a doubtful command is not a command, and rebellion against a doubtful command is not rebellion. Rebellion is defined only when we defy a definite command. If we now add the fact that, apart from command, de-rabbanan prohibitions have no layer of essence, at least on the de-oraita plane, then it is clear that in a doubtful case we violate nothing at all: a doubtful command does not obligate, and essence is absent altogether from de-rabbanan prohibitions on the de-oraita plane.

If we continue along this line of thought and ask ourselves why a doubtful de-oraita case, that is, a case involving laws explicitly written in the Torah, is judged stringently, we should recall our assumption that de-oraita laws contain both essence and command. Which of these two gives rise to the rule that in a doubtful case one must be stringent? Rabbi Shlomo Zalman’s reasoning shows that the source can only be the essence. In a doubtful de-oraita case the command is absent, because what we have is a doubtful command, and such a doubtful command is not a command. What remains is the essence, namely the concern that we may violate the original essence and thereby create defect or corruption.⁵⁹ Therefore, if one must be stringent in a doubtful case, it is clear that this duty to be stringent derives from the essence and not from the command.

At this point the consequences of our analysis begin to appear. First, according to what we have said, the duty to be stringent in cases of doubt is itself a duty for which there is no command. The command that accompanies the law in question is absent in a doubtful case, as Rabbi Shlomo Zalman argued.⁶⁰ But in Maimonides’ view, as we saw, a duty without a command, that is, a duty that is not written in the Torah, is not a de-oraita duty but is called divrei sofrim. Thus Maimonides’ well-known view, that a doubtful de-oraita case is judged stringently by rabbinic law, emerges naturally from our analysis. According to Maimonides, the duty to be stringent in a doubtful de-oraita case is itself only a de-rabbanan law, because we have no command with respect to it.⁶¹

Precisely here it is important to stress that in Maimonides’ thought in general, and here in particular, when we say that this is a “rabbinic” duty, we do not mean force or severity, but only type. The duty to be stringent can be understood as a full-fledged obligation, what outside Maimonides’ terminology is often called de-oraita force, just as the other early authorities understand it. As we have seen, Maimonides does not use the terms de-oraita and de-rabbanan to describe force but only source, and therefore his classification is different. His practical halakhic treatment of this duty, in terms of its seriousness, may be very similar to that of the other school of early authorities.

We must now discuss the view of the later authorities. See Shev Shema’tata, first section, chapters 1-2; Responsa Maharit, Yoreh De’ah, sec. 1; Peri Megadim, Yoreh De’ah 110, Siftei Da’at 14, and there in the laws of double doubt 25; and many others. They distinguish in Maimonides between a doubtful de-oraita case in which the prohibition has been established, where even according to Maimonides the duty to be stringent is de-oraita, and a doubtful case where the prohibition has not been established, where Maimonides holds the duty to be stringent is only de-rabbanan. As we saw above, some even wished to insert this distinction into Maimonides’ text in the Laws of Defilement by the Dead. According to our approach, this is very difficult. Even if the prohibition has been established, what we still have is no more than a doubtful command, which, as we saw, is not a command. Therefore it remains only a doubt concerning essence, and so in that case too the duty to be stringent should be only de-rabbanan.

What, then, actually compelled those later authorities to say this in Maimonides? In practice, as we have seen, Maimonides himself does not distinguish between the cases with regard to the laws of doubt. He distinguishes only with regard to the laws of the suspensive guilt-offering. See the beginning of chapter 8 of the Laws of Unintentional Transgressions. Nor is the emended text in the Laws of Defilement by the Dead really well founded. It is quite clear that these later authorities changed it because of the difficulty of explaining how Maimonides can require a suspensive guilt-offering if all doubtful prohibitions, even those involving de-oraita law, are only de-rabbanan. If so, this would be slaughtering an unconsecrated animal in the Temple courtyard. As noted above, this is also the main reason for changing the text there. Therefore they concluded that doubtful cases for which a suspensive guilt-offering is brought, that is, doubtful cases where the prohibition has been established, must also be forbidden by Torah law according to Maimonides. As stated, this is strained, because Maimonides speaks in a general and sweeping way about the laws of doubt, and it does not appear that he distinguishes for this purpose between established and unestablished prohibition.

Our conclusion, therefore, is that according to Maimonides every doubtful de-oraita case is judged stringently only by de-rabbanan law, across all types of law. That is what emerges from the plain text of Maimonides and from his reasoning. As for the suspensive guilt-offering, which created the main difficulty in this passage, there is no difficulty at all. The reason is that a guilt-offering is brought even when there is no sin whatsoever. Therefore there is no need to assume that a doubtful case of established prohibition is a de-oraita transgression in order to explain why a suspensive guilt-offering is brought. We explain and prove this in Appendix H at the end of the book. There we show that this is a substantive characteristic of the guilt-offering in general, across all the examples we have. A guilt-offering is always brought for transgressions that contain essence without command. That is, it is brought for the factual or ontological defect, and not for rebellion against the command. As we saw, according to our explanation of Maimonides, doubtful transgressions also involve only essence without command, and therefore a suspensive guilt-offering is brought. For a fuller discussion, see our remarks there.

Application to Laws Derived Through Exposition

Let us now see the consequences of such a picture. With respect to laws derived through exposition, earlier called type 3, we saw that they contain essence without command, or at least without full command. If so, in a doubtful case all we have is a doubt concerning essence, because from the outset these laws do not involve the issue of command. But, as we proved above, even in an ordinary de-oraita transgression, involving a prohibition explicitly written in the Torah, the duty to be stringent in a doubtful case stems only from the essence, because doubt about command does not define rebellion. If so, there is no reason to distinguish doubtful cases involving laws derived through exposition. There too we have a doubt concerning essence without command, and there too we must be stringent. Thus even doubt concerning laws derived through exposition, although Maimonides defines such laws as de-rabbanan, will be judged stringently. It follows that no challenge can be raised against Maimonides from sources showing that doubt concerning laws derived through exposition is judged stringently.

The explanation is that the reason for stringency in cases of doubt is the concern that we may violate a Torah prohibition and create a defect. True, there is no concern of rebellion against a command, even where a command exists, because rebellion in a doubtful case is not rebellion. But there is nevertheless concern that the defect may come about. For example, if a piece of fat lies before us and it may be forbidden fat or permitted fat, then if we eat it there is some chance that we have eaten forbidden fat. That is a matter of reality. Therefore that concern tells us to be stringent and not to enter a doubtful situation. It is not because of the command, but because of the concern over the defect.⁶²

Our analysis implies that laws derived through exposition involve this very same concern, and therefore it is obvious that their doubtful cases too will be judged stringently, rabbinically, of course. The same applies to laws derived from reason, which have no formal command but only essence. In such a case there is room to resolve the doubt stringently. From here we can understand why it is commonly assumed that laws derived from reason are considered de-oraita laws. The reason is that their doubtful cases are judged stringently, and therefore from the standpoint of practical halakhic consideration they function as de-oraita laws. But in terms of source and character, according to Maimonides they are probably de-rabbanan laws.

We thus find that, according to Maimonides, there is a type of de-rabbanan law whose doubtful cases are judged stringently. This proves that the terms de-oraita and de-rabbanan in Maimonides’ language do not reflect some distinct and well-defined halakhic essence. Rather, they involve a collection of characteristics that follow from character and type, and not necessarily from severity.

Application to Halakha le-Moshe mi-Sinai

What, then, happens with regard to halakha le-Moshe mi-Sinai?⁶³ Here everything depends on how we understand the nature of that category. Above we proposed two possible explanations. According to the first, halakha le-Moshe mi-Sinai contains no command and only essence. If so, it is quite clear that doubtful cases involving it will be judged stringently, just as we saw that a doubt concerning essence is judged stringently. According to the second explanation, which we saw to be more plausible, halakha le-Moshe mi-Sinai contains only command and no essence. According to that, a doubt concerning halakha le-Moshe mi-Sinai should be judged leniently and not stringently.

Among the decisors it is generally accepted that doubt concerning halakha le-Moshe mi-Sinai is judged stringently. Maimonides is the only exception. With respect to his view we find disagreement among the commentators. On the one hand, several of his commentators write that doubt concerning halakha le-Moshe mi-Sinai is judged stringently. See, for example, Yad Malakhi, Rules of Maimonides, letter zayin; the Maggid Mishneh at the beginning of the Laws of Marriage; and Megillat Esther here, p. 60, and others. But anyone who examines Maimonides’ Commentary on the Mishnah to tractate Mikvaot 4:7, which we cited above, will see proof that doubt concerning halakha le-Moshe mi-Sinai is judged leniently. His wording there is:

When they said, “because it is from the Torah,” they meant the basic obligation of immersion in a pool of water. And the language of the Tosefta is: an olive-sized quantity of carrion and a lentil-sized quantity of a creeping thing, where there is doubt whether there is the required measure or whether there is not the required measure, the doubtful case is impure, for any matter whose basic rule is from the Torah and whose measure is from divrei sofrim, its doubtful case is impure. And we have already explained that this does not contradict what they said, “measures are a halakha to Moses from Sinai,” because anything that is not found in the written text they call divrei sofrim.

