Lesson 10: Category 4 — The Eighth Root

From the book Roots Outstretched by Rabbi Michael Avraham. Translated from Hebrew using gpt-5.4 (reasoning_effort=high, batch API).

---

The Eighth Root: One Should Not Count the Negation of Obligation Together with Prohibitions

Prescriptivism and Descriptivism

Our present essay deals with Maimonides’ eighth root. This root belongs to the fourth of the five categories we defined in the introduction: rules that concern the character of the mitzvah (commandment) as a command. As we saw in our previous essay, this entire category deals with defining the term “commandment,” and distinguishing it from other kinds of sentences, especially by comparing indicative verses with imperative verses. In the terminology of the essays on the fifth and sixth roots, this category deals with defining the “normative sphere” (to which all command-verses belong), and distinguishing it from other spheres, such as the factual one.

In this root, Maimonides establishes the principle that one must not count negations of obligation together with prohibitions. We will define the term “negation of obligation” below, though we have already mentioned it in the discussions of the fifth and sixth roots. At bottom, it is a kind of indicative verse. Therefore, the main discussion of the distinction between these different sentence-types will take place in the present essay.

At first glance, this root seems technical rather than substantive. It therefore appears to leave little room for disagreement, since it seems unlikely that anyone would claim that the enumeration of the commandments should include statements that are not obligations at all, but merely informational disclosures. Yet we will see that substantive disputes do in fact arise here, and that they embody radically different conceptions of the essence of the Torah’s commandments, especially of their relation to reality. Beyond that, we will see that the positions expressed in this root have philosophical implications, chiefly for the meaning of ethical and evaluative statements in general. In the language of analytic philosophy, one may say that the main issue here is the relation between prescriptivism and descriptivism, as will become clear in the body of the essay.

A. The Course of Maimonides’ Discussion

Introduction

In this chapter we will trace the course of Maimonides’ argument and examine several halakhic examples that illustrate his position. Throughout, one must keep in mind that Maimonides assumes here the existence of some abstract “element” that is present in positive commandments and prohibitions, in contrast to indicative verses. He never defines that element directly, although he offers several linguistic indications of its existence. The rest of this essay will be devoted to clarifying and sharpening this shared element, a kind of “color” that characterizes all halakhic normative verses of every type.

The Common Element of Prohibitions and Positive Commandments: The Constitutive Element of the Normative Sphere

Maimonides opens his discussion of this root by defining the term “warning,” or “decree”:

The eighth root is that it is not proper to count negations of obligation together with prohibitions. Know that a prohibition is one of the two parts of command. For one commands the person addressed either to do something or not to do it. Thus one may command him to eat and say, “Eat,” or command him to refrain from eating and say, “Do not eat.” In Arabic there is no single term that includes both of these meanings together. The logicians have already mentioned this and said, in these words: command and prohibition have no single name in Arabic that encompasses them both, and so we were forced to call both of them by the name of one of them, namely command. It has therefore become clear to you that prohibition belongs to the category of command. The well-known Arabic word used for prohibition is the word “do not.” And this same matter certainly exists in every language. That is, one commands the person addressed either to do something or not to do something. Thus it is clear that positive commandments and negative commandments are both full commands: things we were commanded to do, and things we were warned not to do. The things we were commanded to do are called positive commandments, and the things from which we were warned are called negative commandments. The term that includes them both in Hebrew is “decree.” Thus the Sages called every commandment, whether positive or negative, “the King’s decree.”

This passage attempts to characterize something shared by positive commandments and prohibitions. To us this may seem obvious, but it is not obvious at all. A positive commandment is a sentence that instructs us to do something, whereas a prohibition is a sentence that forbids something. What exactly is common to these two kinds of sentences? As we saw in the essay on the sixth root, they are not even logical negations of one another. Their relation, then, is not a logical one. So what do they nevertheless share? To see that, we must contrast them with another type of sentence, such as factual statements, that is, indicative statements.

This common element is tied to the fact that these are two kinds of norms, as distinct from facts. A sentence like “Moses is wearing brown shoes” states a fact. A sentence like “This picture is beautiful” states a judgment, in this case an aesthetic one. A sentence like “Helping others is an important value” states either a judgment or guidance, that is, a norm, on the ethical plane. Sentences of positive and negative commandments belong to what we have often called “the normative sphere,” meaning that they establish halakhic norms. This is the common element shared by both: both establish halakhic norms.

Yet that by itself does not help us understand their nature. Maimonides is saying here that sentences not of this kind are not included in the enumeration of the commandments. But if we define this kind of sentence simply as those that establish halakhic norms, then at most we have produced a circular claim: sentences that do not establish halakhic norms are not included in halakha (Jewish law). But what is the character of such sentences? What indicates that they are not normative sentences, or at least not halakhic norms? At this point we would expect independent criteria to help us determine that. This is the “element” shared by both positive commandments and prohibitions that we are looking for, and Maimonides devotes this introduction to it.

Maimonides, however, only points here to the existence of a common element; that is, to the fact that the normative category is linguistically distinct. He goes on to note that the common element shared by positive commandments and prohibitions, and absent from other kinds of sentences, is universal:

This same matter certainly exists in every language. That is, one commands the person addressed either to do something or not to do something.

In other words, this is not a feature of some particular language or culture. It exists in all languages, and therefore presumably contains something objective. Maimonides devotes a good number of lines at the opening of this root to proving the existence of this element and characterizing it. As we shall explain below, he is hinting here at a central question in the analytic philosophy of morality. This normative element is the principal subject of our present essay.

The Linguistic Aspect

In the passage above, Maimonides explains that the two types of commandments, positive and negative, fall in Arabic under the heading “command.” This is one type of sentence, normative or imperative, which divides into two species. He notes that Arabic lacks a common term for both kinds of sentences, and therefore both are called “command,” even though in its original meaning that term refers to positive commandments and not prohibitions. In Hebrew, by contrast, there is a common term for both, namely “decree.” In summary, one could say that there are two kinds of decrees: prohibitions and commands. We should note that the term “commandments” also originally means “commands,” but in Hebrew too it functions as a general term, interchangeable with “decrees.”

Maimonides explains here that a prohibition usually appears with a negative word such as “do not.” He is alluding to the rabbinic definition in Babylonian Talmud, Eruvin 96a, and parallels:

Rabbi Avin said in the name of Rabbi Ilai: Wherever it says “beware,” “lest,” or “do not,” it indicates only a prohibition.

The early commentators already wrote that the word “not” should also be added to this list. Usually they explain that Rabbi Avin did not include it simply because it is obvious. As we shall see below, however, Maimonides’ language suggests another explanation: the word “not” does not indicate only prohibitions, for at times it appears in the sense of a negation of obligation, rather than command.

Defining the Term “Negation of Obligation”

Maimonides now moves to define the basic concept at the center of this root, the “negation of obligation”:

Negation of obligation, however, is something else. It is when one negates a predicate of a subject, and there is nothing at all in it of the nature of command. For example, when you say: “So-and-so did not eat last night,” “So-and-so did not drink wine,” “Reuven is not Shimon’s father,” and the like. All of these are negations of obligation; there is not the slightest scent of command in them. The word usually used for negation in Arabic is “ma,” though negation is also expressed by “la” and “laysa.” In Hebrew, however, most negation is by the very word “lo,” the same word by which one prohibits. Hebrew also negates by “ein,” and by forms joined to pronouns, such as “he is not,” “they are not,” “you are not,” and the like. Negation in Hebrew by “lo” appears in such verses as: “No prophet again arose in Israel like Moses,” “God is not a man, that He should lie,” “Affliction shall not rise twice,” “No man stood with him,” “No one rose or stirred because of him,” and many others. Negation by “ein” appears in such verses as: “But there was no human,” and “The dead know nothing,” and many others.

Maimonides notes that in certain contexts the word “not” serves simply to negate a predicate of some subject, and in such cases it has nothing to do with the world of commands, that is, decrees or norms, at all.¹ This is what Maimonides calls a “negation of obligation.” In our terms, we would say that a sentence negating a predicate of a subject does not belong to the normative sphere. It is an indicative verse, not an imperative one, and therefore its content is not included in the enumeration of the commandments.

Maimonides now looks for independent indicators that characterize such sentences, still on the linguistic plane. To that end, he compares prohibition with command, and contrasts both with the negation of obligation:

The difference between prohibition and negation has already become clear to you. A prohibition belongs to the category of command, and exists only in the verbal form of command. I mean that just as command is always future, so too prohibition; it is not possible in language for command to be in the past, nor prohibition either. Nor is it possible to place command within narrative, because narrative requires a predicate and a subject, whereas command is a complete utterance, as has been explained in books composed on this topic. Prohibition too does not enter into narrative. Negation is not so, for negation does enter into narrative, and it negates in the past, in the future, and in the present. All this is self-evident upon reflection.

Maimonides lists several respects in which prohibition and command, that is, all decrees, are similar. A command is expressed in future form—“Do such-and-such”—and so is a prohibition—“Do not do such-and-such”—and never in the past tense.² A command never appears as narrative, since a command sentence has no subject and predicate. It does not assert any proposition about any subject. And again, this is true both of positive commandments and of prohibitions. A negation of obligation, by contrast, may be stated in future, present, and past forms, and may include a subject and predicate, and so on. Thus a negation of obligation differs from both commands and prohibitions alike.

It should be noted, however, that ambiguity remains in the case of sentences phrased in the future. Such sentences can be understood either as commands or prohibitions, which are always formulated in future form, or as future indicative statements, that is, statements of future fact.

Why Not Count Negations of Obligation in the Enumeration of the Commandments?

Maimonides now explains why a negation of obligation cannot be included in the enumeration of the commandments. The reason is that it is not a decree, or command, whereas all commandments are decrees, whether positive or negative. Clearly, if one compares a negation of obligation to any kind of decree, it will resemble a prohibition rather than a positive command, since in both cases the word “not” appears at the beginning of the sentence. Hence these are sentences that resemble prohibitions. Maimonides therefore now explains why a negation of obligation differs from a prohibition:

Since this is so, it is in no way proper to count negative expressions that are negations among the negative commandments. This is a demonstrative matter that requires no proof beyond what we have already said in understanding the meanings of the words, by which one distinguishes prohibition from negation.

Maimonides is satisfied with the linguistic distinction between indicative and imperative verses, that is, decrees, and sees this as sufficient grounds for his ruling. This argument is a priori and needs no textual proof, and therefore, in Maimonides’ view, no one can really dispute it.³

What is the meaning of this linguistic distinction? And on the other hand, what is novel about it? If these verses really do not establish norms, what is new in saying that they are not included in the enumeration of the commandments? Is there any reason one might have thought to include factual verses that merely state facts? There is a strong sense that this too is connected to that same shared “element” present in both kinds of commanding sentences, positive and negative. The linguistic differences are only indications of that element, yet nowhere does Maimonides define it substantively, that is, apart from such linguistic indicators. He takes it for granted.

Behag’s View: The Example of “She Shall Not Go Free as the Male Slaves Do”

Despite that certainty, Maimonides immediately adds that someone did dispute this determination:

This escaped others besides us, to the point that he counted “She shall not go free as the male slaves do,” not realizing that this is a negation, not a prohibition. I will explain this.

Maimonides says that Behag erred on this point and counted even negations of obligation as commandments. From this point on, the discussion turns to various examples, and here we will examine one of them in order to clarify the dispute. Maimonides cites as an example of Behag’s error the verse, “She shall not go free as the male slaves do.” Behag counts it as a commandment, that is, a prohibition, while Maimonides argues that it is a negation of obligation and not a prohibition.

First, one should note that even Maimonides understands that Behag does not disagree with him at the level of principle. After all, Maimonides himself explains Behag’s inclusion of this verse by saying that he apparently did not see the verse as a negation of obligation, but as a commandment. In other words, Maimonides assumes that Behag too agrees that negations of obligation are not counted, and that his dispute with Maimonides is interpretive. In Behag’s view, this specific verse is not a negation of obligation but a prohibition, and should therefore be counted.

To understand the dispute, we should first note that this verse unquestionably deals with halakha. It is not like Maimonides’ earlier examples, such as “God is not a man, that He should lie.” This verse has direct halakhic significance, and in that simple sense it certainly belongs to the normative sphere. So why does Maimonides nevertheless think it is not a commandment? What does it mean to say that this is a negation of obligation rather than an ordinary legal prohibition? Is it really just an indicative verse rather than an imperative one?

Maimonides now explains:

God has already ruled regarding one who strikes his Canaanite slave, male or female, and in the blow destroys one of the extremities, that the slave goes free. We might have thought that if this is so for a Canaanite slave, then all the more so for a Hebrew maidservant, and that if one of her extremities were destroyed she too should go free. Scripture therefore denied her this rule by saying, “She shall not go free as the male slaves do,” as if to say: she is not obligated to go free through the loss of one of her limbs. This is the negation of a legal rule with respect to her, not a prohibition. So too the transmitters of tradition explained it, saying in the Mekhilta: “‘She shall not go free as the male slaves do’—she does not go free through injury to the extremities in the way the Canaanite slaves do.” Thus it has already become clear to you that this is the negation of a single legal rule denied with respect to her, not a warning that we were warned against something.

A basic precondition of every negation of obligation is the existence of some prior positive rule that the Torah is now negating. Here “obligation” does not necessarily mean “duty”; it means the positive application of a legal rule. Moreover, there must be some reason that would have led us to apply that positive rule to the case before us, and the negation of obligation comes to say that in this case the rule does not apply.

