Lesson 9: Category 4 — The Fifth Root

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

---

With God’s help

The Fifth Root

It Is Not Proper to Count the Reason for a Commandment as a Commandment in Its Own Right

On the nature of the reasons for the commandments

Our present essay deals with Maimonides’ fifth root. This is the first root from the fourth category, out of the five we defined in the introduction: rules that concern the nature of a mitzvah (commandment) as a command. This category also includes the eighth and tenth roots, and apparently the fourteenth as well; therefore this will be the order of our discussion in the essays that follow. All these roots deal with different kinds of verses that seem to command something, but in fact do not. Thus this category is founded on defining the term “commandment” and distinguishing it from other kinds of statements. In the terminology of the essay on the sixth root, 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. As we shall see, these discussions really revolve around the distinction between imperative verses and declarative verses.

In this root Maimonides establishes the principle that the reason for a commandment should not be counted as a commandment in its own right. He refers to verses that, although they are “prohibition-like” in form, are in fact meant only to give a reason for a commandment. For example: “Her first husband, who sent her away, may not take her again… and you shall not bring sin upon the land” (Deuteronomy 24:4), or: “Do not profane your daughter by making her a prostitute, so that the land not become prostituted” (Leviticus 19:29), and so forth. These verses contain a command followed by a rationale. Regarding the latter half of the verse—the part that gives the rationale or reason—Maimonides claims that it is only a statement of reason, not a command. Therefore, despite its prohibition-like form, it contains no commandment, and it should not be included in the count of the commandments beyond the first half, which is the commandment and to which the second half gives the reason.

The question whether this is a substantive root or merely a technical one—that is, whether we have here a classificatory principle determining what deserves to be counted, so that the non-counting is not for a substantive reason but because the commandment is already included within another commandment, or whether this is a substantive principle determining the actual status of these verses—will also be discussed below. At first glance it seems to be a substantive principle, for anyone who disagrees with Maimonides claims that there are laws which, according to Maimonides, are not only uncounted but do not exist at all in halakha (Jewish law).

Here too Maimonides is contending with the view of Behag, who counted several such verses as commandments. The exact point of disagreement between them is not entirely clear, nor is Nahmanides’ position when he comes to defend Behag’s method.

In the course of our discussion we will also need to deal extensively with the broader issue of the reasons for the commandments and their status, which underlies the discussions of this root, and we will point to several surprising insights concerning them.

A. The Course of the Discussion

Maimonides’ Formulation

Maimonides’ words in this root are brief, so as usual we will cite them in full:

The fifth root: that it is not proper to count the reason for a commandment as a commandment in its own right.

Sometimes, in the reasons given for the commandments, there appears something resembling a prohibition, and one may think that it belongs among those to be counted separately. This is like the verse, “Her first husband who sent her away may not take her again… and you shall not bring sin upon the land” (Deuteronomy 24; negative commandment 392). The words “and you shall not bring sin upon the land” are the reason for the preceding prohibition, as if to say: if you do this, you will increase corruption in the land. Likewise, “Do not profane your daughter by making her a prostitute, and the land shall not become prostituted” (Leviticus 19; negative commandment 355). The words “and the land shall not become prostituted” are the reason, as if to say that the purpose of this prohibition is that the land should not become prostituted. Likewise, “Do not defile yourselves by them, and you shall be defiled by them” (end of Leviticus 11; negative commandment 179). After mentioning the prohibited species whose eating was forbidden, Scripture gave the reason and said: do not defile yourselves by eating them—as if it were merely telling us that doing what it warned against is a defilement of the soul.

Up to this point Maimonides presents several examples of reason-verses, which are the focus of this root. In every case the reason comes after the command itself, and it has the appearance of a prohibition. Maimonides rules that despite that appearance, these are reasons for commandments and not commands, and therefore they are not to be counted.

He then turns to two proofs from the Sifrei, the halakhic midrash (rabbinic legal exposition):

And by way of clarification they said in Sifrei, regarding His exalted statement after the earlier warning against taking ransom for a murderer, “and you shall not defile the land” (Numbers 35; negative commandment 295): Scripture teaches that bloodshed defiles the land. Thus it has been clarified that this prohibition is the reason for the preceding prohibition, not a separate matter.