Maimonides here rules that where there is a doubt regarding a Torah law whose measure is from divrei sofrim, that is, from halakha le-Moshe mi-Sinai, we rule stringently. But that is only because the matter has an original de-oraita basis and foundation. It follows clearly that in a law that is entirely based on halakha le-Moshe mi-Sinai, the doubtful case is judged leniently. The same is implied by Maimonides in the Laws of Slaughter, 5:3. See there also the Ramakh, the other commentators there, and the Rivash, sec. 163.

However, in the Laws of Defilement by the Dead, 2:10, in the discussion of a single bone the size of a barley grain that imparts impurity by touch and by carrying, Maimonides writes as follows:

The impurity of a single bone is a halakha from oral tradition, for it is said: “Whoever touches a bone.” They learned through oral tradition that even a bone the size of a barley grain imparts impurity by touch and by carrying. And since its impurity is a halakha, it is Torah law and not divrei sofrim.

At first glance, this would seem to indicate that Maimonides holds that doubt concerning halakha le-Moshe mi-Sinai is judged stringently. See the Kesef Mishneh and the other commentators there, as well as Tosafot Yom Tov, Ohalot 7:3. Yet one may reject this, because perhaps Maimonides is speaking there of a different concept, “halakha” or “oral tradition,” and not specifically of halakha le-Moshe mi-Sinai.⁶⁴ Beyond that, since the laws of corpse impurity are present in the Torah and the exposition merely specifies the measure involved, this is similar to what we saw above in the Commentary on the Mishnah to Mikvaot: a matter whose basis is in the Torah and whose measure is derived through halakha is de-oraita, and therefore one must be stringent in its doubtful cases.

This also resolves many of Nahmanides’ questions, where he tried to prove from several Talmudic passages that halakha le-Moshe mi-Sinai is considered de-oraita and that its doubtful cases are judged stringently, in the previous gate, section 12. The terminology de-oraita poses no difficulty for Maimonides, because it indicates source and not force. And the severity of the law, that is, the fact that its doubtful case is judged stringently, is, as we have seen, not necessarily a matter of severity but of type. A doubtful case of halakha le-Moshe mi-Sinai that interprets existing laws will indeed be judged stringently, because those laws are included in the text. In such cases halakha le-Moshe mi-Sinai functions interpretively and not innovatively.

For example, Nahmanides, in his glosses to the first principle, p. 22, brings the discussion in Babylonian Talmud, Kiddushin 39a, and writes:

And they treated matters that are a halakha to Moses from Sinai stringently, like Torah law. As they said in the first chapter of Kiddushin: orlah outside the Land is a halakha to Moses from Sinai. And they asked: but is it not taught that doubtful orlah in the Land is forbidden, whereas in Syria it is permitted? This implies that if it were a halakha to Moses from Sinai, its doubtful case would be forbidden.

According to our approach, one can resolve this by saying that here the halakha le-Moshe mi-Sinai introduces a detail within the laws of orlah, and therefore its doubtful case should indeed be judged stringently.⁶⁵ See also Nahmanides on the first principle, p. 24, where he brings another proof from the exposition in Babylonian Talmud, Yoma 28a, though that proof is not necessary.

See the general introduction of the author of Peri Megadim, as well as Lev Sameach in his comments on this principle, who explain the matter in light of this distinction. Compare also our earlier discussion of Nahmanides’ seventh difficulty, where we raised such a possibility in understanding his words.

According to the understanding of Yad Malakhi and his school, it seems that halakha le-Moshe mi-Sinai has essence, and therefore in a doubtful case we must be stringent, just as with de-oraita laws. On that view halakha le-Moshe mi-Sinai would be just like laws derived through exposition. But according to the Commentary on the Mishnah here, and according to the Laws of Slaughter, doubt concerning halakha le-Moshe mi-Sinai is judged leniently. That means that it has no essence but only command. These are exactly the two approaches we proposed above in understanding the concept halakha le-Moshe mi-Sinai.

In the plain sense, Maimonides seems more likely to hold the second view, namely that halakha le-Moshe mi-Sinai contains command without essence, and therefore its doubtful cases are judged leniently. This understanding is also plausible for another reason: halakhot le-Moshe mi-Sinai are usually not intelligible and seem to lack an explanation.⁶⁶ As we already noted above, this points to the second approach and not to the view of Yad Malakhi.

Interim Summary and the Words of Sho’el u-Meshiv

To summarize this section: we have seen that in Maimonides there is no sweeping principle of “a doubt concerning de-oraita law is judged stringently” or “a doubt concerning de-rabbanan law is judged leniently.” The concepts de-oraita and de-rabbanan do not describe force and severity but different types of law. Therefore, in cases of doubt each matter must be judged on its own and not by a sweeping rule based on source. Source as such is not relevant to severity or to how doubtful cases are handled. Its consequence is that laws for which we have a source in the Torah come with command, whereas laws with no source in the Torah come without command. This affects the laws of doubt, which, as we have seen, are governed by the principle that doubt concerning essence is judged stringently. Therefore there are de-rabbanan laws with respect to which we must be stringent, because they contain an ontological dimension, that is, essence, such as laws derived through exposition and perhaps also halakhot le-Moshe mi-Sinai. By contrast, there are de-rabbanan laws in which we are lenient because they contain only command, rabbinically and perhaps also as a weak de-oraita command, but no essence. With halakha le-Moshe mi-Sinai we saw a dispute, yet from at least two places in Maimonides it appears that he understands its doubtful cases leniently.

I was shown the responsum of Sho’el u-Meshiv, sixth edition, question 51, and this is his language:[^263]

But in my humble opinion a new idea seems to emerge: Maimonides too holds that all doubtful cases are governed by Torah law, and the only difference between them is a difference of name. For it is already well known what our great teacher established as a principle, that everything not explicit in the Torah is called by him divrei sofrim, although in truth its law is treated in every respect as Torah law. As he wrote in his principles, principle 2, and as Zohar ha-Rakia and Megillat Esther explained the matter, and so too the Maggid Mishneh wrote at the beginning of the Laws of Marriage, that Maimonides calls betrothal by money “divrei sofrim,” even though in law it has the status of Torah law in every respect, because it is not explicit in the Torah.

According to this, here too, although its doubtful case is judged stringently by Torah law, it is nevertheless not explicit in the Torah that in such a doubtful case we must rule stringently by Torah law. Therefore our master wrote that it is only divrei sofrim, because that is what he calls divrei sofrim.

According to this, our master wrote very well in chapter 18 of the Laws of Forbidden Intercourse, and in Kilayim, and in the above-cited Laws of Defilement by the Dead, that these are only divrei sofrim. That means that as to the name they are called only divrei sofrim, but in law they are Torah law in every respect. And now that reading is very well explained. For our master already established in his principles there that whatever we find the sages calling “words of Torah” is indeed called words of Torah, even if it is not explicit in Scripture. If so, then all the more so with regard to a matter for which one is liable to a suspensive guilt-offering, where we find explicitly in Scripture that the Torah imposed a suspensive guilt-offering. Such a matter is certainly also called Torah law, and thus it is true Torah law. If so, all the structures, difficulties, and contradictions raised against our master from many places collapse, and according to what I have written everything is well resolved. Consider this carefully, for it is a new observation.

This passage summarizes almost all the principles we have discussed here, and so it is worth examining in some detail. First, the author of Sho’el u-Meshiv states that for Maimonides the term de-oraita means “present in the Torah,” exactly as we proved above. He then assumes, as nearly all the commentators did, that according to Maimonides divrei sofrim is only an expression of source and not of legal force. This too accords with our approach, and indeed it is necessary, because if the distinction is not between God’s will and the sages’ will, but between what is present in the Torah and what is not, then clearly the distinction is not one of force and severity, as in the view of the other early authorities. Rather, it is a distinction of type.

However, the author of Sho’el u-Meshiv adds that the legal force of these laws is de-oraita. That is not our position here, because we explained that according to Maimonides there is no sweeping classification according to legal force at all, and every practical implication must be discussed on its own terms. The distinction between de-oraita and de-rabbanan is a distinction of source, while legal force is determined specifically in each context. From this, Sho’el u-Meshiv concludes that one can also explain Maimonides’ determination that the rule requiring stringency in a doubtful de-oraita case is itself de-rabbanan, because for Maimonides the term de-rabbanan indicates source rather than force. Thus this determination as well refers only to source and not to force. This is certainly correct according to our approach too. He concludes that in practice the force of the duty to be stringent in a doubtful de-oraita case is de-oraita, whereas its being called divrei sofrim concerns only source. This is exactly what we proposed above: the source of the duty to be stringent is indeed divrei sofrim, but its force must be determined in each context independently.

The author of Sho’el u-Meshiv assumes, like the Tashbetz, that everything is de-oraita, and at this point we diverge from him. We explain that according to Maimonides there is no sweeping concept of legal force at all, and that in each context the matter must be judged independently. For example, we propose a possible distinction between halakha le-Moshe mi-Sinai and laws derived through exposition, something that his approach apparently cannot allow.