Let us illustrate this in our case. The Torah contains a basic law that if a master strikes his Canaanite slave, male or female, in a way that injures one of the extremities, the slave goes free. That is the underlying positive rule. A priori, one might have thought that this rule should apply to a Hebrew maidservant as well; indeed, there might even be an a fortiori inference in that direction. More generally, there is very good reason to think that this law should apply even to another case not explicitly mentioned in the Torah, such as a Hebrew maidservant. Here comes the negation of obligation and removes that positive rule from the situation, or the person, now under discussion—in our example, the Hebrew maidservant. The verse thus teaches that a Hebrew maidservant is not subject to emancipation through injury to the extremities. It negates the rule in relation to this legal subject. Maimonides also cites the Mekhilta as support for his reading, since it understands the verse as a limitation upon the law of emancipation through injury.

Two Difficulties

Even so, two main difficulties remain:

1. What is the alternative against which Maimonides is arguing? How can this verse be interpreted differently, that is, in a way that is not a negation of obligation?
2. Why does such a reading strip the verse of the status of a commandment, or even of halakha? At first glance, this verse certainly belongs to the normative sphere.

These may seem to be two sides of the same coin: once we understand the alternative, we will see that it reads the verse as a prohibition, whereas the reading as a negation of obligation is different, and accordingly the verse is not a prohibition but something else. As we shall see in the next example, however, these are actually two different questions.

Another Example: “He Shall Not Be Like Korah and His Company”

To sharpen matters further, let us consider another example, one of many brought by Maimonides later on:

Likewise, they said that “He shall not be like Korah and his company” is a negation. The Sages explained that it is a negation and interpreted it to mean that the Blessed One informed us that anyone who disputes the priesthood and stirs rebellion concerning it will not suffer what befell Korah and his company—swallowing up and burning—but rather his punishment will be as the Lord spoke through Moses, namely leprosy, as it says, “Put now your hand into your bosom.” They brought proof from what happened to Uzziah king of Judah. And although we find another statement of theirs in the Babylonian Talmud, Sanhedrin 110a—“Whoever persists in dispute transgresses a negative commandment, as it says, ‘He shall not be like Korah and his company’”—that is only by way of scriptural support, not because that is the plain meaning of the verse. The prohibition against this matter is included under another negative commandment, as I will explain in its place.

This example is clearer and sharper. First of all, here the alternative is obvious. The prohibition that appears in the Talmudic discussion in Sanhedrin—persisting in dispute—is the normative reading of the verse, according to which we have an ordinary prohibition. Maimonides’ proposed reading is that this is a promise: even if in the future others dispute the priesthood, they will not be punished as Korah and his company were. Here it is obvious that this is not any sort of command. Therefore, if that reading is correct, there is clearly no place to include this verse in the enumeration of the commandments. It is not a normative verse at all; it lacks the “color” characteristic of normative verses.

Returning to “She Shall Not Go Free as the Male Slaves Do”

This example may also clarify the alternative reading we were seeking in the previous example. From the comparison it seems that Maimonides there is arguing against the possibility of reading the verse as a prohibition against emancipating a Hebrew maidservant through injury to the extremities. Against that, he claims that the verse should be read as a negation, namely that a Hebrew maidservant is not subject to the law of emancipation through injury to the extremities. That is not a prohibition but a negation. So the first difficulty is resolved: we have identified the normative reading available for that verse.

But the reading as a negation of obligation, which Maimonides proposes, is very different in these two examples. As we saw, in the case of “He shall not be like Korah,” this reading does not belong to the normative plane at all. But in the previous example, even if we do not read it as a prohibition but as a negation of the law of emancipation through injury to the extremities, the verse still plainly belongs to the normative plane. It teaches us that this law does not apply to a Hebrew maidservant. Why is that not a halakhic verse? It therefore seems that the second difficulty remains unresolved, and it is now also clear that these are not just two sides of one problem.

Put differently: the “element,” that is, the “color,” that we are looking for cannot simply be identified with the term “halakha.” There are verses that have halakhic significance and yet do not possess the color of a positive or negative commandment, and as such they are not included in the enumeration of the commandments. There is something about “commandments” that is more specific than “laws.” Put differently still, such a verse—“She shall not go free as the male slaves do”—would appear in the Shulhan Arukh, but not in the enumeration of the commandments. This suggests that the common element we are seeking has to do not with the content of the verses—whether they belong to halakha—but with their character as motivating us to action, whether by commanding or forbidding. We will expand on this below.

Behag’s View

If we return now to Behag, who disagrees with Maimonides and counts “She shall not go free as the male slaves do” as a commandment, it appears that he can disagree with Maimonides on two different planes. He may disagree on the interpretive plane, claiming that the verse prohibits releasing a Hebrew maidservant through injury to the extremities. But he can also agree with Maimonides on the interpretive plane, and still claim that even according to Maimonides’ own reading, the verse is a commandment that should be counted. This is unlike the case of “He shall not be like Korah,” where the disagreement is plainly on the first plane. According to this second possibility, Behag would hold that any verse with halakhic significance can be included in the enumeration of the commandments, even if it does not command us to do anything.

The Root of the Error

Later in his remarks, Maimonides identifies the root of the error that leads Behag astray: the use of the negative word “not” in two different senses:

Nothing distinguishes negation from prohibition except the meaning of the statement, not the word “not” itself. For in Hebrew the word used for negation and the word used for prohibition are one and the same, namely “not.” Therefore the discerning person must understand the meaning of the statement, and he will quickly grasp which “not” is a negation and which “not” is a prohibition, in light of what we have already explained.

The Sages too hinted at this point. For we find them in disagreement over one of the negative expressions in the Torah as to whether it is a negation or a prohibition. This is with regard to the bird sin-offering: “He shall pinch off its head at the nape, but he shall not separate it.” The Tanna of our Mishnah, who speaks in Mishnaic idiom, held that this is a prohibition, and therefore said that if he did separate it, he rendered the offering invalid. According to this, this negative expression is a negative commandment, because when he separated the head he invalidated the offering, just as if he had offered leaven or honey. But Rabbi Elazar son of Rabbi Shimon held that this negative expression is a negation, not a prohibition, and that “he shall not separate” means only that he need not separate the head, but may cut it to whatever extent. Therefore, if he did separate it, according to his view the offering remains valid.

As we saw above, prohibitive verses and negative indicative verses resemble one another because in both the word “not” appears. On the other hand, we also saw differences between them, and Maimonides argues that those differences show that the word “not” has different meanings in these two contexts. That is the essence of the distinction. Failure to distinguish between these two meanings is what lies at the root of the errors in this area.

Maimonides explains that the difficulty of distinguishing the two meanings is reflected in tannaitic disagreements regarding the bird sin-offering. One view reads “he shall not separate” as a prohibition, while another reads it as a negation of obligation. Here too one can see the two possible ways of reading the verse—normatively, or as a negation—and in this respect the case resembles “She shall not go free as the male slaves do,” because in both possibilities the verse has halakhic implications.

What are the two possibilities? One possibility is that there is a prohibition against separating the head from the body in the melikah procedure, and the consequence is that if he did separate it, he invalidated the offering. The other possibility, the negation, is that there is no obligation to separate it, though of course there is no problem in separating it, and therefore if he did separate it the offering is not invalidated.

In practice, we rule that this is a prohibition and not a negation, but the tannaitic disagreement clarifies the ambiguity involved in such verses. Incidentally, with regard to the bird burnt-offering there is a parallel verse, and there it is interpreted as a negation of obligation. That verse comes to negate the application of the law of separation from the bird sin-offering, or from the animal burnt-offering, to the bird burnt-offering. There, accordingly, there is indeed no prohibition against separating it, and doing so does not invalidate the offering.

How, then, does one distinguish these two meanings? Maimonides answers: by the biblical context. Yet our main question still stands: why is the reading as a negation of obligation different from the normative reading? Is not every halakha a norm? Put differently: why, even according to this mode of interpretation, is this not a commandment that ought to enter the count?

The Conclusion of the Discussion

Maimonides concludes the discussion by saying:

We have clarified this matter to the utmost, to the point that no doubt remains—not even for the dullest mind among men.

Immediately afterward, he comments on the negative formulations of the Torah:

And after speaking with this intention, know that the words through which prohibition appears in the Torah are four words: “beware,” “lest,” “do not,” and “not.” Whatever the Torah prohibits through one of these four is called a negative commandment. The Sages explicitly said: wherever it says “beware,” “lest,” or “do not,” it indicates only a prohibition.

His intention seems to be that not everywhere a negative word appears do we have a prohibition, but every prohibition must contain some negative expression. In other words, the rabbinic rule states a necessary condition, not a sufficient one.

Maimonides concludes with a qualification to the sweeping principle of this root, by bringing an exceptional case in which the word “not” appears in the sense of negation of obligation, and yet we do learn from it that a prohibition exists:

One matter remains for us to clarify in order to complete the aim of this gate. When the Torah recounts and decrees that we are to clear ourselves by denying that we performed a certain act, then that act is counted among the negative commandments, even though the negative expression used there is a negation and not a prohibition. Since we were commanded to deny it with regard to ourselves and say, “I did not do such-and-such, and I did not do such-and-such,” we necessarily know that the act in question is prohibited. Thus Scripture commanded us to say: “I have not eaten of it in my mourning, I have not removed any of it while impure, and I have not given any of it for the dead.” This indicates that each of these acts is prohibited. The explanation of this will come in its proper place, in our discussion of these commandments.

His claim is that from the context it is clear that if a person is required to declare that he did not sin by doing something, then clearly there is a prohibition against doing it. That is, the prohibition here is not inferred from the mere presence of the word “not,” since in this case it is a negation of obligation. The existence of a prohibition is learned here indirectly, from the overall meaning of the verse.

We should note that Nahmanides, at the end of his comments in his glosses, says that Behag too agrees with Maimonides’ qualification concerning negations that we are commanded to state about ourselves, such as the declaration regarding second tithe, and that in his opinion too they are counted as prohibitions. Nahmanides himself, however, argues that both of them are wrong. In his view, this is a prohibition inferred from a positive commandment to eat the second tithe in purity and joy. This has practical consequences regarding lashes for someone who violates it. See there for the proofs Nahmanides adduces; we cannot elaborate further here.

B. Nahmanides’ Remarks and His Dispute with Maimonides

Nahmanides’ Position: The Basic Thesis

Nahmanides opens his comments by agreeing in principle with Maimonides on this root:

The eighth principle: the Rabbi, of blessed memory, said that negations should not be counted together with prohibitions. The Rabbi explained this principle and the matter of negations: when we negate a predicate of a subject and there is nothing in it of the nature of command at all…

All these things are clear and hidden from no intelligent student. But we must add one further stipulation: according to our Rabbis, there is nothing in the Torah said in the language of prohibition or in the language of positive command that is merely the negation of a command—that is, a permission—or the negation of one of the laws, unless Scripture had need of that negation.

Nahmanides adds an important clarification. We should never interpret the word “not” as a negation of obligation unless there is some initial assumption that the verse comes to negate. In other words, a negation always appears against the background of some alternative possibility; otherwise the verse would contain no novelty at all, and the Talmud would ask in astonishment: “Isn’t that obvious?”

Nahmanides further claims that this is an interpretive tool. That is, when we do not find any initial assumption that needs to be negated, we should interpret the verse as a command. Only when such an assumption exists may—or, from Nahmanides’ wording, must—we read it as a negation of obligation.

Nahmanides argues that all the medieval authorities, including Behag, agree that negations of obligation are not to be counted. He therefore explains all the examples Maimonides brings to show that Behag counted negations as prohibitions as results of interpretive disagreements. According to Nahmanides, Behag counts those commandments because he does not read them as negations but as prohibitions.

We may add here that Rav Yerucham Fishel Perla, in his introduction to Saadia Gaon’s Sefer HaMitzvot, in his comments on the eighth root, notes that Saadia Gaon too agrees with the principle of this root. He also brings examples in support of Nahmanides’ addition, namely that there must be an initial assumption that the rule would apply in order for a verse to be interpreted as a negation of obligation. He claims that this is explicit in Jerusalem Talmud, Rosh Hashanah 1:1, which says:

Anything that comes to permit—one does not transgress it.

Two Types of Initial Assumption

As we have seen, Nahmanides explains that the existence of an initial assumption is necessary in order to interpret a verse as a negation of obligation. This initial assumption is an a priori consideration that would lead us to apply the rule in this case as well. That consideration may indeed result from logic, moral principles, and the like, but it may also result from an interpretive consideration, that is, from the reading of one verse or another. As an example, Nahmanides writes regarding “She shall not go free as the male slaves do”:

I am close to deciding the law here and ruling that in truth this is indeed a prohibition, but one may nevertheless discuss it as a negation, as the Rabbi said—not for the reason he gave, but for another reason. The section on injury to the extremities was stated generally: “If a man strikes the eye of his male slave or the eye of his female slave…” One might therefore have thought it includes all slaves, or that Scripture speaks only of Hebrew slaves, who are more worthy of compassion. Therefore it was necessary to say here, “She shall not go free,” that Hebrew slaves are not included in this rule, but only Canaanites.

Here the initial assumption is connected not to logic or morality but to interpretation that appears more plausible. The verse states the principle concerning slaves in general, and we would therefore think it applies to all kinds of slaves. That is an interpretive assumption. Or perhaps it might even apply especially to Hebrew slaves, who are more deserving of compassion; that is a logical-moral assumption. As noted above, Maimonides himself brought an a fortiori inference from a Canaanite slave to a Hebrew maidservant.

There is room to discuss the nature of the a fortiori reasoning Maimonides mentions. Behind it lies logic similar to the compassion argument, and yet an a fortiori inference is also an interpretive tool. This is a good illustration of how far from sharp and clear the distinction is between textual considerations and conceptual ones. These matters have implications for questions such as the relation between the plain sense of Scripture and homiletic interpretation, but we cannot elaborate on that here.⁴

Maimonides and Nahmanides on the Need for an Initial Assumption

As we noted, from the examples Maimonides himself brings of reading the word “not” as a negation, it emerges that he too always points to an initial assumption from which the Torah comes to exclude the case at hand. This suggests that Maimonides too agrees with Nahmanides on this point. That is also clear in light of Nahmanides’ own argument that without such an assumption, the negating verse would be redundant.