Likewise they said, at the beginning of Parashat Emor (negative commandment 165), concerning the verse “He shall not go out of the sanctuary, and he shall not profane”: from this we infer that if he did go out, he has profaned.

It appears that Maimonides intends to prove that these parts of the verses are not commands but reasons. The midrash does not read them as two separate commands, but as a command followed by a reason—or by a consequence of the transgression. See below on the relation between those two categories.

Maimonides concludes by addressing Behag’s method, which counted these reasons as commandments:

Others too erred in this root and counted all these prohibitions without reflection. But whoever counted them will be embarrassed when asked: What exactly does this prohibition forbid? He will then have no answer at all. In this way the nullity of his count becomes clear. This is what we intended to explain in this root.

Maimonides’ main claim against Behag is that these are only command-like formulations; they command nothing, and therefore there is no basis for treating them as commandments. These are declarative verses, not imperative ones, for they reveal facts to us but do not command us to do anything.

Clarifying the Argument and the Different Ways of Understanding It

It seems that Maimonides’ argument here has two components, and this is also how Nahmanides understood him:

1. The interpretive component: these verses, despite their prohibition-like form, are statements of reason.
2. The normative component: reasons that the Torah gives for its commandments should not be included in the count of the commandments.

The proofs from the Sifrei are meant to establish the first, interpretive component—to show that these verses are indeed merely statements of reason. The explanation Maimonides gives at the end of his remarks is the basis for the second, normative component.

In principle, the normative claim can be understood in one of two ways:

1. These verses do not carry the force of command at all, and therefore they cannot be counted among the Torah’s commandments. This seems to be the straightforward sense of Maimonides’ language.
2. These verses are not essentially different from commands, but even if they are commands there is no reason to count them, because they add nothing new. They do not warn against anything beyond what the original commandment—the one they come to explain—already warns against.

Clearly, according to the second option, there would be no obstacle to counting only the reason and omitting the formal command. No duplication would result in the count of the commandments. Below, in the discussion of a king’s multiplying wives, we will see that there are tannaitic views that chose such a path. By contrast, according to the first option, we must count specifically the command and omit the reason.

Moreover, this discussion sends us back to the first, interpretive component of the argument, since that component suggests that the normative component must indeed be understood in the first way, as is also implied by Maimonides’ closing formulation.

Nahmanides’ Objections

Nahmanides opens the discussion by saying that at first glance Maimonides appears correct, but in truth that is not so:

Concerning this principle, I say what the Sages say in Babylonian Talmud, Ketubot 54a: at first glance it seems to support him, but when one examines it carefully, that is not so.

From what follows it is clear that he refers to the interpretive component of Maimonides’ claim. That is: at first glance Maimonides is right that these are reason-verses and not command-verses, but on closer examination he is clearly not right. Nahmanides now brings proof for his view:

For we have not seen our masters in the Talmud treat the Torah’s prohibitions as superfluous on the ground that they are merely the reason for a preceding prohibition, unless they come to add or subtract—that is, unless we learn some novel law from that reason. This is like their dispute in Babylonian Talmud, Sanhedrin 21a, over “He shall not multiply wives for himself, and his heart shall not turn away.” This second prohibition is certainly the reason for the earlier restriction on multiplying wives, and nevertheless the sages of Israel disputed about it…

Nahmanides’ point is that we do not find the Sages in the Talmud interpreting such verses as mere reasons, except where they come to narrow or broaden the command itself—in other words, where they introduce some legal novelty. His example is the sugya in Sanhedrin 21a, concerning the prohibition against a king’s multiplying wives. We will examine that sugya in greater detail below, since it contains an explicit mishnah and talmudic discussion about the very issues addressed by this root, and the entire dispute seems to begin there.

Nahmanides then argues against Maimonides on the basis of the well-known talmudic rule that every expression of “beware,” “lest,” “do not,” and “not” has the force of a prohibition. These reasons usually come in such language. Hence they are prohibitions, not reasons; that is, Maimonides’ interpretive premise is mistaken.

Here again we see the basic contrast between Maimonides, who tends toward rationalism—that is, imposing his a priori reasoning on the sources—and Nahmanides, who tends toward empiricism—that is, clinging to the sources and the facts that emerge from them. We have already noted this in the introduction and in several earlier essays.