In the end he also assumes that the suspensive guilt-offering is brought only for sin, and from this he proves that the rule requiring stringency is de-oraita. But, as we already noted, this is not necessary at all, because guilt-offerings are brought for corruption, that is, essence, and not for sin in the sense of rebellion against command.

Other Halakhic Consequences

After seeing how our explanation of Maimonides applies to the issue of doubt, we must now review all the other places in which halakha distinguishes between de-oraita and de-rabbanan, and examine how these are to be explained according to Maimonides on the substantive plane, without using the concepts of legal force and severity. In every such context we must find a substantive explanation of the halakhic rule in question, according to the nature of the laws involved, and not necessarily according to their severity. In all these cases there is room to distinguish between the different types of de-rabbanan law, depending on the issue.

Several such items are listed in Nahmanides’ glosses on Maimonides’ first principle, pp. 18-24. There Nahmanides brings all the places where the sages distinguish between de-oraita and de-rabbanan. There this material is brought in order to challenge Maimonides: why should there be a difference if, on Nahmanides’ understanding of Maimonides, all rabbinic commandments are really derived from “You shall not turn aside,” and thus are all de-oraita? Here we must review all those places and ask what implications they would have for the de-rabbanan laws of the types that have emerged here, which, for convenience, we have called divrei sofrim. Will the treatment appropriate to de-rabbanan laws also be applied to laws derived through exposition and to halakha le-Moshe mi-Sinai? And in what respects will they be treated like de-oraita laws? We begin precisely with the law of the rebellious elder, which is central to this issue.

2. The Law of the Rebellious Elder

It is an accepted rule, see Babylonian Talmud, Sanhedrin 88a, that the law of the rebellious elder applies only if he ruled against the court on a matter whose deliberate violation incurs karet and whose inadvertent violation requires a sin offering. See Maimonides, beginning of chapter 4 of the Laws of Rebels. It also applies in a matter that can lead to a practical difference in a case whose deliberate violation incurs karet and whose inadvertent violation requires a sin offering, such as monetary laws that can lead to adultery in the case of betrothal by money, whose deliberate violation incurs karet. See halakha 2 there. At first glance, such a law would seem to apply only to de-oraita laws, and Nahmanides and most commentators indeed assume this as obvious. But Maimonides, there in 4:1, rules that a person may become a rebellious elder even with respect to de-rabbanan decrees in a matter whose deliberate violation incurs karet and whose inadvertent violation incurs a sin offering, such as one who permits leaven on the eve of Passover during the sixth hour, or forbids benefit during the fifth hour.

On the straightforward reading, Maimonides does not distinguish between de-rabbanan and de-oraita laws, so long as the subject itself is tied to karet and sin offering, as in the case of leaven on Passover. This is in fact how Nahmanides understood him in his glosses on the first principle: Maimonides’ intention includes even the rabbinic prohibition concerning leaven, because the prohibition of leaven itself carries karet and sin offering. However, the Kesef Mishneh there explains the possibility of rebellious elder in the case of leaven during the sixth hour by saying that leaven is rabbinically forbidden for benefit at that time, and one who effects betrothal with it does not effect valid betrothal, as Maimonides himself rules there. This leads to adultery by Torah law in the case of betrothal by money, a matter whose deliberate violation incurs karet and whose inadvertent violation incurs a sin offering. Rabbi Elchanan Wasserman notes at the beginning of his Kuntres that according to the Kesef Mishneh it follows that, in Maimonides’ view, there is no rebellion in a genuinely de-rabbanan matter as such. Rather, sometimes even de-rabbanan laws can lead to Torah-level penalties of karet and sin offering, and therefore the law of rebellious elder applies to them too. According to this, the section on the rebellious elder was stated only with respect to matters whose consequences are de-oraita.

In principle, Nahmanides too could agree in the specific case of leaven during the sixth hour, because it clearly has de-oraita consequences. Yet in his glosses on the first principle he states that the law of rebellious elder does not apply at all to de-rabbanan prohibitions.

Nahmanides himself, p. 19, and likewise the Kuntres there, explains that the root of the dispute between him and Maimonides is the question whether the commandment “You shall not turn aside” includes de-rabbanan commandments or not. According to Maimonides it does include them, and therefore rebellion applies even there. According to Nahmanides himself, “You shall not turn aside” refers only to laws derived through exposition, and therefore the law of rebellion does not apply to de-rabbanan commandments.

But according to what we established in the first unit, Maimonides does not hold that in every de-rabbanan commandment a person violates the de-oraita prohibition “You shall not turn aside,” contrary to Nahmanides’ understanding of him. It therefore seems that we must explain Maimonides differently. According to Maimonides, the criterion in a matter that can lead to a case whose deliberate violation incurs karet and whose inadvertent violation requires a sin offering is that the immediate issue itself need not even be a law. Any situation in which there is a dispute with the Great Court, and from that dispute a de-oraita consequence of karet and sin offering can emerge, creates liability for rebellious elder. In principle, even if the elder disagreed with the court on a factual question, but that disagreement had consequences involving karet and sin offering, he would be a rebellious elder. In practice, however, one cannot disagree with the court about facts, because with respect to facts they have no authority and their ruling is not a legal ruling. Therefore we need some halakhic subject so that there will be a ruling and a possible disagreement about it, and so we require a dispute about some de-rabbanan prohibition. In other words, under Maimonides’ view there is no need to say that “You shall not turn aside” applies to de-rabbanan prohibitions. The point is only that there must be a disagreement with the court, and therefore it must concern a halakhic issue, not a factual one, and the disagreement must have consequences in a matter whose deliberate violation incurs karet and whose inadvertent violation requires a sin offering.

All this holds if we understand Maimonides according to the Kesef Mishneh, namely that in the case of leaven during the sixth hour there is liability as a rebellious elder because it leads to karet in the case of betrothal by money. But if we understand him like Nahmanides, namely that the basis of the rule is that leaven during the sixth hour is a sub-clause of the laws of leaven, and the subject of leaven is in principle one that involves karet and sin offering, and therefore rebellion concerning leaven during the sixth hour counts as rebellion, then our explanation would not work. According to Nahmanides, one who rules permissively on leaven during the sixth hour is liable as a rebellious elder even if no actual consequence involving karet and sin offering follows, for example in the case of betrothal by money. In his view, the de-rabbanan prohibition as such generates the liability of the rebellious elder.⁶⁷

Even here, however, this is not exactly an identification of de-rabbanan with de-oraita, because rebellion applies only in subjects that are sub-clauses of a de-oraita commandment carrying karet and sin offering. In other words, this approach understands such prohibitions as appended to, and forming part of, the general law of leaven, and therefore the law of rebellion applies to them.⁶⁸

What about laws derived through exposition or halakha le-Moshe mi-Sinai? Does the law of rebellious elder apply to them? According to Maimonides it seems clear that it does. Even though he classifies such laws as divrei sofrim, they are certainly no less than outright de-rabbanan laws. According to the Kesef Mishneh on Maimonides, rebellion applies only in a case that leads to a penalty of karet and sin offering. According to Nahmanides, the penalty applies when the law created through exposition belongs to a de-oraita “heading” under which karet and sin offering apply. According to Nahmanides himself, however, it is obvious that rebellion applies to laws that emerge from exposition, because these are de-oraita laws in every respect. As we saw, with respect to them all agree that the prohibition “You shall not turn aside” applies.

Nahmanides, in his glosses on the first principle, pp. 22-24, raises the objection that according to Maimonides it follows that the law of rebellious elder applies even to de-rabbanan prohibitions. Yet this entire law is learned from the bull brought for communal error, see Babylonian Talmud, Sanhedrin 87a, and there it is clear that liability exists only for a ruling on a genuinely de-oraita matter involving karet and sin offering, see Babylonian Talmud, Horayot 8a. It seems that precisely because of this, the Kesef Mishneh felt compelled to interpret Maimonides in his way, even though the plain language of Maimonides points more toward Nahmanides’ understanding of him. Compare our interpretation above.

Nahmanides there also cites the view of Rabbi Yehudah, Babylonian Talmud, Sanhedrin 87a, that one can be liable even with respect to tefillin, though tefillin is a positive commandment, because its basic rule is from the Torah and its explanation from divrei sofrim. Nahmanides adds that the phrase divrei sofrim there does not mean de-rabbanan decrees and enactments, but rather interpretations from the Oral Torah, such as the sages’ derivations through the thirteen hermeneutic rules, that is, explanations of the sages, or halakha le-Moshe mi-Sinai, that is, tradition that we received from the sages.

Nahmanides brings from Babylonian Talmud, Horayot 4a, that any matter acknowledged by the Sadducees does not generate liability for rebellious elder, because that is not a judicial ruling. One may simply say, “Go and read it in the schoolhouse.” Only matters not acknowledged by the Sadducees generate the law of rebellious elder, because only those count as judicial rulings. Nahmanides concludes:

That is, one is liable only for an interpretation of the sages within words of Torah, and not for a matter whose very basis is the words of the sages.