Even so, Nahmanides’ emphasis suggests that he himself understood that there is some disagreement between him and Maimonides on this issue. Apparently for that reason, the author of Pekudei Yesharim, in his commentary on this root, explains that according to Maimonides no initial assumption is required in order to interpret a verse as a negation of obligation. Such an interpretation, he claims, emerges from the context of the surrounding verses and does not require the prior existence of some assumption that must be negated. According to Nahmanides, by contrast, only when there is an initial assumption to apply the rule does the verse become a negation of obligation.

But it is difficult to read Maimonides that way. First, as we saw, he too always makes sure that the negation removes some initial assumption. Moreover, this is nearly explicit in his discussion of Positive Commandment 129, where Maimonides addresses our very root:

The one hundred and twenty-ninth commandment is that the Levites were commanded to separate a tithe from the tithe that they receive from Israel and give it to the priests, as He, exalted be He, said: “Speak to the Levites and say to them: when you take from the children of Israel the tithe that I have given you from them as your inheritance, you shall raise up from it the Lord’s gift, a tithe from the tithe.” Scripture explained to us that this tithe, called terumat maaser, is to be given to the priest, as it says: “You shall give from it the gift of the Lord to Aaron the priest.” Scripture further explained that this tithe must be taken from its best and finest part, as it says: “From all its choicest part, its sacred portion from it.” And it further hinted that they sin if they do not separate it from the best, as it says: “You shall not bear sin because of it when you raise up its best part from it.” This negative expression is a negation of obligation, for it says: you bear no sin when you separate it from the best, and this indicates that if you separate it from the inferior part, you do sin. Its sense is like that of a prohibition inferred from a positive commandment, which is not counted among the negative commandments. That is, after commanding that it be separated from the best, Scripture thereby indicated that it must not be separated from the inferior. And the language of the Sifrei is: From where do you know that if you separated it not from the best, you bear guilt? Scripture says: “You shall not bear sin because of it when you raise up its best part from it.” The laws of this commandment have already been explained in Tractate Terumot, in Tractate Maaserot, and in places in Demai as well.

We thus see here a verse read as a negation of obligation and yet still counted in the enumeration of the commandments, as a positive commandment. How does that fit with the principle of this root?

It appears from Maimonides’ words that in this case there is no initial assumption from which the verse excludes us. Before this verse, we did not know that there was sin in separating from inferior produce, and even if we had known it, we certainly would not have thought there would also be sin in separating from superior produce. So clearly no verse was needed to teach that there is no sin in doing so. Here the situation is the opposite: the verse that says there is no sin when one separates superior produce is itself what teaches us that there is sin in separating inferior produce. That is its main novelty. Since that is what the verse itself introduces, we do not have here a negation of obligation but rather a prohibition inferred from a positive commandment. The conclusion is that even according to Maimonides an initial assumption is required before we read a verse as a negation of obligation. Without such an assumption, it will be interpreted as a prohibition inferred from a positive commandment.

A Consequence: “And He Shall Not Profane”

In the course of his gloss to the fifth root, Nahmanides writes as follows:

I saw that the Rabbi, in Commandment 165, said something else. He explained that the phrase “and he shall not profane” is a negation and not a prohibition; it teaches that his service is not profaned even though he is an onen. The implication is that another who did not go out—namely, an ordinary priest who performed the service while an onen—his service is invalid. And he cited what they said in Babylonian Talmud, Sanhedrin 84a: “From where do we know this for an onen? For it is written: ‘And from the sanctuary he shall not go out and he shall not profane’; the implication is that another, who did not go out, does profane.”

Maimonides sees this verse as a negation of obligation. In other words, a High Priest who becomes an onen—that is, a mourner before burial—and continues the service has not profaned his service. He brought proof from the Talmudic discussion in Sanhedrin, which derives from the verse that in the opposite case, an ordinary priest who is an onen and does not go out has indeed profaned the service. That is, the first half is merely a negation of obligation, and the second is a kind of prohibition inferred from a positive commandment, as we just saw in the discussion of Positive Commandment 129, where there is a clear connection between negations of obligation and prohibitions inferred from positive commandments.

As we already noted in our essay on the fifth root, Maimonides’ words here show a close affinity between negations of obligation, our present root, and reason-verses, the subject of the fifth root, for he writes:

It has thus become clear to you that their reading of “and he shall not profane” treats it as a negation and not a prohibition, meaning that his service does not become profaned even though he is an onen. And according to the plain meaning of the verse, “and he shall not profane” is the reason for the previous warning: he shall not go out, so that he not profane. According to both of these understandings, this prohibition should not be counted separately, as is clear to one who understood the earlier roots in this treatise.

That is, according to the plain sense of Scripture, “and he shall not profane” is a reason for what precedes it, and therefore it is not counted by virtue of the fifth root. According to the rabbinic interpretation, from which we learn that an ordinary priest who serves while an onen does profane the service, then with respect to the High Priest this is a negation of obligation—that in his case there is no profanation—and therefore it is not counted by virtue of the eighth root.

Here too we have an example of a negation of obligation that has no parallel independent source for the corresponding prohibition, and therefore there was room to see it as a prohibition inferred from a positive commandment—namely, as a warning to the ordinary priest. If there had been some independent source teaching that the service of an ordinary priest who serves while an onen is profaned, then this verse would have been read only as the negation of obligation with regard to the High Priest, namely that he does not profane the service.

Nahmanides then notes that the Tannaim disagreed on this in the Talmudic discussion in Zevachim:

They already disagreed on this very rationale in the second chapter of Zevachim, where they said: From where do we know that an onen profanes the service? Because it is written, “And from the sanctuary he shall not go out and he shall not profane”—the implication is that another, who did not go out, does profane. Rabbi Elai said: from here—“They offered, but I offered”—the implication is that had they offered, it would rightly have been burned.

That is, there is a tannaitic dispute whether this is a negation of obligation or not. The Gemara there asks why Rabbi Elai does not interpret it as a negation of obligation:

And why did Rabbi Elai not say: “And from the sanctuary he shall not go out”? He would answer: Had Scripture written “the other, if he did not go out, profaned,” then indeed. That is to say, had it been a negation, meaning that the High Priest is given permission and need not go out of the sanctuary, and may serve while an onen without profaning the service, then it would have been proper to infer from it that another priest, if he did not go out, does profane. But now that it is a prohibition to the High Priest not to profane, an ordinary priest is not implied here as one who profanes. And the entire passage there fits this view, namely Rabbi Elai’s. According to his view, then, it is not a negation but a prohibition.

Nahmanides explains the Gemara’s question as follows: had the verse been a negation of obligation, meaning that the High Priest is permitted to serve while an onen and his service remains valid, then one could indeed infer that another priest who did not go out while an onen does profane the service. But if the verse is read as a prohibition addressed to the High Priest, then there is no basis for inferring anything about the ordinary priest.

According to Maimonides, however, this is not difficult at all. As we already noted, if there were some other source telling us that an ordinary priest who serves while an onen profanes the service, then this verse would indeed be read only as a negation of obligation. If so, Rabbi Elai is simply saying that there is such another source—“They offered, but I offered”—and therefore this verse is certainly only a negation of obligation. Maimonides thus emerges as correct according to both Tannaim. Rabbi Elai too reads this verse as a negation of obligation, and even more decisively than the first Tanna.

Difficulties in Nahmanides’ Approach

We saw that according to Nahmanides an initial assumption is required in order to read a verse as a negation of obligation. According to Maimonides, some wished to argue that such a reading exists even where there is no initial assumption to negate. But if so, it is indeed unclear why any verse is needed to negate the rule, as we already noted. This strongly suggests that Maimonides too requires such an initial assumption.

Yet this now gives rise to an opposite problem, precisely because such an initial assumption does exist. We should preface by saying that this discussion is relevant only where the wording of the verse itself can bear both interpretations, namely negation of obligation and prohibition. Only then does the question arise which of the two interpretations is correct. But if both possibilities do exist, and if there is in fact an a priori reason to apply the rule to the present case, then the more plausible reading would seem to be precisely not the negation of obligation. If our a priori reasoning says that the rule should apply here too, why should we prefer the interpretation that moves away from that reasoning? Why not choose the option closer to our a priori understanding?

It would therefore seem, on the face of it, that specifically when there is no initial reason to apply the rule, it would be more reasonable to interpret the verse as a negation of obligation rather than as a prohibition, because that would fit our prior reasoning. In such a case, the negation of obligation is sensible and accords with our a priori judgment. By contrast, when there is an initial reason to apply the rule, a priori logic would seem to tell us not to interpret the verse as a negation of obligation, since that runs against our reasoning.

Returning to the Distinction Between Two Kinds of Examples

It seems that here we must distinguish between the two kinds of negation of obligation that we presented above. For the first type, we gave the example of “He shall not be like Korah and his company.” This example can be interpreted in two ways:

1. Normatively—as a prohibition against dispute.
2. As a negation of obligation—as a promise that future rebels will not receive Korah’s punishment.

In the background stands the a priori reasoning that any rebel who behaves like Korah would presumably also be punished like Korah; there is no reason to assume Korah was a special case. According to Nahmanides’ line of thought, if there is an a priori assumption from which the verse comes to exclude us, then here we should indeed interpret it as a negation of obligation. But this is exactly the difficulty: since logic says that anyone rebelling like Korah should also be punished like him, why suppose that the verse is promising the opposite? Would it not be more reasonable to interpret it as a prohibition against dispute, and leave the prior reasoning in place?

And in fact, regarding this example, Nahmanides himself in his glosses here explains that it is a prohibition and not a negation. According to his proposal, the verse prohibits disputing the priesthood as Korah did. The broader prohibition against dispute is an extension made by the Sages beyond the plain meaning of the verse, which deals only with challenging the priesthood.

By contrast, as we saw above, in the example of “She shall not go free as the male slaves do,” both possibilities are halakhic and lie on the same axis:

1. The normative possibility—there is a prohibition against emancipating a Hebrew maidservant through injury to the extremities.
2. The possibility of negation of obligation—there is no duty to emancipate her through injury to the extremities, which is a neutral state: neither prohibition nor obligation.

Now, as we saw, even in this case there is an a priori assumption that the rule of emancipation through injury to the extremities should apply to a Hebrew maidservant as well; indeed, Maimonides himself mentions an a fortiori inference from the Canaanite slave. If so, as we argued, there is no reason to prefer the interpretation of negation of obligation, according to which the emancipation rule does not apply here, since our a priori reasoning suggests that it should apply. On the other hand, the other possibility is to read the verse as a prohibition, that is, that it is forbidden to release her through injury to the extremities. But that option is even farther from the a priori reasoning, since the reasoning suggested not merely that emancipation is permitted, but that it is required.

Therefore, at least in cases of this type, where both options lie on the same scale of obligation, permission, and prohibition, Nahmanides is indeed right—and, as we said, it seems Maimonides agrees—that when there is an initial assumption to apply the rule, we interpret the verse as a negation of obligation rather than as a prohibitory warning. And when there is no initial assumption to apply the rule to our case, the verse presumably teaches not merely permission but an actual prohibition. Nahmanides’ remarks must therefore be limited to precisely those cases where both options have a halakhic cast and lie on the same axis. Cases parallel to “He shall not be like Korah” do not belong to that discussion, and with respect to them the consideration we raised above indeed applies: absent special textual pressures, the verse should be interpreted in the way that best fits the a priori reasoning.

Was This Really Nahmanides’ Intention?

At first glance, our explanation cannot fully account for Nahmanides’ position. As we have formulated it, the principle of choice between the two interpretations does not arise from the consideration of “Isn’t this obvious?” as Nahmanides himself explained, but from the exact opposite consideration: precisely because the a priori default points toward applying the rule, it is more reasonable to read the verse as a negation of obligation than to read it as a full prohibition. In other words, the choice of the negation reading, and the rejection of the prohibitory reading, is not because we want to move away from the a priori reasoning, as Nahmanides explained, but rather because we want to remain as close to it as possible. It is possible that at this point Maimonides does disagree with Nahmanides, despite sharing his basic assumption that the interpretation as negation of obligation depends on the existence of an a priori reason to apply the rule.

Perhaps one should distinguish between a situation where there is an a priori reason to apply the rule, and one where the a priori reasoning points in neither direction. Where, without the verse, we also would not have applied the rule, there is indeed no point in writing a verse that merely negates it. It would add nothing. But where there is reason to apply the rule, we saw that logic itself makes it plausible to interpret the verse as a negation of obligation, more plausible than reading it as a prohibition.

The natural conclusion is that Nahmanides’ intention was specifically the case in which there is no decisive reasoning one way or the other, and a priori there was room both to apply the rule and not to apply it. In such a case there is point in writing a negation of obligation, since without the verse we would not have known what to do. The negation is then not vulnerable to the objection that it is obvious. Yet we still must choose between reading the verse as a negation of obligation and reading it as a prohibition. At that stage, textual considerations of the sort Maimonides discusses, attention to the biblical context, or considerations like the one just mentioned, namely minimizing novelty and staying as close as possible to what seems likely, may enter. It is also possible that when the two readings are equally balanced, we choose the one involving the least novelty, namely the negation of obligation. In that case, the preference is not because negation fits the a priori reasoning better, but because it minimizes the verse’s innovation.

The Example of the Dispute Between Rabbi Akiva and Rabbi Ishmael in Sotah: Positive Indicative Verses

Up to this point we have seen that this root deals with negative indicative verses. A negation of obligation is an indicative verse, not an imperative one, and it reports a negative fact. Are there parallel cases of positive indicative verses that can be read in two ways, either normatively as commands or indicatively as statements? It now appears that there are.