Immediately afterward Nahmanides adds a further argument:

Moreover, on his view we should also say here: why may the first husband not take her back? Because of “and you shall not bring sin upon the land which the Lord your God gives you as an inheritance.” If so, the prohibition would not apply outside the Land. Just as they said there: why “he shall not multiply”? Because “his heart shall not turn away.” And why “he shall not take”? Because “he shall not profane”; and if he did not consummate the marriage, he does not receive lashes (Babylonian Talmud, Kiddushin 78a). Likewise: “You shall not pollute the land,” and “you shall not defile the land in which you dwell, in whose midst I dwell.” If so, we would permit all these things outside the Land, because the reason stated in the prohibition applies only in the Land. Rather, all of them are independent prohibitions.

His claim is that if these really are reasons, then we should derive legal conclusions from them—for example, that such prohibitions would not apply outside the Land of Israel. Here Nahmanides continues the line he suggested above, according to which the repetition of a reason is always meant to produce a legal novelty. Maimonides, of course, would answer that the reason is not meant to introduce anything on the legal plane, since we do not derive law from the reason of the verse.

Nahmanides now presses what he sees as an internal inconsistency in Maimonides, since Maimonides himself counts such reason-verses in his own list:

And the Rabbi himself, of blessed memory, counted prohibitions that contain reasons even more than these. But our rabbis treat them as prohibitions, as they said at the end of Babylonian Talmud, Yoma 72a: “Rachava said in the name of Rav Yehudah: one who tears the priestly garments incurs lashes, for it is said, ‘it shall not be torn.’ Rav Aha bar Yaakov challenged him: perhaps the Merciful One is only saying, ‘make it with a border so that it not tear.’ Is it written, ‘so that it not tear’?” Likewise they said there that one who dislodges the breastplate from the ephod incurs lashes, and one who removes the poles of the Ark incurs lashes. In all these cases they objected that perhaps they were only reasons for the preceding command, as if the Merciful One were saying: make them properly so that it not be displaced and so that they not be removed. And they answered: Is it written, ‘so that it not be displaced’ and ‘so that they not be removed’?”

He concludes from this:

From these we learn that whenever the Torah states a prohibition in addition to the command, and it is not an explicit reason for what came before, we treat that prohibition as independent—like “it shall not be torn” and “it shall not be displaced.” There are many such verses. For example: “You shall not uncover the nakedness of your daughter-in-law; she is your son’s wife; you shall not uncover her nakedness” (Leviticus 18). According to the Rabbi’s approach, the second clause would be only a reason for what came before: it is a great evil because she is your son’s wife, he owes you honor and you owe him love, and therefore it is not fitting that you uncover her nakedness. But according to the view of the sages in Babylonian Talmud, Sanhedrin 53b, the second prohibition is expounded to impose liability even after the death of the son; and “she is your son’s wife” means: I said this only when your son has a valid marital bond with her, excluding a raped woman, a slave woman, and a foreign woman (Sifra, Kedoshim, chapter 10, halakha 10).

He then continues with many more rabbinic derivations that extract legal details from such “reasons.”

Thus Nahmanides does in fact agree with Maimonides regarding the normative component of the claim—that reason-verses should not be counted among the commandments. But he disputes Maimonides’ interpretive component, arguing that in all these cases the second half of the verse should not be viewed as a reason but as a command: either an independent command or a command that introduces a detail into the command found in the first half of the verse. In any event, this second half still belongs to the normative, that is, halakhic sphere; it is not a mere declarative verse, as Maimonides sees it.

It therefore appears, at first glance, that the interpretive dispute is total. According to Nahmanides, within the Torah’s command-verses there are no parts that are genuine reason-statements, read as declarative rather than imperative. Verses of command contain only commandments or laws, and nothing else. Hence the very totality of the interpretive disagreement neutralizes the practical significance of their agreement on the normative component. Nahmanides does agree that if these were reason-verses there would be no place to count them—but in his view, the Torah simply does not contain such non-normative reason-verses.