At first glance, this discussion is a splendid proof for Maimonides’ view. We see here that things explicitly stated in the Torah are not judicial rulings at all. Of course the prohibition “You shall not turn aside” was not said about them, because that prohibition concerns obedience to the rulings of the court, and this is not the court’s ruling. That is not merely because the matter is obvious, but because it is a ruling of the Torah and not of the court. The rulings of the sages include only interpretations through the thirteen hermeneutic rules or the tradition of halakha le-Moshe mi-Sinai. These are exactly the two categories that Maimonides brought in the second principle under the term divrei sofrim. And that is also his language in the Commentary on the Mishnah to Kelim, which we saw above, where he says that the phrase divrei sofrim has two meanings:

When they say divrei sofrim, they mean either that the matter is a received tradition of the sages, like all the interpretations and laws received from Moses, or a scribal enactment, like all enactments and decrees. Remember this as well.

If so, it is clear that laws generated through exposition are precisely the sort of laws to which the law of rebellious elder applies, and the same is true of halakha le-Moshe mi-Sinai. In other words, these are de-rabbanan laws, yet the law of rebellious elder applies to them. The reason is that the law of rebellious elder applies precisely to those things that are not included in the Torah, because things explicitly stated in the Torah are not rulings of the sages but simply matters one can go and read in school. The law of rebellious elder has nothing to do with the severity of the prohibition, and not even with its essence. What it requires is a ruling, so that there is something against which to rebel. This is a question of type and not of severity.

According to our approach, an additional important halakhic consequence follows from this. With regard to laws of type 1, those received by tradition and having a midrashic anchor, it appears that even though these are de-oraita laws, the law of rebellious elder should still apply to them. One who disputes them is in effect claiming that they are not written in the Torah. That means he is disputing the sages and not the Torah directly, and therefore he counts as a rebellious elder.

To summarize: this is an example of a unique parameter that generates a difference between de-oraita and de-rabbanan laws not by virtue of severity, and certainly not by virtue of any sweeping classification of laws as de-oraita rather than de-rabbanan. The discussion of the rebellious elder is an important proof for Maimonides’ approach here, and even the terminology used in that discussion matches Maimonides’ terminology. The phrase divrei sofrim in that sugya refers to laws derived through exposition and halakha le-Moshe mi-Sinai, and not to decrees or enactments. As noted above, even Nahmanides would agree to that here.

3. Punishment and Warning for de-rabbanan Laws

As is well known, one does not impose punishments such as lashes, death, and the like for de-rabbanan prohibitions, with the exception of excommunication. See Nahmanides, near the end of p. 20.⁶⁹ What about laws learned through exposition? Here the question depends on how one understands punishment. Punishment in halakha can be understood in terms of deterrence, or punishment for the sake of atonement, and so forth.⁷⁰ Different explanations of punishment may dictate different scopes of application. For example, if the reason is deterrence, what matters is the public tendency toward the offense and not only its severity. If the reason is atonement, what matters is the severity and nature of the transgression, because the atonement must correspond to the defect. Or, put differently, one may ask whether punishment is a repair for the defect created by the transgression, or punishment for rebellion against God. The consequences of this for our purposes are obvious.

With respect to commandments learned through exposition, we saw that in substantive terms they are just as serious as commandments from the Torah. The defect created by transgressing them may be on the same level. The difference lies in the question whether there is a command or a warning. Therefore punishment certainly belongs to them as well, though there may still be room to be lenient because there is no warning written in the Torah. Regarding halakha le-Moshe mi-Sinai, the matter depends on how one understands that concept. If it is command without essence, then there is no parameter of substantive severity, only rebellion. If it is essence without command, then the situation resembles what we described regarding laws derived through exposition.

As we already noted, Maimonides himself addresses this issue explicitly in his introduction to the Book of Commandments, after the fourteenth principle. There he writes that a warning must be explicit in the Torah, like a matter that the Sadducees acknowledge. Apparently, if it is not explicit, a person can claim that he did not know, because he was not sufficiently learned. Maimonides adds there, apparently for that reason, that a warning learned through exposition is not a warning for purposes of punishment, even though the prohibition certainly exists. But if the punishment is explicit in the text, then a warning learned through exposition is sufficient. We already explained his words by saying that when the punishment is explicit in the text, the matter is considered written in the Torah, and therefore there is warning and one may punish. According to our approach here, however, one may explain the matter somewhat differently. Presence in the Torah means that both essence and command are present, and both are required for punishment. That is, presence in the Torah is required not only because punishment requires prior warning, but in order to define the type of transgression for which punishment is imposed.

With respect to decrees and enactments, the situation is different. There may indeed be a warning in the Torah, by virtue of “You shall not turn aside,” though this is not necessary, because according to our conclusion in the first unit that prohibition does not directly prohibit de-rabbanan transgressions. But the severity of the transgression is lower. In other words, the term de-rabbanan in its ordinary sense also serves as a criterion of lower severity. Where the term de-rabbanan in Maimonides indicates source rather than severity, there would be room for punishment, and the only problem would be lack of warning. If so, the reason one does not punish for decrees and enactments is not because they are de-rabbanan laws, but for another reason, such as absence of command or absence of essence. By contrast, if a person violates a de-rabbanan rule as an act of rebellion, that is, from principled rejection of the sages’ authority, then he violates the de-oraita prohibition “You shall not turn aside.” In that case he should in principle have been liable to lashes, were it not for the fact that this prohibition serves as the warning for the death penalty of the court, in the law of the rebellious elder. At the beginning of the first unit we cited Maimonides’ explicit statement to that effect.

In places where we do find punishment for de-rabbanan prohibitions of the type of laws derived through exposition or halakha le-Moshe mi-Sinai, we will have to explain that the punishment is explicit in the Torah, or that the law learned through exposition or halakha le-Moshe mi-Sinai is merely a detail within an already existing de-oraita commandment, and is therefore itself de-oraita. This is like a matter whose basic rule is from the Torah and whose measure, or some other detail in it, is from divrei sofrim, as we saw in the previous section.

Thus, precisely with respect to punishments, a sweeping rule does emerge. In principle, all de-rabbanan laws, such as enactments, decrees, laws derived through exposition, and halakhot le-Moshe mi-Sinai, do not generate punishment. It may be that the reason is one and the same: punishment requires warning, and as far as warning is concerned all de-rabbanan commandments are equal, because warning must be written in the Torah. If so, we can explain why Maimonides cites specifically the issue of warning as a consequence of the principle presented in the second principle, and not anything else. Only this consequence is sweeping, and therefore it can be stated with respect to all kinds of de-rabbanan laws. But one could also explain the absence of punishment by different reasons in the different cases, even if in the end the practical result is the same: no punishment is imposed for any type of de-rabbanan transgression.

4. A Decree Upon a Decree

In negative commandment 135 Maimonides writes that the prohibition against an uncircumcised person eating terumah, which is learned through a gezerah shavah from the uncircumcised person in the Passover offering, see Babylonian Talmud, Yevamot 70a, is counted as a de-oraita prohibition. The reason is that in Babylonian Talmud, Yevamot 72a, it is clear that this prohibition is from the Torah. The proof is that the Gemara there rabbinically forbids a meshukh, one whose circumcision has been drawn forward so that he appears uncircumcised, to eat terumah because he looks like an uncircumcised man. Maimonides proves from this that the uncircumcised man is forbidden to eat terumah by Torah law, while the meshukh is forbidden only by rabbinic law.

Rabbi Yerucham Fishel Perla, in his comments to the second principle, p. 20, remarks on these words of Maimonides that if we interpret him according to the Tashbetz, then they are utterly impossible to explain. According to the Tashbetz, a law learned through gezerah shavah is de-oraita, so why should the sages not make a decree because of it? And even if one were to say that this is a lighter kind of Torah law, it still certainly is not a de-rabbanan enactment, so it remains unclear why the sages should not make a protective decree concerning it. From here Perla proves that according to Maimonides laws derived through exposition are genuine de-rabbanan laws, as Nahmanides understood him, and only on that basis can one understand why the sages do not make an additional decree because of them.

It should be noted that from these words of Maimonides we find an additional halakhic consequence, beyond the consequence regarding warning that we already encountered, to the claim that laws derived through exposition are divrei sofrim: the sages do not make a decree because of them. In this respect they are like all de-rabbanan prohibitions. It should also be noted that the proof Maimonides brings here is halakhic and not linguistic. We prove from the Gemara that the force of the law is de-oraita, not merely that it belongs to the linguistic category de-oraita. That is precisely Perla’s difficulty.

In any event, at first glance Perla seems correct: this does appear to prove that Maimonides is talking about the legal force of laws derived through exposition. Can this also be explained according to our interpretation of Maimonides?

It seems that it can. According to our approach, a law derived through exposition is a law with essence but no command. If so, Maimonides can perhaps be explained in two ways:

1. He may hold that one makes a protective decree only with respect to a law that contains command. According to this, perhaps the sages would make decrees with respect to halakha le-Moshe mi-Sinai, depending on how we understand its nature.

2. Or he may hold that one makes a protective decree only with respect to a law that contains both essence and command. According to this, the sages would not make decrees with respect to laws derived through exposition, and likewise not with respect to halakha le-Moshe mi-Sinai.