Immediately after his introduction, Nahmanides brings as an example for his claim—that an initial assumption is required in order to interpret something as a negation of obligation—the series of disputes between Rabbi Akiva and Rabbi Ishmael in Babylonian Talmud, Sotah 3a:

Another example is their statement in Sotah: “‘And he shall be jealous of his wife’—permission, according to Rabbi Ishmael; Rabbi Akiva says obligation. ‘He shall become impure for her’—permission; Rabbi Akiva says obligation. ‘You may work them forever’—permission; Rabbi Akiva says obligation.” They asked there: Shall we say that Rabbi Ishmael and Rabbi Akiva disagree throughout the whole Torah, one saying obligation and the other saying permission? And they explained: they disagree in scriptural interpretation. What is Rabbi Ishmael’s reason? As it was taught: since the Torah said, “You shall not hate your brother in your heart,” one might have thought that this includes such a case; therefore Scripture says, “A spirit of jealousy comes over him, and he is jealous of his wife.” Rabbi Akiva says: another verse about jealousy is written. Likewise with “He shall become impure for her”—what is Rabbi Ishmael’s reason? Since it is written, “Speak to the priests, sons of Aaron: none shall become impure for a dead person among his people,” I might have said not at all; therefore it was necessary to write, “For her he shall become impure.” Rabbi Akiva says: this is derived from “except for his close relative.” And similarly with “You may work them forever”: since the Torah said, “You shall not let any soul remain alive,” it was necessary to say, “You may work them forever,” to permit one of the nations who had relations with a Canaanite slave woman. Rabbi Akiva says: this is derived from “from them you may acquire.”

Thus all agree regarding negations where Scripture needs that negation, meaning where it must permit us something and grant us leave to do it. But where there is no need to grant such permission at all, they determine that it is a positive commandment or a prohibition.

All these examples concern positive commandments. The Talmud’s reasoning points precisely to the fact that where there is an opposite initial assumption, we interpret the verses as permission, that is, as a negation of prohibition, as distinct from a negation of obligation. And where there is no such initial assumption, we interpret the verses as positive commandments.

It appears from the Talmud that the simple meaning of every verse is normative. That is why the question is directed against Rabbi Ishmael and not against Rabbi Akiva. Rabbi Ishmael, who says that these verses indicate permission, is the one who must explain his position. And what is his explanation? Since there is an a priori reason against the commandment, the principle of minimizing novelty leads us to interpret the verse as negating the prohibition rather than introducing a new positive commandment. For example, in the case of the suspected adulteress, the a priori assumption is that one may not hate any fellow Israelite, and therefore one may not hate even one’s wife who has gone astray. The verse then comes and says, “and he shall be jealous of his wife,” that is, it excludes this case from the prohibition of hate. We now have two possibilities: either it commands him to be jealous, or it merely cancels the prohibition against hating her, and therefore jealousy is not obligatory but merely permitted. Rabbi Ishmael prefers the second interpretation because it is less novel than the first, that is, less distant from the a priori assumption. This is exactly our earlier explanation of negations of obligation.

If there were no such initial assumption—if it were already clear a priori that the principle forbidding hatred of one’s fellow does not apply to a wife who has gone astray—then when we come to read the verse, we could not possibly read it as negating a prohibition, because there would have been no initial reason to think the act was prohibited. We would therefore be forced to conclude that the verse commands him to be jealous as an obligation and not merely as a permission.

It follows that according both to Rabbi Akiva and to Rabbi Ishmael, one prefers to interpret the verse on the normative plane, whether as a prohibition or as a positive commandment. But when there is an a priori reason in the opposite direction, one moderates the interpretation and keeps it as close as possible to the a priori reasoning. And where there is no such prior reason, the correct reading is obviously the normative one. The dispute between Rabbi Akiva and Rabbi Ishmael concerns only whether there is in fact such an a priori reason in those cases. But the general principle of interpretive priority is agreed upon by both of them.

Nahmanides later goes on to explain Behag’s view regarding “She shall not go free as the male slaves do” in light of this, but that need not concern us here.

Why Does Maimonides Deal Only with Negative Indicative Verses?

Why does Maimonides deal only with verses of negation, and innovate the concept of “negation of obligation,” while ignoring positive verses that could just as well produce a parallel concept of “negation of prohibition”? At first glance, one might say that in negative verses the word “not” appears, and because that word can bear two interpretations confusion arises. But from the Talmudic discussion in Sotah it seems that no less confusion is possible even in the context of positive indicative verses, and it is reasonable to assume that Maimonides would apply his principle in those reverse cases as well.

It may be that Maimonides simply did not encounter anyone who erred in the context of positive commandments, and therefore he did not address the removal of the error that confuses positive commandments with positive indicatives. But the matter requires further clarification.

The Dispute Between Maimonides and Nahmanides

As we have seen, Nahmanides agrees with Maimonides’ definition that any verse such that one who acts against it commits no transgression, and whose entire point is merely to grant permission to perform some act rather than to impose an obligation, is an indicative verse—namely, a negation and not a warning—and therefore should not be included in the enumeration of the commandments. Their dispute concerns only which verses have the character of indication and which have the character of command.

Nahmanides claims that Maimonides expands too far the number of verses interpreted as negations of obligation, saying:

This matter of negations requires great precision and should not be expanded as broadly as the Rabbi expands it.

Yet he offers no alternative definition, and it is not clear exactly what the dispute is. As we have seen up to this point, Nahmanides seems not to have a different definition of negation, but rather an interpretive ideology committed to reducing as much as possible the number of verses that we read as negations. For that reason, Nahmanides begins by explaining that the interpretation of a verse as negation should take place only when we have an initial assumption from which the Torah wishes to exclude us. Mere literal interpretation does not suffice, as he thinks Maimonides assumed; there must also be the background of a mistaken initial assumption.

Up to this point we have seen that Maimonides too most likely agrees with this requirement. Without such an assumption, the interpretation as negation of obligation is vulnerable to the objection that it is obvious. Even so, it seems no accident that Nahmanides makes very sparing use of this interpretive tool, and that all his disputes with Maimonides move in the same direction: Nahmanides claims a normative reading of the verses in question, while Maimonides explains that they are indicative verses.

Between Permission and Indication

There are two possible ways to formulate the principle expressed in this root, both of which have been implicit in our discussion so far. At first glance, it seems that these verses are not counted because they neither command nor forbid, but merely permit. They are verses of permission, and therefore have no place in the enumeration of the commandments. But what exactly are verses of permission? Are they indicative verses? One could see them as normative verses, and yet because of their content they are not included in the count.

As we have seen, negation verses are verses that negate the existence of a norm in a given context. It is possible that even a verse that negates the existence of a norm is itself a normative verse; that is, it too belongs to the normative sphere, except that its content is neutral. It does establish some halakha, but it neither commands nor forbids. Such a verse merely informs us that there is no binding norm here, and that itself is a legal determination. On the other hand, one might also understand that such a verse differs not only in content but in kind. It is not a normative verse at all, but an indicative verse. It informs us of facts and is not “colored” with the color of normative verses. We will define this more clearly below.

The question raised here is whether the difference between permission and prohibition lies on the same plane as the difference between prohibition and command. If it does, then all three are different kinds of normative verse. Some normative verses establish prohibitions, some establish permission, that is, the absence of a norm, and some establish obligations. But it is also possible that the difference between prohibition and command exists within the normative sphere, whereas verses of permission are indicative verses that lie outside it. According to this proposal, the difference between negation of obligation and decrees is categorical: it concerns the type of verse, not merely its content.

One may present the matter this way. Clearly, every indicative verse is a permission verse, because a commanding verse is obviously not merely descriptive. But it is not entirely clear whether every permission verse is an indicative verse. There may be permission verses that, in terms of content, do not command anything, yet in terms of grammatical type still belong to the normative sphere. The two possibilities just mentioned disagree on this point. Below, after we define more clearly the “color” that colors normative verses, we will return to this question.

C. Various Problems in Understanding the Principle of This Root

Introduction

In this chapter we will present the main problems raised by the principle put forward in this root. In the next chapter we will propose a solution to those problems.

The Relation to Conditional Commandments

The first point we must examine is the relation between the principle of this root and the topic of conditional commandments. We will first define the concept and then examine it briefly.

The commentators distinguish between two types of commandments: obligatory and existential.⁵ An obligatory commandment is one that one is fully required to fulfill. Whoever performs it thereby fulfills a commandment, and whoever does not perform it commits the transgression of neglecting a positive command. An existential commandment, by contrast, is one that it is worthwhile to fulfill. Whoever fulfills it has fulfilled a positive commandment, but one who does not fulfill it has committed no transgression. With regard to such commandments there is no such thing as neglecting a positive command.⁶

It is also important to distinguish between an existential commandment and a conditional commandment, a distinction that many commentators ignore. An existential commandment is one that one is under no obligation to perform, but if one does perform it one fulfills a commandment. Examples include Torah study beyond the minimum of reciting the Shema morning and evening, or giving charity beyond the amount and in the context in which one is obligated.⁷ A conditional commandment, by contrast, is one whose obligation is fully binding, but only on the condition that certain circumstances arise. For example, the commandment of fringes applies only when one wears a four-cornered garment, and the commandment of ritual slaughter applies only when one wishes to eat meat. In such cases there is a full-fledged commandment, and one is entirely obligated to fulfill it, except that the obligation is conditional on a certain factual situation. If one eats without slaughter or wears a four-cornered garment without fringes, one neglects a positive commandment. These are therefore not existential commandments but obligatory commandments that are conditional, conditional upon the circumstances in which the obligation takes effect.⁸

Conditional commandments are, in a certain sense, nothing but prohibitions inferred from a positive commandment. Put differently: the conditional command can be formulated as a negation. For example, ritual slaughter is a conditional commandment, since there is no commandment to slaughter as such; the command only determines that if we want to eat meat, we must first slaughter the animal. In other words, one could say that this command means not that there is a commandment to slaughter, but that there is a prohibition against eating unslaughtered meat. Up to this point, this still does not touch on the topic of negations. But in another formulation, one may view the conditional commandment like this: the command to slaughter means that the prohibition of “unslaughtered” does not apply when the animal has been slaughtered. In that formulation, it appears that we have a negation and not a command. The “obligation” is the prohibition against unslaughtered meat, and the negation says that this prohibition does not apply when slaughter has been performed.

In this example, however, the command to slaughter is itself what establishes the prohibition of unslaughtered meat. There is no independent prohibition of unslaughtered meat from some source other than the commandment of slaughter. If so, this commandment is not negating another prohibition, and we do not have here a negation of obligation but the definition of a prohibition. This may be the root of Nahmanides’ claim, discussed in our essay on the ninth root, that a prohibition inferred from a positive commandment is counted as a positive commandment, but if it comes in parallel to an explicit prohibition, it is not counted at all. We can now offer another explanation: if there is an explicit prohibition in the background, then this verse is interpreted as a negation and not as a command, and is therefore not counted. But if there is no explicit prohibition in the background, then this command is a prohibition inferred from a positive commandment, which is itself counted as a positive commandment.

Eating in the sukkah, apart from the first night when there is an obligatory commandment to eat bread in the sukkah, can also apparently be formulated as a negation, namely that the prohibition against eating does not apply when one eats in the sukkah. Here it is clearer that this is only a formal similarity, because in reality there is a command to eat in the sukkah, and the prohibition against eating outside the sukkah is derived from this commandment. There is no independent source for that prohibition; it is nothing but an expression of the positive commandment’s existence. The commandment to eat in the sukkah is not really conditional, except in a merely formal sense.⁹

Giving a bill of divorce is a more difficult example, because one can plausibly argue that it amounts to nothing more than the negation of the state of marriage, together with the consequences of that state, such as the prohibition on a married woman and the mutual obligations of husband and wife, which no longer exist after a divorce. More precisely, one might say that all the verse does is tell us that the prohibition of a married woman, which attached to this woman, no longer exists once she has been given a bill of divorce. If so, this is a negation of obligation and not a command. Here it will be hard to say that the prohibition exists because no bill of divorce was given, since the prohibition exists because she is married, even if the procedure of permitting her through a bill of divorce did not exist. For example, with respect to the wives of a rapist or seducer, whose husbands are forbidden to divorce them, the prohibition of a married woman obviously still exists. Our conclusion is therefore that one can indeed say that the commandment of divorce through a bill of divorce is actually a factual disclosure that the woman ceases to be married once the bill is given. If so, there is no commandment here at all, but only a negation.

We should note that in the two examples we gave, slaughter and the bill of divorce, there is an initial assumption to prohibit even without the verse. In the case of slaughter, there is an initial assumption that the prohibition of unslaughtered meat would still apply, and so a verse is needed to permit the meat through slaughter. In the case of divorce, there is an initial assumption that the prohibition of a married woman would remain even if she received a bill of divorce, and therefore a verse is needed to permit her to others through that bill. Only because such initial assumptions exist is there room to interpret both commandments as negations rather than commands. And in fact, in the case of slaughter, the argument we raised—that there is no independent source for the prohibition of unslaughtered meat—means precisely that no such initial assumption exists apart from the verse, and therefore it should not be interpreted as a negation but as a positive commandment. If so, this principle is nothing but an expression of Nahmanides’ earlier remark.

And indeed we find in Maimonides himself, in Positive Commandment 129, that he had already noted the relation between this rule and a prohibition inferred from a positive commandment:

The one hundred and twenty-ninth commandment is that the Levites were commanded to separate a tithe from the tithe that they receive from Israel and give it to the priests, as He, exalted be He, said: “Speak to the Levites and say to them: when you take from the children of Israel the tithe that I have given you from them as your inheritance, you shall raise up from it the Lord’s gift, a tithe from the tithe.” Scripture explained to us that this tithe, called terumat maaser, is to be given to the priest, as it says: “You shall give from it the gift of the Lord to Aaron the priest.”

Scripture further explained that this tithe must be taken from its best and finest part, as it says: “From all its choicest part, its sacred portion from it.” And it further hinted that they sin if they do not separate it from the best, as it says: “You shall not bear sin because of it when you raise up its best part from it.”