Later in his remarks Nahmanides qualifies this totality. He points out that sometimes reason-verses occur without the language of prohibition—without “beware,” “lest,” or “do not”—for example: “Do not swear falsely by My name, thereby profaning the name of your God” (Leviticus 19:12). On this the Sifra comments: “This teaches that a false oath is a profanation of the Divine Name.” That is, the aim is only to teach us the gravity of the offense, not a legal detail—though in a certain sense even here there is an assertion in the normative sphere, especially given that profanation of the Divine Name is itself a legal transgression, not merely a reason.

From this, as Nahmanides writes, the objection to Behag also disappears:

It is indeed proper to ask the author of Halakhot Gedolot, as the Rabbi said, “What does this prohibition forbid?” But he will not be disgraced or embarrassed by that question. For they already expounded in Sifrei, Parashat Ki Tetze, the words “and you shall not bring sin upon the land” as a warning to the court concerning this matter. Thus they did not see it as a mere reason, as the Rabbi did, but as a prohibition and warning to the court. Therefore he counted it as an independent prohibition: we are warned to compel him to divorce her, and all the more so to prevent him from marrying her, just as they count the commandments concerning those liable to death or lashes, where we are warned to execute the law regarding them.

If every such clause teaches a legal detail, then the question of what it commands never arises. It too belongs to the normative sphere, and therefore Behag includes it in his count.

Nahmanides then continues with still more examples of such verses and the way the Sages treat them. In some places his readings are somewhat forced, and Maimonides’ approach seems more natural—but that is not our concern here.

In the course of his remarks Nahmanides also deals at length with the question of negations, which belongs to the eighth root, and we will address it there. The very fact that he links it to our present topic points to the connection between these two roots. In both cases Maimonides sees the verses in question as belonging not to the normative layer but to another layer, and Nahmanides disputes him on exactly this point. The entire fourth category of our essays on the roots deals with such disputes.

Remarks on Maimonides’ Position

It should be noted that Maimonides himself is aware of the distinction between reasons phrased in the language of prohibitions and other kinds of reasons, yet he chooses not to make anything of it. At the beginning of the root, when he describes the verses that he takes to be reasons and not commands, he writes:

Sometimes, in the reasons for the commandments, there appears something resembling prohibitions, and one may think that they belong among those to be counted separately.

By the phrase “resembling prohibitions,” Maimonides means that the reason contains the word “not,” and therefore looks like a prohibition, yet he still holds that it is not one. In his view there is no difference between reason-verses that come in prohibition-like language and reason-verses of other kinds. How do we know that a verse is a reason-verse rather than a command? Usually from the interpretive context. According to Maimonides, when a verse itself links the command to what appears in its second half, then the second half is read as the reason for the command, not as a separate command.

Qualifications to Maimonides’ View

Before proceeding, we should note two further points in Nahmanides’ remarks. Later he challenges Maimonides for counting, under negative commandment 76, the rule that when a priest who had immersed that day but must still wait until nightfall served in the Temple, he profaned the service—even though that rule is derived from a verse which, according to Maimonides, would seem to be a reason-verse: “They shall be holy to their God and shall not profane the name of their God.” It should be noted, however, that Maimonides does not count the fact that the service is invalidated—that is the reason—but rather the prohibition itself against such a priest serving in the Temple. That is clearly a prohibition, not a reason. True, it is derived from a verse that at first glance looks like a reason-verse; but according to Maimonides the Sages evidently had reasons not to read that particular verse as a mere reason. In this root Maimonides’ main point is that the invalidation of the service itself should not be counted, because it is the reason for the prohibition on such service, not an independent prohibition.

Thus Maimonides himself is prepared to read such verses as command-verses; he does not always read them as reason-verses. Below the disagreement between Maimonides and Nahmanides will become clearer, and the course of the discussion here will be easier to understand.

A Qualification to Nahmanides’ View

Toward the end of his remarks Nahmanides says that the prohibition against defiling the land, which follows the prohibition against taking ransom for the life of a murderer, is indeed a prohibition, because, according to his own earlier criterion, it contains the word “not.” Yet here Nahmanides admits that it has no practical legal consequence, and therefore in truth it should not be counted. Here we see that Nahmanides too can in principle accept the second interpretation mentioned above. Even verses that look like prohibitions and not reasons are sometimes not counted because they add nothing beyond already counted prohibitions, especially in relation to the command in the first half of the verse. In general, however, he argues that there are practical consequences, as he attempts to show throughout his remarks.