Perhaps the matter can be explained more precisely. We should note that here we are dealing with a special kind of decree. This is not a decree lest in this very act one come to violate a de-oraita prohibition. Rather, it is a decree concerning a situation that resembles the de-oraita case, since the meshukh resembles an uncircumcised person. In other words, it is a decree because one may come to violate a de-oraita prohibition in some other case. It is not so plausible that the concern is that people may come to permit terumah or the Passover offering for a genuinely uncircumcised person. Rather, this is a rabbinic expansion of the de-oraita law. Compare our distinction in the first unit between different kinds of rabbinic decrees and enactments. If so, it is quite plausible that the sages would not make an expansive decree on the basis of a law that is itself an expansion of de-oraita. According to our approach, the principle that one does not make a “decree upon a decree” means that one does not make an expansion upon an expansion. If the original law is itself an expansion beyond what is in the verse, we do not add to it a further rabbinic expansion. The practical difference would be that a decree lest one actually come to violate that prohibition directly would indeed be made even in such a case. This is quite plausible, because it makes more sense to make a decree where there is concern that one may come to violate a prohibition that has essence, and thereby create real defect and corruption, than to make a decree concerning a pure matter of command, where no defect is created and only a violation by the person occurs.

In any event, what emerges is that Maimonides holds that the sages do not make decrees on the basis of laws derived through exposition. And this remains true even on our understanding of Maimonides, so Perla’s question need not trouble us.

It should be noted that in his code, Laws of Terumot 7:10, Maimonides rules that the meshukh is permitted to eat terumah. At first glance, he would seem to have retracted, perhaps precisely because he really understood the law derived through exposition there as de-rabbanan and not de-oraita. But this seems incorrect, because as we have already seen several times, he did not retract on this point. If one examines negative commandment 135 carefully, one sees that Maimonides did not necessarily intend to claim that the meshukh is forbidden to eat terumah. Rather, he wanted only to prove that there is an opinion that would decree and forbid him to eat terumah, and that alone is sufficient to prove that the law derived through exposition is not merely a de-rabbanan law. Even if it is a de-oraita law, it does not necessarily follow that the sages would in fact make a decree concerning it. Only the converse is certain: if it is de-rabbanan, then they certainly would not. See also Perla himself, who notes that the flow of the sugya proves that according to the conclusion there is in fact no decree concerning the meshukh, according to the version cited by the Kesef Mishneh and Mahar”י Korkos in Laws of Terumot there. If so, there is certainly no need to conclude that Maimonides retracted.

5. Dimensions of Leniency and Severity: An Oath Concerning Laws Derived Through Exposition⁷¹

In Babylonian Talmud, Nedarim 8a, it is stated that an oath to study Torah takes effect, even though an oath to fulfill a commandment does not take effect, because one already stands sworn from Mount Sinai, and an oath cannot take effect upon an oath. The early authorities offer several explanations of this passage. Some write that Torah study is not an obligation at all, apart from the recitation of the Shema in the morning and evening, and therefore this is not an oath concerning a commandment. See the Rosh’s commentary there. The Ran there writes that because the obligation to study Torah beyond the recitation of Shema morning and evening is learned through exposition, from the word “and you shall teach them diligently,” see Babylonian Talmud, Kiddushin 30a, the oath takes effect.

In explaining the Ran’s words, one can proceed in two ways.⁷² Some understood that the Ran holds, like Maimonides, that laws derived through exposition are de-rabbanan laws, and they assumed that Maimonides too intended this, following Nahmanides’ understanding of him and not that of the Tashbetz. Others wrote that the Ran means that laws derived through exposition are de-oraita laws, but lighter ones, and therefore an oath can take effect with respect to them. So writes the Kuntres, sec. 1, notes 19-20. See further discussion below. That would apparently correspond to the Tashbetz’s reading of Maimonides, according to which laws derived through exposition are de-oraita, except that the Ran would hold that they are lighter in severity.

But the plain reading is as we have proposed: these laws are de-rabbanan in the sense that there is no command concerning them, because they have no source in the Torah. The Ran explicitly writes that their force is de-oraita. His language there is:

From here it seems to me proof for what I wrote at the end of the chapter Shevuot Shetayim, that anything that comes through exposition, although it is from the Torah, since it is not explicit in the verse itself, an oath takes effect with respect to it.

The Ran is not speaking specifically about Torah study, but stating a general principle. It is clear from his words that laws learned through exposition are de-oraita laws, except that they are not explicit in the Torah.

If so, it appears that the common assumption of the commentators, that the reason an oath takes effect with respect to such commandments is their lesser severity, is incorrect. The explanation lies in their character and essence, not in severity or leniency. The rule that one oath does not take effect upon another appears in the Mishnah, Shevuot 27a. It is based on a technical principle: the second oath has no “place” in which to take effect, because the first oath is already there. This may be understood in two ways:

1. It follows from the general principle that one prohibition does not take effect upon another, as with all Torah prohibitions. This is the implication of Tosafot, “to fulfill,” in Shevuot 27a, and others. In that case the matter may, though not necessarily, involve severity and leniency.

2. Or one may understand it as a law unique to oaths: because we were already sworn at Sinai concerning the commandments, an oath has already taken effect upon all of them, and therefore another oath cannot take effect upon them. This is the implication of Tosafot, “commandment,” in Shevuot 27a, and others.

The practical difference concerns commandments that do not appear in the Torah, such as halakha le-Moshe mi-Sinai. On the one hand, according to most of the early authorities these are full commandments for purposes of the rule that one prohibition does not take effect upon another. On the other hand, it may be that we were not sworn upon them at Sinai, because they are not included in the Torah itself, and therefore an oath can take effect with respect to them. Again, this would not be because of any lesser severity, but because of their special character, namely that they are not written in the Torah. Even if, with respect to halakha le-Moshe mi-Sinai, we say that an oath does not take effect because these laws were indeed given to Moses at Sinai by God and therefore included in the Sinai oath, one can still say that laws derived through exposition, which do not appear in the Torah, are different. They were not given to us at all, but were created by the sages. This is apparently the Ran’s position.

In light of all that we have said, one can say that laws derived through exposition are not written in the Torah, and therefore Maimonides calls them divrei sofrim. If so, it becomes clear why an oath takes effect with respect to them. It should be emphasized that the fact that they are not written in the Torah is here not an indication of leniency, nor of the absence of warning, nor of the fact that the matter is a judicial ruling, as in the case of the rebellious elder. Rather, it indicates that we were not sworn on them at the revelation at Sinai, and therefore, since they were not included in that original oath, the later oath can take effect upon them.

As noted, according to this it is entirely possible that an oath would not take effect with respect to halakha le-Moshe mi-Sinai, because that law was given to Moses at Sinai. If so, for this purpose halakha le-Moshe mi-Sinai may be exactly like a commandment explicitly written in the Torah. Once again we see that in every context the matter must be judged on its own. There is no sweeping rule regarding the difference in legal force between de-oraita and de-rabbanan, and no sweeping rule regarding the halakhic consequences of that difference. The differences are differences of type, and they are not necessarily dependent on broad and general concepts.

It should be noted that this example differs from the earlier ones. Above we dealt with laws that apparently create difficulties for Maimonides. Here we are dealing with the Ran’s view, because the Gemara itself could have been explained differently. But we saw that in the Ran himself one can certainly understand the matter in line with what we have argued concerning Maimonides. Beyond that, this sugya yields another consequence to the question whether a given commandment is written in the Torah or not: whether an oath can take effect with respect to it.

It should also be noted that the Kuntres there brings several additional examples from the early authorities who claimed that laws derived through exposition are lighter. He explained this by a principle that the more space something occupies in Scripture, the more important and severe it is. He tried to prove this from the sages’ remark about Eliezer, Abraham’s servant, that the conversation of the servants of the patriarchs is more pleasing than the Torah of the descendants. See Rashi on the section of Chayei Sarah.

This principle is difficult. We do not find, for example, that the rule of marrying off the older daughter before the younger one, which is brought as law in Magen Avraham, Orach Chayim, is more severe than the prohibition of murder, which is commanded in two words. And there are many other examples where severity is not proportional to the length of the description in the Torah.⁷³ We are therefore forced back to our explanation: the consideration here is not severity but type. This fits exactly with our explanation of Maimonides’ method.

6. Additional Examples of Differences Between de-oraita and de-rabbanan

First, we will briefly bring here several additional examples, three brought by the Kuntres and two more, in which we find leniency in laws derived through exposition. After that, we will bring two more examples of differences between de-oraita laws and laws derived through exposition where no issue of severity or leniency is involved.

• A. In Tosafot, “Rabbi Ulla said,” Babylonian Talmud, Yevamot 7b, it is written that it is preferable to permit the metzora to insert his extremities into the Temple courtyard, because this is only partial entry into the courtyard and its prohibition is learned through juxtaposition, rather than to permit full entry, which is stated explicitly in the verse “He shall not enter the Sanctuary.” We thus see that the former is regarded as a lighter prohibition.

In truth, this does not fully accord with our explanation of Maimonides, because it seems to prove that a prohibition written explicitly is more severe than a law derived through exposition, whereas according to our explanation the difference is one of type. But Maimonides can explain the Gemara differently. One can say that a law derived through exposition is not in fact lighter, but that it lacks command, and that alone is sufficient to prefer permitting it rather than permitting the law explicitly written.