This negative expression is a negation of obligation, for it says: you bear no sin when you separate it from the best, and this indicates that if you separate it from the inferior part, you do sin. Its sense is like that of a prohibition inferred from a positive commandment, which is not counted among the negative commandments. That is, after commanding that it be separated from the best, Scripture thereby indicated that it must not be separated from the inferior. And the language of the Sifrei is: From where do you know that if you separated it not from the best, you bear guilt? Scripture says: “You shall not bear sin because of it when you raise up its best part from it.” The laws of this commandment have already been explained in Tractate Terumot, in Tractate Maaserot, and in places in Demai as well.

This is a prohibition inferred from a positive commandment, but it is not included in the enumeration of the commandments, not even as a separate positive commandment. Rather, it is included within the positive commandment of separating the priestly tithe from the tithe. The reason is that it is a negation and not a warning.

Descriptive Verses Counted in the Enumeration of the Commandments: Explanatory Commandments

We saw above that Maimonides argues that verses of negation are indicative or permissive rather than imperative, and therefore should not be counted in the enumeration of the commandments. In this context, it is interesting to cite Maimonides himself on two consecutive positive commandments, although there is no direct connection between them, where he appears to include even indicative, or permissive, verses in his count. We will discuss this, God willing, in our essay on the twelfth root.

In Positive Commandment 95, Maimonides counts the commandment to adjudicate the annulment of vows. At the beginning of his remarks there he writes:

The commandment that we were commanded to judge the laws of annulment of vows. That is, the Torah instructed us to judge according to those laws. The point is not that we are under an obligation to annul in every case. This same point should be understood whenever you hear me count one of the laws, for there is not necessarily a command regarding an action; rather, the commandment is that we are commanded to judge in this law concerning this matter.

Maimonides states that the content of this commandment does not obligate performance, but is rather the disclosure of a legal reality. The same is true in the next commandment, Positive Commandment 96, where Maimonides counts as a commandment that anyone touching a carcass becomes impure. There too he writes:

I will now mention an introduction that you should remember regarding all the types of impurity that we will mention. When each type of impurity is counted as a positive commandment, this does not mean that we are obligated to become impure with that impurity, nor that we are warned against becoming impure so that this would be a negative commandment. Rather, the Torah’s statement that whoever touches this type becomes impure, or that this thing imparts impurity in such-and-such a way to one who touches it, is itself a positive commandment. That is, the law with which we were commanded is a positive commandment—namely, that one who touches this in such-and-such a way becomes impure, and that one who is in such-and-such a condition does not become impure. Becoming impure itself is optional: if he wishes he may become impure, and if he wishes he may avoid impurity… The commandment is what we were told in this law: that one who touches this becomes impure and will be impure, and that the obligations incumbent on the impure person will then apply to him, such as leaving the camp of the Divine Presence, not eating sacred food, not touching it, and the like. This is the commandment, namely that he is impure with this type when he touches it or is near it in such-and-such a way. Remember this point with respect to every type of impurity.

Here too Maimonides says that a principle revealing a legal reality is counted within the enumeration of the commandments. The commandments regarding impurity reveal when a person becomes impure and when not, whereas the laws that apply to him as a result are not themselves part of these commandments. The commandments that apply to the impure person are counted separately in the enumeration.¹⁰

In light of all this, the problematic nature of the principle stated in this root becomes even sharper: why should negations not be counted in the enumeration of the commandments? For example, one could count the principle that a Hebrew maidservant does not go free through injury to the extremities, even if there is no prohibition against releasing her. The very definition that this rule does not apply to her would seem to deserve enumeration as a commandment. Maimonides himself states in the quotations above that commandments are not only commands but also disclosures of reality relevant to the applicability of other commandments, namely to whom or to what they do or do not apply.

Perhaps here we must return to the distinction we raised above between permission verses and indicative verses. If we assume that Maimonides regards these two commandments as belonging to the normative sphere even though their content is not prohibition or obligation but merely permission, we can understand why he counts them. See below for more on this.

The Relation to the Roots of the Fifth Category

In the seventh root, as well as in the eleventh and twelfth roots, which complete the fifth category of roots dealing with sorting and classification, Maimonides argues that details of a commandment should not be counted as independent commandments. The details are included within the general commandment and therefore should not be designated as separate commandments.

One may now ask why such details are not really just negations. Whenever there is some detail in a commandment, that means that in one circumstance the commandment does not apply and in another it does. In effect, this is a negation of the commandment’s applicability in the case under discussion.

And conversely: why should negation not be included under the principle that one does not count details as independent commandments? For example, “She shall not go free as the male slaves do” is a detail within the commandment of emancipation through injury to the extremities, namely that it does not apply to a Hebrew maidservant. If so, it is not clear why Maimonides formulated two different roots for such cases.

On closer inspection, however, it seems that whenever the Torah determines that a certain commandment does not apply to someone or to a particular situation, this will not be counted as a commandment because the verse is a negation, not because it is a detail of a commandment. The roots that deal with not counting details of commandments address parts of one act that together make it up, such as stages of sacrificial procedure, or different appearances of one commandment in different circumstances, or component parts of commandments. Statements that negate the applicability of a commandment in a certain case are indeed the subject of this root.

According to this argument, the present root is indeed another root dealing with details of commandments, except that these details are not particular laws within the commandment, but indications of the scope of its applicability, that is, a determination that it does or does not apply in a given case.

But that does not exhaust the matter. From Maimonides’ language, and especially from his linguistic and logical introduction, which we discussed above, it appears that his main claim here is that negations are not counted because they are indicative verses. According to this, the distinction between this root and the roots of the fifth category is that the roots dealing with details of commandments concern imperative verses that are not counted because they form part of a broader commandment already counted. Those are problems of duplication, sorting, and classification. By contrast, in this root the principle preventing enumeration rests on the fact that negation verses are indicative and not imperative. In other words, it is rooted in the type of verse, not in its content or in overlap with other verses. It therefore still seems that we have here a different kind of principle. These verses are indicatives because they negate other norms, whereas verses that add details are imperative verses but are subsumed within broader commandments.

This distinction sharpens our earlier claim that the basis of this root differs from that of those others. Here the basis is the distinction between indicative and imperative verses. For that reason we included this root in the fourth category rather than the fifth. This also serves as proof that Maimonides understood the basis of this root to be that these are indicative verses, not merely that their content is permission. We will explain this further below.

The Relation to the Fifth Root

The relation to the fifth root is certainly closer. As we have seen, the present root belongs to the same category as the fifth root, because in both cases the reason for non-enumeration is that we are dealing with indicative verses.¹¹ In our essay on the fifth root we pointed out this duplication between the two roots, and saw that it already arises in Maimonides’ own words. We will return briefly to that subject below.¹²

As already noted, the fundamental discussion in the present category, namely the discussion of the principled distinction between indicative and imperative verses, takes place in this root. Clarifying that discussion will be the subject of the next chapter.

D. Prescription and Description

Introduction

In this chapter we will deal with a problem much discussed in the analytic philosophy of morality: the meaning of commands, moral and otherwise; the relation between commands and facts; and the relation between the two corresponding types of sentences. We will see that this issue underlies Maimonides’ determination in this root, and in fact the distinction itself is the point of Maimonides’ logical-linguistic introduction to this root.

The Reality Underlying the Command

The relation between command and reality was also discussed in our essay on the fifth root. There we saw that one can understand a command as derived from some underlying reality, which is the reason for the commandment. One might perhaps even identify the commandment with the reason itself, and claim that the command is not what appears in the command-verse but what appears in the verse giving the reason. For example, if the Torah commands: “Her first husband who sent her away may not take her again… and you shall not bring sin upon the land,” then perhaps the Torah is not forbidding the divorcer to remarry his former wife after she married another, but rather forbidding him to bring sin upon the land. Remarrying one’s divorced wife after her subsequent marriage merely causes sin to be brought upon the land, and therefore it is a forbidden act; but the real prohibition is the reason itself, bringing sin upon the land.

According to this conception, the same is true even with regard to commandments whose reason is not explicitly stated in the Torah. The assumption remains that there is a reason underlying the commandment, unlike the “weak-minded” described by Maimonides in the Guide of the Perplexed, who think that commandments have no reasons. For example, the prohibition against taking a widow’s garment in pledge rests on the fact that doing so gives her a bad reputation among her neighbors. There too one could certainly argue that the real prohibition is giving her that bad reputation, while taking the pledge is only the act likely to lead to that prohibition, and not the prohibition itself.

Thus, at the basis of every prohibition or commandment there stands some factual or evaluative determination.¹³ Usually such a determination concerns the consequences of the act: if such-and-such an act is done, these consequences will result. Yet one may also understand that these reasons are not facts in the ordinary sense, but reasons—moral or otherwise. Bringing sin upon the land is not necessarily a fact in the ordinary sense, but the moral consequence of the forbidden act. Still, from the logical-linguistic point of view, bringing sin upon the land is a fact; otherwise it would not function as a reason for the commandment, but would merely repeat the command in other words.

Natural Law

We may now ask whether the Torah’s disclosure that the land is morally tainted when a person remarries his former wife after her subsequent marriage immediately translates into a command not to do so, or whether we still need an explicit command of the Torah forbidding one to taint the land. If we knew that the land becomes tainted when a man remarries his former wife after her remarriage, would that alone suffice to say that it is forbidden? And conversely: is a command nothing more than a factual disclosure of the factual or moral consequences of the act, or is there something in command beyond the statement of that fact?

On one side of this question, a verse stating that Sabbath desecration has problematic consequences of one sort or another would be entirely equivalent to a verse forbidding us to desecrate the Sabbath. But from our discussion in the fifth root it is quite clear that this is not so. These are two different types of verse: the first is factual, an indicative verse, and the second normative, an imperative verse. What is the relation between them? What is the connection between the fact that underlies the command, that is, the reason, and the command itself? What carries us from the one to the other?

This question leads us to the dispute between two approaches in legal theory: natural law, usually associated with Thomas Aquinas, and positivism. The natural-law approach regards the laws of a legal system as commands whose validity derives from their being correct, usually morally correct. According to this approach, a law draws its authority from the fact that it expresses something true and right, not necessarily from the fact that it was enacted by some formally proper procedure. The legislative process, on this view, is secondary; its main function is to bring the moral norm to the citizens’ attention. According to this view, the law is really just an indicative sentence, informing us that this form of conduct is good and beneficial. The obligation to obey derives from the very fact that it is good and proper. This is, de facto, an identification of law with morality, albeit in a broader sense of that term.

By contrast, the more common contemporary approach, strengthened significantly after the Second World War, is the more formal approach, which sees the basis of a law’s validity in the fact that it was duly enacted. A law binds not necessarily because it is right or moral, but because it was properly legislated by the competent institutions and through the authorized procedures. This approach makes a sharp distinction between indicative sentences, which tell us facts, and imperative sentences, which command us—that is, laws. A sentence that teaches us that some act is right expresses only a fact, albeit a moral fact. But it still does not instruct us to do it. There is thus a clear difference between indicative and imperative sentences, and in the legal context what connects them and leads from one to the other is the process of legislation. Here a clear distinction arises between morality and law.

There are many differences between these two approaches. For example, the first naturally supports purposive interpretation of the law, since the moral purposes of the law stand at the basis of its validity. The second does not necessarily support such methods. It will tend more toward textual-formal interpretation, that is, interpretation aimed at deducing instructions from the enacted text as far as possible. What is found in the language of the statute is what has been duly enacted, and therefore that is the valid and binding law. Purposes belong to interpretation and extend the law in light of its goals rather than exposing its content, and it is therefore difficult to view them as norms that were themselves enacted in a binding way.¹⁴

Of course, there are several intermediate positions. Today there is certainly some recognition of basic moral norms as binding even within the legal sphere—for example, through principles such as the refusal to recognize a manifestly illegal order as binding. This is an expression of a moral principle that never underwent formal legislation and yet affects the law, even limiting it. It is therefore hard today to find anyone who advocates either of these two positions in perfectly pure form. But these two poles underlie the different schools in legal theory, which endorse combinations of them in varying proportions. At bottom, however, it is now clear that some act of legislation is generally required in order for a law to have force, though as noted there are various qualifications.¹⁵

The Naturalistic Fallacy

The dilemma we noted above is rooted in the question of the relation between norms and judgments, on the one hand, and facts on the other. The Scottish philosopher David Hume already pointed out that one cannot derive judgments or norms from factual premises. For example, from the premise that the sky is blue one cannot derive the conclusion that the sky is beautiful. The reason is that the premise is factual in nature, while the conclusion is a judgment. In Hume’s language, one cannot derive an ought from an is. Therefore one also cannot say of evaluative statements that they are true in the factual sense.

Another example: from the fact that a dog licks us, one cannot infer that he is our best friend. Such an inference would be invalid. To validate it, one needs an additional premise, namely that whoever licks us probably loves us and is our friend. Likewise, one cannot infer the conclusion that the picture before me is beautiful merely from the fact that it is rich in colors or arranged in some specific manner. To validate that inference one needs an additional premise, such as the claim that variety of color is beautiful. In exactly the same way, one cannot infer from the fact that giving charity improves the poor person’s condition the ethical conclusion that it is morally good to give charity.

All these cases are transitions from factual premises to moral or evaluative conclusions. The conclusion is that norms and facts are different spheres that do not “speak” to each other. One cannot automatically pass from the factual sphere to the ethical, aesthetic, legal, or halakhic sphere—that is, neither to norm nor to judgment.

Bridge Principles and Legislation

How, then, can one bridge this gap? How do we move from facts to judgments, norms, or values? The transition occurs through bridge principles—principles that bridge these two alien spheres. Such principles combine factual and evaluative elements and connect them. In the example above concerning the beauty of the sky, in order for the inference to be valid one must add a bridge principle, that is, a principle that is not purely factual but already contains an evaluative component, such as: everything blue is beautiful. This is a principle that links the fact, blueness, with the judgment, beauty. Only if one accepts this bridge principle as an additional premise does the inference become valid. One should note that in such a case we have not derived a judgment from facts alone, but from premises that themselves already include an evaluative component. In this way, we avoid the naturalistic fallacy.