Several explanations are possible:

1. Even if there is no difference in substantive severity, command is also one of the parameters in the balance, and therefore it is preferable to permit only essence rather than both essence and command.⁷⁴
2. There may be considerations connected to preserving the honor of the Written Torah.
3. Perhaps this is an interpretive consideration. The very fact that the Torah explicitly says that the metzora must insert his extremities proves that the exposition does not apply to this case. The matter is therefore viewed as fully permitted and not merely as overridden.

All of these explanations could also be correct with respect to halakha le-Moshe mi-Sinai, but this is not the place to elaborate.

We may also note that in Nahmanides, in the passage cited above, see pp. 19, 20, and 22, we find several leniencies that are applied to de-rabbanan laws more readily than to de-oraita laws,⁷⁵ and perhaps those leniencies really would apply as well to laws derived through exposition for reasons of this kind. But this is not the place.

• B. In the Mishnah, Zevahim 9a, it is explained that the sin offering takes precedence over the guilt offering because its blood is applied to the four horns and to the base of the altar. Tosafot Yom Tov there asks that the guilt offering also requires the base of the altar, and explains that in the sin offering this is explicit in the text, whereas in the guilt offering it is learned by exposition. Rashi there, “and to the base,” also hints at this. Therefore, for this purpose the sin offering is considered holier than the guilt offering.

Again, according to Maimonides it does not seem that the obligation in the sin offering is more severe. Rather, there is a rule of precedence, as though the obligation in the sin offering were holier. Again, the rule of precedence can be explained in various ways according to Maimonides, as above, and not necessarily on grounds of severity.

It is certainly possible to say that with respect to halakha le-Moshe mi-Sinai, a law explicit in the verse would not necessarily take precedence, since in both cases we are dealing with an explicit divine will and therefore they are equally holy.

• C. In Tosafot, “The verse says,” Babylonian Talmud, Kiddushin 15b, and in Gilyon ha-Shas there, which also cites Rashi on Babylonian Talmud, Berakhot 21a, “which does not require,”⁷⁶ and from there proceeds to Babylonian Talmud, Bava Metzia 88b, “as for wheat and barley, threshing floor is written there explicitly,” it is shown that one may draw an a fortiori inference from what is not explicit to what is explicit. There, however, it does not seem that the point is that what is explicit is more severe.

According to one interpretation in Tosafot there, the Gemara reasons as follows. One sold to an Israelite does not have explicit mention that he is redeemed through relatives, yet he nevertheless goes free after six years. Therefore, one sold to a gentile, with respect to whom Scripture explicitly says that he is redeemed through relatives, should all the more so go free after six years. Thus we are not dealing here with a commandment, but with a possibility of redemption, and it is therefore impossible to speak in terms of severity and leniency. Apparently the issue is who more properly belongs within the framework of redemption, and the inference assumes that what is stated more explicitly belongs there more centrally.

The explanation is probably that there are several possible reasons why God chose to write some laws and not others. According to the explanation of the Kuntres, based on severity, the matter is simple. But according to Maimonides, who does not see the issue in terms of severity, there must be some other consideration. It may be one of the factors in deciding whether to write a law or not, depending on how central a certain quality is to the subject under discussion.

• D. Babylonian Talmud, Kiddushin 24b, concludes, following “those who decide before the sages,” that if one knocks out his slave’s tooth or eye, the slave goes free without a bill of release. But if he injured one of the other extremities, a bill of release is still required, because that law is learned from the sages’ exposition. Here we see that a law derived through exposition is treated more leniently than a de-oraita law, that is, its force is lower. An injury that frees the slave by virtue of the sages’ exposition cannot terminate the slave’s status without a bill, whereas an injury recognized de-oraita terminates the slavery immediately and without a bill. See the discussion in Responsa Rashba, part 1, sec. 1185, and below in Appendix C at the end of the book.

• E. In Babylonian Talmud, Yoma 34a-b, a baraita is cited containing a tannaitic dispute about the meal offering and libations. According to the first tanna, the verse “and its libation, a quarter of a hin,” in Numbers 28, refers to the morning daily offering, and from it one learns by juxtaposition to the afternoon daily offering. According to Rabbi, the reverse is true: the verse refers to the afternoon daily offering, and one learns from it by juxtaposition to the morning daily offering.

Tosafot there, “Rabbi,” writes that the practical difference between them arises in a case where they have libations sufficient for only one offering. According to the first tanna, they should be brought with the morning daily offering, because that is the one explicitly mentioned in Scripture. According to Rabbi, they should be brought with the afternoon daily offering, because that is the one explicitly mentioned in Scripture according to him. What both sides share is the assumption that what is explicit in Scripture is weightier than what is learned through juxtaposition. Even though the two are linked to one another, the one explicit in Scripture is weightier and takes precedence over the one inferred from it.

• F. In Babylonian Talmud, Shabbat 40a, Rava states that one who violates a de-rabbanan law may still be called an offender. The very fact that this had to be stated implies that one might have thought that only one who violates de-oraita law can be called an offender. At first glance, this seems to depend on the severity of the prohibition, and therefore it would not apply to laws derived through exposition or halakha le-Moshe mi-Sinai. According to that reading, Rava’s novelty is that even though de-rabbanan prohibitions are less severe, they are still serious enough for one who violates them to be called an offender.

But according to our approach to Maimonides, one might instead have explained that an offender is only someone who violates a prohibition accompanied by an explicit warning in the Torah. Such a person rebels and does not merely create a defect. According to this, the idea should also be relevant to laws derived through exposition, and perhaps also to halakha le-Moshe mi-Sinai, depending on whether one sees command there. We saw above that this is disputed. As we saw, without command and warning there is no punishment, and one might therefore have thought that the transgressor would not be called an offender. There is no element of rebellion here, and so he should not bear that title. But in the end that is not so. He too is called an offender. That shows that command is not the criterion for the title.

• G. In Babylonian Talmud, Berakhot 20b, it is explained that one who is obligated only by de-rabbanan law cannot discharge the obligation of one who is obligated by de-oraita law. The question is whether this depends on the strength of the obligation. One could say that every commandment contains two obligations: to obey God’s voice and to fulfill the content of the command, that is, to repair what requires repair. Accordingly, one who is obligated under one type of obligation cannot discharge one who is obligated under the other, not necessarily because of severity or leniency, but because type is decisive. According to this, for example, one obligated by an explicit verse could discharge one obligated by exposition or by halakha le-Moshe mi-Sinai, but not vice versa. More than that, one must ask whether one obligated by exposition can discharge one obligated by halakha le-Moshe mi-Sinai, or the reverse. If one understands halakha le-Moshe mi-Sinai as command without essence, then neither side can discharge the other. If one understands it as essence without command, then each can discharge the other.

It is possible, however, that discharging another’s obligation is different, and that specifically the plane of command is decisive there. But this is not the place to elaborate.

Chapter 5: Three Types of Operation of the Hermeneutic Tools⁷⁷

Introduction: Can Expositions Yield Lenient Results?

To conclude the discussion of Maimonides’ view, we cannot ignore a serious problem that exists within it, and we must ask whether the explanation we have proposed succeeds in resolving it. There are quite a few cases in which exposition uproots a de-oraita law, or qualifies it. Put differently, the conclusion reached through exposition is not stricter than the plain meaning, but more lenient. Here are a few examples:

1. The exemption of women from time-bound positive commandments is learned through juxtaposition. See Babylonian Talmud, Kiddushin 35a.
2. In Babylonian Talmud, Zevahim 57a, the Gemara learns through exposition that the firstborn offering may be eaten for two days, and not only for one.
3. The law permits a mixture of wool and linen in tzitzit by virtue of juxtaposition. From here one also learns that a positive commandment overrides a prohibition. See Babylonian Talmud, Yevamot 4a.
4. Rabbi Eliezer learns through exposition that the preparatory acts of a commandment override the Sabbath. See Babylonian Talmud, Shabbat 131b.
5. “An eye for an eye” means monetary payment. Here too there is leniency and not only stringency. The court does not carry out its duty to punish the assailant as the Torah’s command would suggest. It is admittedly unclear whether this is an exposition at all or rather plain interpretation, because no explicit hermeneutic rule is used here.
6. There are many expositions that seem to create stringency, but their consequences include leniency. For example, expositions that disqualify certain witnesses make the law of adultery with a married woman more lenient in the case of a woman betrothed through such witnesses. Another example: any exposition that makes prohibitions stricter in areas involving lashes or death makes the law of bodily injury more lenient,⁷⁸ and where a sacrifice is involved, it makes the law of unconsecrated slaughter in the Temple courtyard more lenient. And so on; there are many more examples of this sort.

If an exposition comes to create some new law or new stringency, one can say that its status is that of divrei sofrim. But if it comes to create leniency, then we have a de-oraita obligation and a leniency that comes from divrei sofrim. What, then, is the status of such a law?