The same applies to moral judgment. One cannot infer that striking another person is morally wrong merely from the fact that the blow causes pain and suffering. In order to derive the value judgment, we must add as a premise another bridge principle linking fact with value, such as: causing suffering to another person is morally reprehensible. That bridge principle is what underlies the moral argument. In fact, it is itself the moral premise of the argument; all the rest are facts.

The bridge principle of the legal world is legislation. The fact that there is a moral fact does not mean that we thereby have a legal obligation. The problem here resembles the one expressed by the naturalistic fallacy, because it derives a norm from a fact, in this case a moral fact. The legal and moral spheres too do not “speak” to one another, at least in the positivist conception. What moves us from one to the other is legislation. Legislation is a type of bridge principle that validates legal norms, so to speak carving them out of the moral sphere.

Between Norm and Obligation

So far we have dealt with the relation between a factual determination and the norm supposedly derived from it. We have seen that one cannot derive a norm from an indicative verse unless one adds some bridge principle.

But we must now examine another seam that cannot simply be skipped. Suppose we have adopted some norm-sentence—for example, “Sabbath desecration is bad.” One may say that this norm is accepted by us as correct and fitting. Let us set aside for the moment the question of how such a norm is derived from factual statements, if at all. Another problem remains: what is the relation between accepting a norm-sentence and being obligated to do what it directs? Does the statement “Sabbath observance is good and proper” automatically mean that one is obligated to keep the Sabbath? Does the claim that something is a “normative fact,” as distinct from an ordinary fact, itself mean that there is an obligation to realize it? Is there not a gap between validating a norm and imposing an obligation to fulfill it? Here too there seems to be a Humean gap, which does not allow us to derive obligation from a normative determination, even though that determination is itself only a certain kind of fact, a “normative fact.” This transition too seems to involve a version of the naturalistic fallacy.

This is the problem of the descriptivity and prescriptivity of legal and moral sentences, and it arises in other contexts as well, such as philosophical ethics and the philosophy of law, when one discusses the meaning of sentences in these fields.¹⁶

The Emptiness of the Concept of “Good”

Some philosophers of ethics noted that the statement “You should do this act because it is good” is empty. The very meaning of the concept “good” is that one should do it. If so, the earlier statement, which appears to present a reason for performing the act, provides no reason at all but simply repeats the same thing in different words. Its real meaning is: you should do this act because it is an act that should be done. A similar question can be raised regarding all evaluative concepts, ethical or aesthetic. “You should appreciate this picture because it is beautiful” also appears to be an empty sentence, since “beautiful” means that which is worthy of aesthetic appreciation.

This line of thought presents the concept “good,” or “beautiful,” as empty. It contains nothing beyond the claim that this is what ought to be done. The term “good” is simply the set of norms or acts that ought to be done and by which one ought to conduct oneself.

And yet the sense remains that this concept does have content of its own. That is, the feeling is that the argument that one should do some act because it is good is not an empty sentence. We grasp the concept “good” as a concept with independent content, from which one may derive the obligation to perform certain acts, or at least the claim that such acts are positive and worthy norms. Below we will clarify how such an argument can still have meaning, even though at first glance it seems circular.

Essentialism and Conventionalism

In the book Two Carts and a Balloon, note 8, this dilemma is presented as an example of an essentialist approach to concepts. The basic claim put forward there is that concepts have an existence of their own; that is, they are not merely conventions, agreements within a community of speakers, that serve as shorthand for some bundle of properties and nothing more. With regard to objects such as a table, it is fairly clear that we do not identify them with the totality of their properties. For example, we do not say that a table is nothing more than the collection of its properties, such as having four legs, being brown, being of a given height, and so on. Clearly there is an object that is the table itself, and that object has various properties. This is the distinction between substance and accident, or between matter and form.

A similar distinction is made there with respect to concepts. For example, the concept “democracy” is not merely a shorthand label for a form of government with elected representatives, separation of powers, civil rights, and so on. Democracy itself is a kind of entity, the very essence of the concept, and that concept has several characteristics or properties: elected representatives, separation of powers, and the like. These are its accidents, or its form. As explained there, otherwise one could not alter the definition of a concept or argue about its meaning.

One of the illustrative arguments offered there concerns moral relativism. Sometimes there is a dispute about an act: is it moral or not? For example, taking the elderly out into the snow to freeze to death, which the Eskimos regard as proper behavior, while in our culture it is seen as an unacceptable act. We must then ask ourselves whether the term “good” is being used in the same sense by both communities. The question is not whether it has the same properties or applications, for that is clearly not so. The two communities sharply disagree about how to apply it. The question is about the meaning of the concept “good” in the sentence “Taking the elderly out to freeze in the snow is a good act,” a valid sentence within Eskimo ethics, as opposed to the sentence “Taking the elderly out to freeze in the snow is not a good act,” valid within our ethics. Is the term “good” functioning in the same sense in both contexts?

One could say that the two cultures use the same word in two different senses; in reality, these would be two different concepts, and the shared name would be merely accidental. But if that were so, then there would be no disagreement at all. The Eskimos would agree that taking the elderly out to the snow is “not good” in the sense we assign to “good,” and we would agree that it is good in their sense. The solution to such a “dispute” would simply be to distinguish the concepts. The concept “good” in Eskimo culture could be renamed by some unrelated word, and then there would be no problem and complete agreement: taking the elderly out to the snow is indeed not good, but it does satisfy that other concept.

This solution seems rather absurd. The reason is that our intuition tells us that there is a real disagreement here, and not merely semantic confusion. But if there truly is a genuine disagreement, and if it really cannot be dissolved by this sort of semantic separation, then we are forced to assume that the concept “good” has the same meaning in both contexts. Now it becomes clear that if I say act X is good and another person says it is not good, then there is a disagreement between us about some feature or application of the concept “good.” The very existence of a disagreement is thus strong evidence for the universality of the meaning of the concept under dispute.

This universal dimension is what we there called “the concept as such.” The concept “good,” in its universal sense, is the category shared by all cultures. One may say that this is the connotation accompanying statements of the form “X is good” in all cultures, even when there are sharp and polarized disputes over which X’s should be inserted there in order to produce an ethically valid statement.

Thinking and Cognition

In the book Enosh KeHatzir, one further step is taken in the direction of making sense of these disputes. Because a full justification of that picture would require much more detail, I will present here only the conclusions.

Ordinarily, we regard moral judgment as the product of thinking. As such, it is a subjective mechanism operating within the human mind, with no relation to the objective world outside. This gives rise to the approach that insists on a separation between facts and values or judgments. But in light of our earlier suggestions about the existence of the concept “good” as such in some objective sense, beyond its specific characteristics and applications, we can now understand moral judgment as a kind of contemplation. Moral norms are the result of contemplating the concept “good,” or the idea of the Good, and its characteristics. Disagreements are nothing but different perspectives on that concept.

Such contemplation is not done with the eyes, of course, but with “the eyes of the intellect,” in Maimonides’ phrase at the beginning of the Guide of the Perplexed; or as “contemplation of ideas,” in Edmund Husserl; or as “auditory logic,” in the language of the Nazir. All these expressions are different ways of blending elements of thought and cognition. If one pays attention, one will notice that each such phrase combines an epistemological notion with an intellectual one.

If so, the formation of moral norms is the result of a process of cognition rather than mere thought. It is a process based on a relation to some objective concept outside ourselves, in the world of ideas, and not on a subjective and relative process taking place solely within us.

We can now see that the argument that one should do act X because it is “good” is not empty. It has meaning, because it means that contemplation of the concept “good” yields the norm that obligates us to perform that act. The concept “good” has significance beyond the collection of its specific applications.

Once Again: Facts and Norms, Prescription and Description

If the result of ordinary observation is a fact, then the result of this sort of moral contemplation is a norm, or a judgment. Norms and judgments are indeed not the result of facts in the ordinary sense, and in that Hume is certainly correct. But they are the result of another kind of contemplation, whose objects are “moral realities,” and whose outcomes are “moral facts.”

Ordinary facts, such as the color of a table or the height of a tree, are neutral facts. The sentences that describe them are indicative or descriptive sentences. These facts, and these sentences, are transparent and devoid of special charge. They stand on their own and, as such, have no evaluative or normative consequence, in light of Hume’s naturalistic-fallacy argument. Moral or normative facts, however, are of a different character. These are “facts” charged with value. They are not neutral, but move us to act or refrain from acting. They are not merely neutral descriptions.

We asked above how the “fact” of the existence of a norm is connected to our obligation regarding that norm. The answer is that a normative fact differs in essence from an ordinary fact. It is not neutral, but charged with a motivating force, and anyone who apprehends it thereby grasps his obligation toward it. If he does not experience such an obligation, then he has not actually seen that normative fact. This “color” or “charge” bridges the naturalistic fallacy. A moral or legal fact does not require a bridge principle; once we recognize it, our obligation toward it is clear. Here the norm does indeed flow from the fact, because such “facts” are of a different character.

In fact, the distinction begins even deeper. The “object” we contemplate in moral judgment is the idea of the Good. That “object” is different from ordinary objects. It is colored or charged with a special force, so that whoever perceives it understands that an obligation rests upon him. Therefore its “features,” or applications, are charged with that same force. Its features are not color or height or other physical characteristics, but specific moral or legal directives applying to specific situations. These are norm-sentences, colored with the normative color. And from that color there follows the practical obligation of fulfillment and obedience.

This distinction was already made by the philosopher R. M. Hare, who distinguished between prescriptive and descriptive sentences. According to Hare, the value statement is not merely descriptive; implicit in it are both an evaluative judgment and a motivation to action. The sentence “You shall love your neighbor as yourself” does not merely describe love of others as a positive moral experience, but claims that there is a moral obligation to love one’s fellow, and that this obligation rests on everyone. Hidden inside the apparently factual claim that love of others is a moral obligation is also a command to realize it. In other words: it is not an indicative sentence but an imperative one. An imperative sentence is charged with a motivating “load,” or colored with a special “color.”

If we consider a descriptive sentence such as: “In the accepted moral code of such-and-such a tribe there is an obligation to kill the old man and eat his flesh,” it is clear that no obligation to do so follows for us. Therefore this is description, not command. But a person who belongs to that tribe, that is, someone who accepts the evaluative judgments of that moral system, when he sees such a sentence will say that the obligation to kill the old man and eat his flesh is not a descriptive sentence but an imperative one. From his perspective, the sentence does not merely state a neutral fact but a moral fact, and therefore we have here a command, from which follows an obligation of obedience and practical realization. Hare’s implicit assumption is that one cannot derive the value judgment from the mere facts, and therefore when one utters a sentence containing a value judgment, one does not mean to describe but to command.

In these terms, one may say that normative sentences are prescriptive and not descriptive. They are not descriptions of neutral facts, but directives that move us to action. Such sentences are “colored” in a non-neutral way, and one can indeed derive from them halakhic or moral obligation, or ethical or aesthetic judgment.¹⁷

This is the “color” Maimonides is trying to define in his opening sentences of this root. It is the common element of positive and negative commandments alike, because both kinds of sentences are colored with this motivating force. The command directs us toward action and has practical consequences. It is not a neutral and indifferent fact.

“For Its Own Sake”

In the final gate of the book Enosh KeHatzir, the concept of doing something “for its own sake” is discussed at length. When a person asks himself why he fulfills commandments, the answer cannot be some moral or other principle, such as gratitude, and the like.¹⁸ Commitment to the service of God is the most basic reason, from which all other commitments derive their authority. As explained there, this is the meaning of serving God “for its own sake.” A person serves God for no prior reason at all—simply because.

Our moral obligations, like our Torah obligations, are not fulfilled because of some super-principle, but by virtue of the fact that they are binding norms or commands. That itself is a sufficient reason for our commitment to them; no other reason is needed. This “reason” is the charge that accompanies these normative sentences or commandments. That charge is what moves us to action, and it is the most basic reason for our obligations in these domains. The search for some other, more external reason is doomed to failure. For example, the question why I am obligated to moral principles has no answer and cannot have one. I am obligated because they are moral principles. The determinations that establish them are colored and charged with a force that obligates me to action by virtue of their being norms and commands, and so no further justification is required. One may even say that if a person is committed to these norms for some other reason, he is not a genuinely moral person or a true servant of God. He serves that other reason which causes his commitment. Service of God or moral commitment means unconditional commitment, arising from the very fact that these norms are binding norms.¹⁹

We saw that questions about the authority of moral principles are sometimes answered analytically and conceptually: the very meaning of the term “good” is that we ought to obey whatever falls under it. In this terminology, one may say that we fulfill it “for its own sake.” This is another way of saying that the concept “good” is prescriptive and not descriptive, that is, that it is “colored” with the color of commandments, or charged with the force that ought to move us to action. The term “for its own sake” expresses the normative color or motivating force that belongs to imperative sentences. Performing a commandment for its own sake means performing it from that motivation of commitment to halakha.

Interim Summary

Up to this point we have seen three stages in the formation of a norm:

1. The fact. This is the reason for the norm.
2. The norm-sentence derived from that fact. This is a “normative fact,” whether moral or legal.
3. The practical obligation, derived from the prescriptivity of the norm-sentence, that is, from its special “color” or “charge.”

We explained the emergence of stage 2 by saying that contemplation of the idea of the Good yields norm-sentences. In that sense, normative sentences do have an anchor in objective reality. Not in ordinary objects, to be sure, but certainly in moral objects, that is, ideas.

We then asked how the existence of norm-sentences, which are after all also a kind of fact, produces motivation to action or obligation. The answer is that these facts, learned through contemplation of the moral object, are not neutral. They are colored in a special way and carry a “charge” that moves the knower toward action and commitment. In our terms, these are not neutral facts or descriptions, but directives and commands. A command is not merely a different grammatical tense, but a different type of sentence.

This entire analysis is relevant to every system of norms or judgments, and not only to morality or aesthetics or any other particular field. Among other things, it can be applied to halakha, as we shall now see.

E. Returning to the Dispute Between Maimonides and Nahmanides

Introduction

After the philosophical preliminaries given in the previous chapter, we can now return to the words of Maimonides and Nahmanides and understand their positions and the disagreement between them.

Applying This to Maimonides’ Discussion of the Root

Above we distinguished between two kinds of verses: indicative and imperative. But afterward we raised the possibility of a third category: normative verses. The conclusion is that the Torah contains three kinds of verses:

1. Indicative verses that describe a fact. This obviously includes ordinary facts, such as the Israelites encamping at some place in the wilderness. But here our main concern is with a different kind of indicative verse: moral facts. Examples of this category are the reason-verses discussed in the fifth root. These verses describe moral facts, such as the fact that remarrying one’s former wife or accepting ransom for a murderer pollutes the land.
2. Normative verses that state a halakhic fact. Among these are verses that command us through positive commandments or prohibitions, and also verses that negate some halakhic obligation. All verses whose content has halakhic significance belong to this category. A verse such as “Honor your father and your mother,” as well as “Do not boil a kid in its mother’s milk,” and also “She shall not go free as the male slaves do”—all these belong to this category. All have halakhic meaning. They impose obligation, prohibition, or present the absence of a norm, that is, permission to do something.
3. Within normative sentences, there is a subset Maimonides calls “decrees.” These are what we call commandments, both positive and negative. Such verses are charged with the energy or motivating force to act or refrain from acting. They are not neutral facts but charged facts, precisely like the charged facts of the moral domain discussed above.

A negation of obligation is indeed a normative sentence. But it is not charged with the special force that characterizes commandments. It does not move us either to action or to restraint. It merely determines that in the context under discussion there is no binding norm. It is an uncharged fact. As such, it certainly has halakhic significance, and therefore belongs to the normative sphere, but it is not a commandment. That is why Maimonides rules that it should not be included in the enumeration of the commandments.

We can now distinguish between sentences that do not belong to the normative sphere at all, namely those of the first category alone, and sentences that do belong to that sphere but describe permission or allowance, such as negations of obligation or negations of prohibition. Our answer here to the dilemma raised above is that there really is a difference between a permission-sentence and a non-normative sentence. A negation of obligation is a permission-sentence, but it certainly belongs to the normative sphere, because it has halakhic significance. Therefore it enters the Shulhan Arukh, but not the enumeration of the commandments. Only charged and colored verses enter the enumeration. Even normative verses, if they are not charged and colored with this special color, will not be included in the count.

Against this background, we may perhaps try to understand the relation between negations of obligation and explanatory commandments such as Positive Commandments 95 and 96 mentioned above. We saw that Maimonides includes in his count commandments that do not impose any practical obligation on us and do not prohibit anything, such as the determination that one who touches a corpse is impure. In our terminology, these are verses with halakhic significance, but on the face of it they do not seem to belong to the category of commandments. They are not charged with the force that moves the reader to action, since they contain neither positive nor negative command. They are descriptive rather than prescriptive.

Nevertheless, there may be room to distinguish them from negations of obligation. A negation of obligation determines the absence of some norm, and is therefore merely a neutral fact, that is, a fact rather than a “moral fact,” and thus we have here an indicative verse. For that reason it is not included in the enumeration of the commandments. By contrast, these verses establish the existence of a norm which, though it does not directly impose a practical duty, may still perhaps be viewed as colored with the “color” of imperative verses. We will explain this more fully in the next subsection.

The Dispute with Nahmanides: Two Different Conceptions of the Principle of This Root

Nahmanides, as emerges from the whole course of his discussion, understands the principle of this root to be that verses indicating acts of permission are not counted within the enumeration of the commandments, unlike Maimonides, who understands that indicative verses are not counted. In other words, Nahmanides understood this root as a distinction based on the content of the verse—permission versus prohibition or obligation—and not on its grammatical mode. Therefore Nahmanides would not count commandments such as Positive Commandments 95 and 96, which are normative verses whose content is permission or legal option.

It is important to understand that according to this approach there would have been no need at all to engage in the grammatical analysis of sentence structures. If Maimonides had meant the principle as Nahmanides understands it, it would have sufficed for him simply to point out that these are permission-verses. Moreover, positive commandments like 95 and 96 should then not have been included in his count.

It is therefore clear that Maimonides did not intend the distinction Nahmanides attributes to him. In his view, negations of obligation are not counted not because they are permission-verses, but because they are verses that do not belong to the normative sphere at all.²⁰

This distinction returns us to the logical illustration we presented in our essay on the sixth root. There we pointed out that the negation of a norm does not itself belong to the normative sphere. Such a negation takes us outside the normative sphere. The statement that some norm does not apply in a certain situation is not itself a statement on the normative plane.²¹

This is what distinguishes commandments such as 95 and 96, which establish the existence of norms, from a negation of obligation, which denies the existence of a norm. The former lie within the normative sphere; the latter does not.

Nahmanides, however, will not distinguish between these two types of sentence, because in his view the content of those commandments is not obligation or prohibition but permission, and therefore he does not see them as commandments. For the same reason, he also does not see a negation of obligation as a commandment, since it too is only permission.

It is interesting to note that in the fifth root, whose connection to the present root we have already noted both here and in our essay on it, Maimonides rules that reason-verses are not counted. Nahmanides, in his glosses there, argues that what seems to Maimonides to be a reason-verse is often in fact a command-verse. It seems that there too, as here, Nahmanides tends to interpret various verses as command-verses, and only when left with no choice does he interpret them as indicative or permission-verses, whereas Maimonides interprets them as indicative more readily. This continues the pattern we saw above.

Earlier we brought the dispute between Maimonides and Nahmanides, in the continuation of Nahmanides’ gloss to the fifth root, about how to understand the tannaitic dispute in Zevachim regarding “and he shall not profane.” It may be that there too the basis of the disagreement is whether a negation of obligation is a normative verse whose content is permission, or whether it is an indicative verse. According to Nahmanides, this verse is a normative verse, and therefore the discussion revolves around which norm is learned from it. But according to Maimonides it is an indicative verse, and therefore the tannaitic dispute is not about what one learns from it, but whether there is some initial assumption that teaches us that the verse is only a negation and nothing more.

Further Implications of This Presentation of the Dispute

We have seen that Maimonides and Nahmanides disagree on how to conceive the principle of the present root. According to Maimonides, negations are indicative verses and not normative verses, whereas according to Nahmanides they are normative verses whose content is permission.

A. The interpretation of negation-verses. It is possible that the dispute between them over the default interpretation of negation-verses is also connected to these two conceptions. Nahmanides sees negation-verses as normative verses whose content is permission. Why should the Torah inform us that something is permitted? Only if there is an initial assumption that some different norm would apply there. Therefore, if there is no clear reason to interpret the verse as a permission-verse, we prefer to interpret it as a verse that commands or forbids something. By contrast, according to Maimonides, since these are ordinary indicative verses, there is no principled reason to deny their existence. They are verses that do not belong to the normative sphere, and therefore there is no reason to assume that they must be commanding or forbidding.

B. The duplication with the fifth root. Above we raised the problem of duplication between this root and other roots. The conclusion there was that the problem exists only with respect to the fifth root. In light of the present discussion, we may add that according to Nahmanides these are obviously quite different principles. The fifth root says that reasons are not counted, and those are not normative verses at all, whereas the eighth root says that acts of permission are not counted, even if they appear in normative verses. By contrast, according to Maimonides, the eighth root says that indicative verses are not counted, and not specifically acts of permission as Nahmanides thought. If so, from Maimonides’ perspective the fifth root would seem to be a special case of the eighth root.

One should note, however, that according to Maimonides negation-verses are a slightly different kind of indicative verse. Negation-verses tell us of the nonexistence of a norm, whereas reason-verses tell us of some fact, albeit a “moral fact.” It therefore seems that there is still some difference of principle between the present root and the fifth root, which says that reasons are not counted.

C. Conditional commandments. Above we discussed the relation between the principle expressed in this root and the concept of conditional commandments. According to Maimonides, acts of permission can be counted among the commandments, since they are normative verses, as in Positive Commandments 95 and 96. Conditional commandments, however, are not indicative verses but conditional imperatives. The grammatical mode of such verses is imperative, and therefore they are normative verses and should be counted. According to Nahmanides, by contrast, even norms of permission are not counted in the enumeration of the commandments, and therefore the question remains why conditional commandments are counted.

It would seem that conditional commandments are not acts of permission, because permission is two-sided: one is permitted to do and permitted not to do, and whether one does or does not act carries no halakhic significance. A conditional commandment, by contrast, is a commandment by which the person is fully bound once the relevant circumstances arise, and certainly if he fulfills it he thereby fulfills a commandment. It is therefore clear that even according to Nahmanides such commandments must be counted. Their content is not mere permission.

Another Practical Difference: A Note from the Taz

The Taz, Orah Hayyim 588:5, explains the rabbinic decree not to sound the shofar on Rosh Hashanah when it falls on the Sabbath, and brings the question of Rabbi Elijah Mizrachi as to why they did not prohibit sounding it on every Rosh Hashanah. The Taz answers:

We do not blow the shofar—because perhaps one will take it in his hand to go to an expert to learn and will carry it four cubits in the public domain. The Mizrachi asked: on a festival too let us decree because perhaps he will repair a musical instrument. He discusses this at length and in the end solves nothing. It seems to me that the sound of the shofar was not given for music but for commandment-blowing, and therefore they did not decree such a decree. I would also answer that they have no right to decree and thereby entirely uproot a Torah law that commanded us to blow on this day. On the Sabbath, however, it is not entirely uprooted, since when Rosh Hashanah does not fall on the Sabbath we do blow.

He then offers an explanation for his position, namely that rabbinic decrees are supposed to serve as a fence around the Torah and therefore clearly cannot operate against an explicit Torah law:

It seems that this is included in what they said: they made a fence for the Torah, not in order to transgress Torah law. For to uproot it entirely would be to transgress Torah law. This also helps explain what the Beit Yosef cites from the Ran: why does circumcision override the Sabbath, and why do we not decree lest one carry the child four cubits in the public domain? He answers with a strained distinction. According to what I have written, it is clear, for the Torah explicitly included “on the eighth day,” even on the Sabbath, and they did not wish to uproot an explicit Torah law because of a decree.

He repeats this in Yoreh De’ah 117:1, where he proves it from Tosafot on Sukkah 39a:

It seems that the Sages have no power to forbid something that the Torah explicitly permitted. The Sages have power to be stringent only where there is neither an explicit prohibition nor an explicit permission in the Torah, but not where the Torah explicitly permits.

There he brings support from the rule that a court can strike and punish beyond the strict law:

This too is included in their statement: I heard that a court would strike and punish not according to the law, not to transgress Torah law but to make a fence for the Torah. The difficulty is obvious: surely no human being has the power to transgress Torah law. But according to the premise I mentioned, it becomes even clearer, because the point here is that even to be stringent they have no power where thereby they pass against explicit Torah law.

See also the discussions in Torah Temimah.

Now there is room to ask whether negations of obligation count as something explicitly written in the Torah for this purpose. For example, can the Sages enact a duty to emancipate a Hebrew maidservant through injury to the extremities? To be sure, there is no halakhic duty to keep her under the master’s control, but the Torah explicitly states that there is no duty to release her. Prima facie, this would seem to be something explicitly permitted by the Torah, namely that the master may keep her, and therefore the Sages should have no power to impose a rabbinic duty to release her.

At first glance, this may depend on the dispute between Maimonides and Nahmanides. According to Nahmanides, the verse establishes a norm of permission, and therefore one cannot enact a contrary obligation. But according to Maimonides, there is no verse establishing such a norm at all; it only establishes the absence of the opposite biblical norm, namely that there is no obligation to release her. The absence of a biblical norm does not contradict the possibility of a rabbinic ordinance creating such a norm. If so, according to Maimonides there seems to be no obstacle to instituting a rabbinic enactment against a negation of obligation.

Put differently: if one can indeed infer from this verse that the Sages may not enact an ordinance to release her, then the verse has operative normative force with halakhic consequences, and we ought to treat it as a command. And if, according to Maimonides, a negation of obligation is not a command, then it would seem that the Sages may enact a rabbinic ordinance contrary to it.

This is another possible practical consequence of the dispute between Maimonides and Nahmanides.

A Note on Optional Acts

We saw that according to Nahmanides, negation-verses are verses whose content is permission. But does the Torah and halakha actually recognize acts that are simply optional? Are not all acts either prohibited or commanded?

It is commonly said that the medieval authorities disagree on this. Hovot HaLevavot, in Gate of Divine Service, chapter 4, concludes that from a halakhic point of view there are no optional acts at all. By contrast, the Tanya, chapter 7, explains that kelipat nogah, which mixes good and evil, is the source in this world of optional matters. He elaborates there that such optional matters, such as eating and drinking, can be elevated and rectified when done for the sake of Heaven. Yet it seems from his words that without such rectification they are not prohibited acts but remain optional acts lacking value, unlike the position of Hovot HaLevavot and others.²²

It is interesting that specifically Maimonides, in his commentary on the first mishnah of Avot, on the saying “I found nothing better for the body than silence,” explains that there are several types of speech in halakha, one of which is optional speech. His view is therefore that there are optional acts even from a halakhic perspective.

Still, all this is not directly relevant to our present issue. It is very hard to deny that, from the standpoint of pure halakha, there are acts that are optional. The question discussed in the sources just mentioned is whether there are acts that are neutral in value, not merely in halakha. Whenever a person does something unnecessary, he is neglecting Torah study and could have done something of value. Is there a formal transgression here? In many cases, certainly not. Therefore, from a halakhic standpoint this is indeed an optional act, even if from a value-oriented standpoint one may question whether there really is such a category as “optional.”