Initial Considerations

Before explaining the matter, let us briefly analyze the situation according to the various approaches to Maimonides. If laws derived through exposition are de-oraita laws, as the Tashbetz and his school maintain, then matters are simple. There is no difficulty in the fact that exposition shapes the boundaries of de-oraita law. As we saw, this is indeed the accepted conception of halakhic exposition. But if laws derived through exposition are de-rabbanan laws, as Nahmanides and his school understood Maimonides, and as is our own conclusion concerning Maimonides, then matters become very difficult. As we saw, Maimonides is indeed unwilling to punish on the basis of a warning learned through exposition, if the punishment is not explicitly written in Scripture. In other words, he really does hold that exposition does not create a de-oraita prohibition of ordinary force. If so, it is unclear how the sages can come and uproot severe de-oraita laws, at times even through positive action.

Several possible explanations are available:

1. Perhaps these laws were transmitted by tradition, and therefore the expositions concerning them are supportive rather than creative, or revealing rather than expanding. In that case they are de-oraita laws even according to Maimonides.
2. Perhaps these expositions are not made through the thirteen hermeneutic rules or through inclusion, and therefore they count as explicitly written in Scripture. Once again, they would be de-oraita laws.
3. Perhaps there is some other local consideration that makes them de-oraita. For example, perhaps the exposition only creates the warning, while the punishment is explicit in the text, as Maimonides says in the introduction to the Book of Commandments. Or perhaps the law learned is only a detail within an already existing de-oraita commandment, and therefore it too counts as de-oraita. This would be similar to what we suggested above regarding betrothal by money.⁷⁹
4. Sometimes one can find a local explanation that resolves the tension between Maimonides and the Talmud without reference to Maimonides’ unique theory. For example, perhaps Maimonides does not rule in accordance with the particular sugya that derives these laws through exposition.⁸⁰

Clearly these possibilities do not necessarily exclude one another. In every case we must examine the matter individually, and it is entirely possible that each case will have a different reason and a different explanation from among these four possibilities.

And what according to our explanation of Maimonides? Must we also say that in all these laws the sages had a tradition? Recall that Maimonides himself writes in the responsum mentioned above that there are only three or four such cases. Or must we choose one of the other possibilities listed above? This is not an easy question to decide, and we will try to clarify somewhat the difference between this approach and Nahmanides’ approach.⁸¹

We proposed here that according to Maimonides the tools of exposition are expanding tools rather than revealing tools. These tools were indeed given to us at Sinai, but they were given as tools of construction with which we add further stories to what is present in the Torah. The status of laws generated through such processes is not that of ordinary de-oraita law, but neither is it that of de-rabbanan law in the ordinary sense. As stated, Maimonides does not accept this sweeping conceptual and halakhic dichotomy at all. This is another type of law, which in his thought is divided into five types, to which we added three more. It has its own defining features, and not necessarily a single level of severity that characterizes it. It may include several levels of severity, and every halakhic consequence must be examined on its own, on substantive grounds and not necessarily on grounds of severity.

Let us take as an example the exemption of women from time-bound positive commandments. According to this intermediate approach, can we say that the sages, through their expositions, can intervene on the de-oraita plane even toward leniency, by exempting women from commandments in which, absent the exposition, they would seemingly be obligated? That would differ from the conception of laws derived through exposition as straightforward de-rabbanan laws.

First Explanation: Dividing the Above Examples into Two Types

It seems that such a possibility does exist, though the matter is far from simple. First of all, there is a strong intuition that since we are not dealing with de-rabbanan enactments in the ordinary sense, but with something that is part of the divine will itself, there is greater room to say so. Let us explain this more clearly. Above we explained the difference between de-rabbanan enactments and expositions. Enactments stem from the sages’ own will and from the pressures of reality that call for intervention. Expositions, by contrast, are supposed to be made even if there is no concrete motivation for the content of the laws themselves. Even if the sages see no value at all in reverence for Torah scholars, they will still have to expound the word “et” in some way, at least in certain cases, such as a free term, and so on. Thus expositions are actually the result of a constraint or direction that comes from the Torah itself, and not something that begins with the sage who expounds, as in the case of enactments.⁸² The sage merely has to decide what to do in a context that demands exposition. Put differently, the Torah itself, by means of a free term or some other textual context that demands exposition, hints to us that some action is needed here, while leaving us the choice of what action to take, or what to derive.

If so, in that sense exposition is not an expression of the sages’ own will, but an expansion that leads to an understanding of the divine will, though beyond what is expressed in the Torah itself. The decision in which direction to expand what appears in the verse is indeed entrusted to the sages’ reasoning, but the result is supposed to capture God’s will and not to express the sages’ will. It is therefore reasonable to say that this kind of “interpretation” can indeed intervene in what is written in the Torah and shape the boundaries of de-oraita law. In the final accounting, God’s will is the totality composed of what is explicitly written in the Torah plus the layers that the Torah entrusted to the sages to expand.⁸³ What emerges at the end is not necessarily de-oraita law, because according to Maimonides de-oraita means what is written in the Torah and not what is produced by exposition. But it is certainly part of the divine will. If so, there is no principled problem with such exposition producing leniencies in de-oraita laws written in the Torah and even uprooting something from them, as in the case of women’s exemption from time-bound positive commandments.

Put differently, one can connect this to what we saw above: the notion of de-oraita law in the sense of legal force, which is what creates the apparent difficulty regarding lenient additions learned from exposition, simply does not exist in Maimonides’ conceptual system. Once that notion falls away, the difficulty falls away with it.

Second Explanation: Different Types of Operation of the Methods of Exposition

If we look closely at what characterizes the six examples above, we will see that they are not similar to the expositions we discussed when analyzing Maimonides’ second principle. Here we encounter an additional mechanism, different in essence from the previous one, for explaining the possibility of lenient expositions. We will distinguish here between different kinds of expositions and present a somewhat different picture of Maimonides’ determinations in the second principle.

We begin from the law of warning derived by logical inference. As recalled, this is the main halakhic example that Maimonides himself brought to illustrate the principle of the second principle. The issue of warning always concerns the question whether there is or is not a prohibition here, and not the definition of some detail within an already existing prohibition. If so, when a warning is learned through exposition, the law learned from it will be a law of divrei sofrim. But such a law is not more lenient than something already known from the text of the Torah. Rather, it creates an entirely new law. The warning is a new prohibition derived through exposition. Indeed, where the punishment is already written in the Torah, Maimonides innovates that even warning learned through exposition can produce a Torah prohibition, because it joins something that the Torah itself has already revealed to us as existing. In that case the exposition does not create something out of nothing.

Take as another example the exposition “You shall fear the Lord your God,” from which the sages derive reverence for Torah scholars. This exposition also creates a new commandment out of nothing. It does not make the law more lenient or more stringent relative to something written in the Torah. Rather, it invents or creates something new. Here the issue is not one of authority at all, but at most of the prohibition against adding. Yet that difficulty certainly does not arise with expositions, because we received at Sinai permission to use exposition, and that certainly is not included in the prohibition against adding to what was given to us.

Another example is the prohibition against an uncircumcised person eating terumah, negative commandment 135, where again the exposition creates a new prohibition. In that context too, the issue of leniency relative to what is written in the Torah is not relevant. The same is true of all the other examples to which Maimonides refers under the principle of the second principle. In the case of betrothal by money, one may hesitate whether this is a detail within an already existing commandment or legal system, even from the standpoint of prohibitions, or whether a new and additional mode of betrothal has been created. Above we saw that the fact that the halakhic status of betrothal by money is de-oraita can be understood according to either possibility.

But in situations where the exposition deals with some detail within an already existing prohibition, it is difficult to assume that the exposition is only an additional layer. In such situations, exposition interprets the existing command that is written in the Torah. Therefore, in such cases it will be part of de-oraita law, and it will be treated as though it too were explicitly written in the Torah. For example, we saw regarding betrothal by money that because it is part of the commandment of betrothal, and we even inferred this from Maimonides’ language in Sefer ha-Mitzvot, positive commandment 213, its status is de-oraita, even though it is learned through exposition. Another example is Maimonides’ statement in the Commentary on the Mishnah to Kelim 17:12, where we saw that he rules that if a commandment’s basic rule is in the Torah and its measure is from divrei sofrim, one must be stringent in cases of doubt. This means that if a law has a de-oraita root, then even its measure, which comes from halakha le-Moshe mi-Sinai and is divrei sofrim in Maimonides’ system, is considered part of the boundaries of that de-oraita commandment, and therefore must be treated stringently in doubt. This too is an indirect indication that a law of divrei sofrim can attach itself to de-oraita law. Here we again see Maimonides’ general tendency not to classify matters mechanically according to rigid rules, but to decide on the basis of substantive and specific considerations, each case on its own.

Put differently: we saw that laws derived through exposition are laws that contain repair or corruption, that is, essence, but no command. Yet when they spell out something within an already existing commandment, the command that applies to the commandment as a whole applies to everything under it, while exposition merely specifies the essence that lies beneath the command. Therefore, in such cases there is no problem even if the exposition yields leniency, because the command directs us toward the essence that lies beneath it, and exposition is one of the ways of discovering what the true essence is in such cases. The assumption is that exposition reveals real and correct essences. The tools used to do so are hermeneutic tools rather than plain-sense interpretive tools. If so, it is clear that the command applies to the essences disclosed through exposition, and not to what may appear to us, on a superficial reading, to be the simple meaning of the Torah’s text.