F. Additional Halakhic Perspectives

Introduction

We have seen that there is a sharp distinction between an indicative verse and an imperative one. We also saw that even the expression of a value judgment is only an indication. For there to be a command, we need a verse that commands us; mere expression of a negative evaluation of something does not suffice. In this chapter we will examine several additional halakhic aspects connected to this issue.

“No Punishment Unless There Was Prior Warning”

A well-known halakhic rule states: “No punishment is administered unless there was prior warning.” When the Torah presents a verse imposing punishment for some act, the Talmud does not consider that sufficient and always searches for a warning-verse in addition to the punishment-verse.

The author of Sefer HaHinukh, in Commandment 65, explains that without a warning we would think that the punishment is merely a mechanical consequence, that is, there is a condition that whoever performs such an act receives such a punishment, but there is no violation of God’s will involved. Therefore, a warning-verse is required to tell us that such an act stands in contradiction to the command of the Holy One, blessed be He.²³

This could also be explained somewhat differently. Without a warning-verse, we might think that the act is not halakhically prohibited. The Torah does indeed attach a punishment to this act, but perhaps the punishment is imposed for doing something morally or evaluatively wrong, even if it is not formally defined as a legal transgression. The Sages therefore say that there must always be a warning, because punishment is administered only for a halakhic offense. An act that is improper only in the moral or evaluative sense is not something for which a court may punish.

Here too we see the distinction between expressing the fact that the Holy One, blessed be He, does not approve of some act, and a verse warning us and determining that such an act constitutes a halakhic offense. Without a warning-verse, even if we knew that the act was unworthy, it still would not be prohibited in the halakhic sense. This is the gap between an indication of God’s will and a command that forbids it to us.

Even from the term itself, “warning,” it seems simple enough that the role of these verses is to warn and not merely to reveal God’s will. So too it seems from the rabbinic statement cited in this root: “Wherever it says ‘beware,’ ‘lest,’ or ‘do not,’ it indicates only a prohibition.” That is, what is required is a command and a warning; mere disclosure does not suffice.

Positive Commandments

The same is true of positive commandments. A worthy act is not yet a commandment. It may be positive or admirable, but from the formal halakhic point of view that does not make it a commandment. For some act to constitute a commandment, there must be in the Torah, or be derivable from the Torah, a verse that commands us concerning it. There is no commandment without command.

Moreover, fulfilling a commandment on the basis of rational conviction—that is, because the act is fitting and good—and not as submission to command, that is, because we were commanded, is not fulfillment of a commandment. Thus we find in Maimonides, at the end of chapter 8 of the Laws of Kings:

Anyone who accepts the seven commandments and is careful to perform them is one of the pious among the nations of the world and has a share in the world to come—provided that he accepts them and performs them because the Holy One, blessed be He, commanded them in the Torah and informed us through Moses our teacher that the descendants of Noah had previously been commanded concerning them. But if he does them because reason so dictates, he is neither a resident alien, nor one of the pious among the nations of the world, nor one of their wise men.

The thrust of Maimonides’ words here is that observance of commandments not from halakhic commitment but merely out of recognition of their value is an act that has no religious value, only moral value. One who observes the commandments because reason dictates it, and not because of submission to command, is not fulfilling a commandment at all, but only doing a good deed. Hence there is no commandment without command. This is precisely our point.²⁴

The Status of Reasoning

In several places we find that reasoning has a status comparable to biblical law. But if that is so, then we would seem to find here halakhic descriptivism—that is, a natural-law theory that identifies the moral-evaluative fact with the halakhic command. In other words, if something is fitting, then it is commanded. According to such an approach, there is no gap between the factual description and the law or command. On the other hand, if halakha has a positivist character, then even if reasoning tells us that an act is fitting, there is still no commandment without an actual command.

Indeed, there are views according to which reasoning—at least reasoning that creates new law rather than interpreting existing law—cannot constitute biblical law. One might, with some difficulty, distinguish between obligation and commandment, but in the plain sense a halakhic obligation too requires command.

Yet as we saw above, even in ordinary legal systems there are principles considered as though they had been enacted implicitly. Similarly, one may say that reasoning teaches us that the Holy One, blessed be He, certainly commanded this, and therefore it counts as though there were a command here. But this is only in cases where the reasoning really yields the conclusion that God Himself surely commands it. Reasoning that merely says an act is fitting still cannot turn that act into a biblical obligation. One must therefore distinguish among different kinds of reasoning.²⁵

Accordingly, the status of reasoning does not necessarily force us into a natural-law conception of halakha.

Natural Law in Rabbinic Enactments

With regard to rabbinic enactments, it seems even clearer that without legislation there is no halakhic significance at all. God’s will exists and stands whether or not it is expressed in the Torah. But the will of the Sages, by itself, has no halakhic significance. Only if the Sages institute a proper and authorized enactment does their will become part of binding halakha.

On the other hand, enactments usually have a reasonable underlying rationale. If so, the rationale generally exists even before the Sages establish the enactment or decree. According to a natural-law approach, if there is good reason to behave in a certain way, then we ought to be bound by it even without legislation. But an enactment, by its very essence, binds only because it was duly enacted.²⁶ Therefore, in the plane of rabbinic enactments, a natural-law approach would seem not to arise at all.

Against this background, it is interesting to present an example from the end of the chapter Hazkat HaBatim in Babylonian Talmud, Bava Batra 60b. The Gemara says:

It was taught: Rabbi Ishmael son of Elisha said, “From the day the Temple was destroyed, it would have been proper for us to decree upon ourselves not to eat meat and not to drink wine. But one does not decree a decree upon the community unless most of the community can abide by it. And from the day the wicked kingdom spread out over us, decreeing harsh and evil decrees upon us, taking Torah and commandments away from us, and not allowing us even to enter a week of celebration for a son—or, according to another version, the celebration of the son itself—it would have been proper for us to decree upon ourselves not to marry and not to beget children, so that the seed of Abraham our father would come to an end by itself. But leave Israel alone; it is better that they sin unwittingly than deliberately.”

The Gemara is discussing here two decrees that ought to have been enacted upon Israel. The first is not to eat meat and drink wine, and this was not enacted because one does not impose on the community a decree it cannot endure. The second is not to marry and have children after the sufferings of destruction and oppression. But here the factor preventing enactment is a different rationale, not the principle of a decree the community cannot endure, which was stated in the previous case. Instead, the rationale is “better that they be unwitting than deliberate.”

This rationale is highly puzzling. It implies that if they had enacted the decree, the community would have violated it deliberately, because it could not endure it, and therefore they preferred not to enact it. But if the decree was not enacted, then there is no prohibition at all, and one who marries and has children acts fully legitimately, indeed in accordance with halakha. How can such a person be called “unwitting”? At first glance, the rationale that should have been given here is exactly the same as in the previous case: it is a decree the community cannot endure.

On the face of it, this suggests that refraining from procreation is binding even without enactment. Even without the Sages formally decreeing it, a person who understood the matter should have refrained from procreation. The Sages merely did not publicize it, so as not to cause those who continued to procreate to sin deliberately rather than unwittingly. If so, even without a rabbinic enactment, by force of the rationale itself we ought not to procreate—even though there is a biblical commandment to do so. This looks like a conception of natural law even with respect to rabbinic decrees, perhaps in the form of “reasoning has biblical status,” such that it can even uproot a biblical obligation.

We see here a halakhic descriptivism. As noted, according to such a view the verse—or in our case, the decree—serves only to reveal the problem, not to warn us against it. The very existence of the problem is enough to obligate us.²⁷

But these things do not necessarily contradict our claim that there is no room for natural law in the world of rabbinic enactments. The public may have been obligated to act this way even without an enactment because of the force of the rationale, but halakhic validity would still not exist unless there were a proper formal enactment. We cannot expand on this further here.

Footnotes

True, a command is always phrased in future form, and that too is not accidental. A command addresses a person and tells him what to do after the command is issued—there is no point in commanding the past—and therefore it is natural that it should be framed in future language. Even so, this is not a reference to the future in the same sense as a future indicative statement of fact. In imperative or normative sentences, time is not part of the content of the sentence itself, but only a grammatical mode of expression.

Even so, one might distinguish and say that tefillin are also not really conditional. True, they apply only to men, only by day, only when one is clean, and so on. But all these are conditions not in our control and not subject to our will. This commandment is conditioned by objective circumstances, such as time, but we have no real way of avoiding it. That is unlike commandments such as slaughter or fringes.

See Rabbi Lichtenstein’s introduction to his lectures on Taharot, where he discusses whether there is a prohibition for an ordinary Israelite to become impure. Here Maimonides explicitly says that there is not, yet it is still clear that the state of impurity is an independent reality even for an Israelite, and not merely a description of the circumstances in which the other prohibitions apply. In any event, it is clear that according to Maimonides impurity is a reality, and not just a word describing a situation in which one may not enter the sanctuary, touch sacred food, or eat it.

Against this, David Henshke argues in one of his articles that according to Maimonides impurity is not an objective reality at all, but merely a cluster of laws.

This question may also be connected to the well-known Euthyphro dilemma from Plato’s dialogue Euthyphro: does God command an act because it is good, or is the act good because God commands it? From the view that the act is good because God commands it, it would seem that the revelation is not about an objective reality that precedes it, but that the revelation itself determines the reality. Even so, that does not necessarily mean there is no reality underlying the norm. For discussion of this matter, see our essay on the fifth root.

From the legal perspective, one may consult throughout Haim Gans’ book Tziyut VeSiruv, Hakibbutz Hameuchad, 1996, especially chapter 1.

From the perspective of moral philosophy, one may consult the anthology edited by Marcelo Dascal, HaTzodek VeHaBilti Tzodek, University Publishing Projects, 1977, especially part 4, and in particular Leo Rauch’s article there.

From the religious perspective, one may consult Jacob Joshua Ross’s article in the volume Bein Dat LeMוסר, edited by Daniel Statman and Avi Sagi, Bar-Ilan University, 1994, p. 43. See also the opening of Tamar Ross’s article in Daat 24, 1984, especially around note 6, where the moral question is linked with the religious question. Such a linkage was already made by Wittgenstein, though in my opinion not justifiably. There the discussion deals more with the psychological layer of the problem, namely the degree of connection between knowing what one ought to do and actually doing it. Many have noted, for example Anscombe, that psychological questions are relevant to this discussion.

Another interesting aspect may be found in Shlomo Biderman’s article in the anthology Philosophia Yisraelit, Papyrus, 1983, p. 93. That article deals with the “religious” interpretation of the Mimamsa school in Indian philosophy concerning the Vedas. It contains several illuminating parallels to various issues discussed in our essays on the roots, but this is not the place to elaborate.

One might, however, say that every indicative is in some sense a negation, and therefore these verses too fall within our discussion even according to Maimonides. The explanation is that, as we saw, a negation-verse always comes to negate some plausible initial assumption; otherwise it would be redundant and vulnerable to the objection that it is obvious. If so, the verses discussed in Sotah clearly come to negate some initial assumption, and therefore, even though they are phrased positively and without the word “not,” we should nevertheless understand them in principle as negating verses. According to this analysis, every indicative verse is a negating verse, even if positively formulated.

On this, see my article “Halakha and Reality,” submitted to Tzohar.

This is exactly a case of revelation without command. More than that, it is a revealing verse of the first kind, namely a factual disclosure of the actual spiritual consequences of the act, rather than a disclosure that the act stands against God’s will.

Rabbi Dov Lando brings another example of such a conception from Rabbenu Gershom on Temurah 3, where he raises the possibility that someone entrusted with an object for safekeeping that was lost might have to swear by command of the Torah and then be flogged for swearing—not for a false oath, since the oath is true. Here we have a case where a person is obligated by the Torah to perform an act, and yet he is punished for it. This is exactly like the case in Minhat Hinukh. It is important to note that here too this is only an initial possibility that is rejected in the conclusion of the sugya.

In this context one should also look at Sihot Musar – 5733 by Rabbi Hayyim Shmuelevitz, where he cites the Gemara about Rabbi Rehumi, who delayed coming home after being away for a year and thereby caused pain to his wife, and was punished by death—even though the punishment did not achieve its ostensible aim, since it only caused her more pain. Rabbi Hayyim says there that interpersonal sins are like fire: whoever enters them is burned, not because that is how justice reasons, but because that is how the laws of nature are built. This is an example of punishment as a natural reaction to the sinful act, and not as the result of the punisher’s discretionary judgment.

One might say that they did refrain. More plausibly, one should say that they did not, because if the Jewish people as a whole were going to continue anyway, there would be no reason for them not to contribute their part. Refraining from procreation has meaning only if it would in fact bring the existence of the Jewish people to an end.

Still, all this is quite puzzling: would the Sages really have thought it proper to bring about the end of the Jewish people? It may be that the discussion there concerns only a kind of pressure tactic directed toward God: if He does not redeem us, we will cease to exist. In that way, perhaps, they sought to force the fulfillment of the promise that the eternity of Israel does not lie. According to that, it becomes even clearer why there would be no point in the Sages themselves refraining from procreation when the majority of the public would not do so.

In the plain sense, however, this sugya seems to reflect the conception that the value of life is measured only in terms of the fulfillment of commandments. When there is no possibility of fulfilling commandments, life has no point. This recalls the rationale brought in Yoma 85 for desecrating the Sabbath in order to save life: desecrate one Sabbath for him so that he may observe many Sabbaths. That rationale too measures the value of life in terms of Sabbath observance. Perhaps this stands behind the dispute with the other rationale cited there, “and live by them—and not die by them.” See also M. Avraham’s article “Organ Donation in a State of Brain Death,” submitted to Tehumin, but we cannot expand on that here.