In all the examples we brought above, where exposition yields leniency compared with what one would think from the Torah’s wording, we are dealing with one of two types of cases:

1. There are cases where the exposition establishes a principle that is a detail within an already existing commandment. This is precisely what leniency means here: there already existed a statement that would have led to a stricter result, and the exposition comes to qualify it. As we saw, in such cases there is no problem in saying that exposition reveals the true essence that lies beneath the existing command. For example, the rule that a firstborn offering may be eaten for two days. Here the law introduced a detail within an existing rule, and it is obvious that this is God’s will and therefore the true essence. In such a case it is clear that the command to eat the firstborn refers to this essence. There is therefore no difficulty in saying that exposition can lead to leniency relative to what we would have done without it, namely eating it for only one day.⁸⁴

2. In the other examples brought above, the expositions establish a meta-halakhic principle. A meta-halakhic principle is not a “leniency” in the usual sense of the term. It is a general and external interpretation of Torah law. True, such an interpretation can also result in leniency, but it can certainly be made. The reason is that we do not possess a clear statement in the Torah requiring stringency. Rather, there is simply an absence of a statement, which, without exposition, we would have interpreted stringently. Exposition reveals how we ought to understand the law, and therefore in such cases too it can yield leniency. For example, the exemption of women from time-bound positive commandments is an overarching meta-halakhic principle. It is not a detail within some existing commandment, and not even a collection of details within a group of commandments of that kind. It is a rule from which all those details are derived. Yet it is obvious that the leniency that follows from it expresses the divine will. Surely Maimonides too would not say that the Torah really wants women to be obligated in these commandments and that the sages merely added another layer and expanded what is written in the Torah. Here the sages are engaged in interpreting the divine will itself.⁸⁵

What these two types of situations have in common is that in both of them the drash (midrashic interpretation) deals with level one, not with the expansion of level one and the construction of level two. An expansive midrashic interpretation is one that takes what is written in the Torah—for example, the mitzvah (commandment) of fearing God—and adds an additional level to it, such as a commandment to revere Torah scholars. Here there is room to discuss whether this is included within the written commandment or added to it as an extension, or perhaps whether it is even an independent commandment, with the written commandment merely revealing that it exists. But when the discussion concerns the exemption of women from positive commandments that are time-bound, there is no prior level one. True, there is the ruling we would have reached even without the midrashic interpretation, but the interpretation itself determines the basic content, not some additional content. In such cases it can certainly be lenient. The same is true in the case of the firstborn offering that is eaten for two days: the interpretation determines the basic level and does not add an extra layer as a higher tier. Therefore, here too there is no problem in deriving leniencies from midrashic interpretations.

We can explain this from another angle. We saw above that the consideration that led Maimonides to conclude that laws derived by midrash are laws of divrei soferim (rabbinic law) stemmed from his theory of interpretation. His basic assumption is that there cannot be two interpretations of the biblical text. Therefore, if there is a peshat (plain-sense) interpretation, then the midrashic interpretation is not an interpretation but an expansion. It therefore receives the status of divrei soferim, since it is not found in the Torah itself. But in the two types of cases we have encountered here, this is not the situation. In these cases there are not two interpretations competing over the same text. There is only one interpretation, and it is the midrashic one. In such a case there is no reason to assume that its status is that of rabbinic law. Even if we say that this itself is the correct interpretation of the Torah, it is the only interpretation, and this is entirely compatible with Maimonides’ interpretive conception.

As noted, the midrashic tools of expansion were given to us at Sinai. But we have seen that sometimes those very same tools serve revelation rather than expansion. Usually the Sages use received tradition to determine whether a given act of midrashic interpretation is revealing or expansive. But there is also the possibility that they will decide this on the basis of their own understanding of the interpretation itself, as in the approach of the author of Halakhot Gedolot and Saadia Gaon, which we shall see in Appendix C at the end of the book. If the interpretation appears to them to reveal the meaning of the verse itself, then the halakha (Jewish law) that emerges from it will be a de’oraita law (of Torah authority), one that addresses and interprets the verse itself. And this is so even without a tradition. In fact, perhaps such a case describes a plain-sense interpretation rather than drash. The more plausible it seems to us as an interpretation of the verse, the more entitled it is to be called peshat.

It should be emphasized that the “working tools” may be precisely the same tools. The question is how to regard the result obtained by using those tools. Reflection on it may lead to the conclusion that it reveals the Torah’s own content, or that it expands the Torah’s content. The tools in themselves do not determine the halakhic and essential meaning of the result. Something similar appears in Nahmanides’ approach, which views all the tools of drash as tools of revelation rather than tools of expansion, even though he was certainly aware of their analogical character.

Summary: Three Different Contexts in Which the Hermeneutic Tools Operate

To summarize, there are three contexts in which midrashic interpretations appear:

1. A change—addition, subtraction, or otherwise—in a detail within an existing mitzvah.
2. The determination of a meta-halakhic principle.
3. The creation of a new obligation, and perhaps the removal of an existing obligation.

Only in the third case is it relevant to apply the principle established by Maimonides in the second root, and to treat the novel laws created by midrashic interpretations as laws of rabbinic authority. In the first two situations, the midrashic interpretations can certainly change the existing law even in a lenient direction, as explained above, and they take part in shaping de’oraita itself. In the first case, the interpretation will yield a de’oraita rule, as though it were written explicitly in the Torah, since both the command and its substantive content are present, as in the case we saw where there is both tradition and midrashic interpretation. In the second case, the interpretation does not create something that is de’oraita but rather something of rabbinic provenance; however, that determination has no significance. The issue here is not the binding force of an obligation, since no mitzvah is being created. This is a rabbinic interpretation of the Torah’s will, and therefore it is obviously binding even when it yields lenient conclusions.⁸⁶

This picture may also explain the surprising fact that the implications of Maimonides’ sweeping principle in the second root have such rare halakhic applications. Although most Torah laws are learned from midrashic interpretations, only very few places in the Mishneh Torah reflect an application of the determination that these are laws of rabbinic authority. According to our approach, one may say that only those applications of midrashic interpretations that create novel laws are relevant to this principle. And even there, this is true only if we do not have a tradition establishing that the law is nevertheless of Torah authority. The other two types of application of midrashic interpretations are simply irrelevant to the principle of the second root.

The conclusion that emerges from this is that Maimonides’ sweeping principle, despite the conceptual and theoretical revolution concealed behind it, has rather limited practical implications. Even according to Maimonides, midrashic interpretations will usually yield laws of Torah authority. Only interpretations of the type that innovate a mitzvah belong to the category under discussion in the second root. When Maimonides writes in his responsum that there are only three or four cases of midrashically derived laws that count as laws of Torah authority, he means only three or four cases from among the interpretations that belong to the third category. There are many additional interpretations that yield laws of Torah authority even without a tradition. As we have seen, the sage who expounds may classify the halakha that emerges from the midrashic interpretation as de’oraita on the basis of reflection on the meaning and character of that halakha as well. It appears that a considerable portion, if not the majority, of the midrashic interpretations known to us belong to these categories.

Part Three

The Synthetic A Priori in Legal Theory

This part is the heart of the book, and it contains the reason it is included in the quartet. Here we shall deal with several more general contexts that arise from the overall picture presented in the first two parts.

We shall examine here the relation between the first two roots, and the general picture of the halakhic categories that emerges from them, with its underlying branching mechanism that defines categories standing between the familiar categories on both sides—de’oraita and derabbanan (of rabbinic authority), legislation and interpretation, Written Torah and Oral Torah. After that, we shall turn to the philosophical basis of the branching mechanism, and connect it with similar phenomena at the logical level that appear in the philosophy of science and in epistemology. These phenomena were discussed in the previous three books of the quartet, and were linked to the issue of the synthetic a priori that accompanies us throughout. Here we shall see its implications in the field of halakhic and general legal theory.

Contents of Part Three

Gate Five: A General View of the Two Roots: Halakhic and Meta-Halakhic Meanings

• Chapter 1: Summary and a General View of the Two Roots
• Chapter 2: Halakhic Implications: The Parameters of “Do Not Deviate”
• Chapter 3: The Logical and Halakhic Relations Between the Two Mechanisms

Gate Six: Legal Implications: Issues in the Relationship Between Law and Halakha

• Chapter 1: Halakha and Law: Maimonides’ Two Roots in Light of Civil Legal Systems
• Chapter 2: Basic Concepts and Foundational Questions in the Philosophy of Law: A Selective Introduction to Legal Theory and Legal Interpretation
• Chapter 3: Interpretation, Convention, Stability, and Discretion: Between Halakha and Law
• Chapter 4: The Analytic and the Synthetic in Law, in Halakha, and in Philosophy

Gate Seven: The Philosophical Foundation: The Synthetic A Priori

• Chapter 1: Studies in Dialectics
• Chapter 2: The Analytic and the Synthetic
• Chapter 3: Justifying the Dialectical Synthesis: Two Types of the Synthetic A Priori

Footnotes