B. The Linguistic and Logical Nature of Reason-Verses

Introduction

We have seen that Maimonides does not count reason-verses because he does not regard them as commands. Nahmanides, by contrast, sees at least some of them as commandments in their own right, and others as commands that introduce details into other commandments.

In this chapter we will examine the relation between the present root and the eighth root, which also belongs to this same category. The comparison will sharpen Maimonides’ claim itself and bring us back to the distinction drawn in the previous essay, on the sixth root, between command-verses and declarative verses, or between norms—statements that belong to the normative sphere—and facts—statements that do not belong to that sphere.

The Relation to the Eighth Root

As noted, Maimonides’ method in this root should be considered in light of another root. In the eighth root Maimonides rules that negations should not be counted as prohibitions or warnings. Negations are verses that deny some property of an object or some state of affairs, but contain no element of command. For example, the statement “Abraham is not a priest” is not a command-verse but a negative factual statement. Such a verse merely speaks about reality; it commands nothing. We saw something similar in our essay on the sixth root, where we argued that the statement “It is not true that there is a commandment X” is not a normative statement but a factual one. A more detailed discussion of that root will appear in our next essay. For present purposes, however, we must ask whether the principle at stake in the present root—the fifth—is not in fact exactly the same principle as the one in the eighth root.

As we saw, there are reasons that “resemble prohibitions,” because they contain a negative word such as “not.” For example, “and you shall not bring sin upon the land”—which, according to Maimonides, is a factual statement and not a command. Moreover, the present root deals specifically with such reasons, since only in their case is there room for confusion and for treating them as commandments; that is exactly why Maimonides must say that they are not commandments but reasons. At first glance, then, reason-verses of this sort are a kind of negation.

If so, the fact that these reasons are not counted as independent commandments is simply a consequence of the principle that negations are not counted. It follows that this root, at least in one respect, duplicates the eighth root.

One could, of course, say that the novelty of the fifth root is precisely the claim that such reason-verses are indeed negations and not commands. Once that is understood, we then move on to the eighth root and learn that negations are not included in the count of the commandments. But if so, it still remains unclear why these verses were not brought as examples in the eighth root. Moreover, in negative commandment 165 Maimonides explicitly states that these are two different reasons. Let us briefly examine that discussion, since it sheds light both on the connection between reasons and negations and on the distinction between them.

Maimonides in Negative Commandment 165

In negative commandment 165 Maimonides discusses the prohibition on priests leaving the Temple during the service. The basis is the verse:

“He shall not go out of the sanctuary, nor shall he profane the sanctuary of his God, for the crown of the anointing oil of his God is upon him; I am the Lord” (Leviticus 21:12).

There is an interesting and unusual dispute there with Nahmanides, but it is too involved for our purposes. What matters here is a statement Maimonides makes in the course of that discussion—namely, that the High Priest does not invalidate his service even if he serves as an onen (a mourner before burial), and this for two reasons:

It has thus become clear to you that the words “nor shall he profane” were taken as a negation, not as a prohibition—meaning that his service is not rendered profane even though he is an onen. And according to the plain meaning of the verse, the words “nor shall he profane” are the reason for the preceding warning: he shall not go out [of the sanctuary], so that he not profane. According to both these understandings, this prohibition should not be counted independently, as is clear to one who has understood the earlier roots in this treatise.

Maimonides clearly states here that these are two different reasons. One relies on the rabbinic exposition of the verse and says that it is a negation—in other words, that the service of an onen High Priest is not invalidated. That is not a command or prohibition at all, but a factual statement. The other relies on the plain sense of the verse and says that it is a reason, not a command.

Maimonides’ treatment of these two reasons as distinct indicates that a reason is not merely a certain kind of negation, but a special case that would not be counted even if negations were counted. Once we have decided that reason-verses are not to be counted, it indeed follows that they are negations—but that obviously does not make the two reasons identical. As noted, even if negations were counted, there would still apparently be a separate reason not to count reasons as such.

From this it emerges that, in Maimonides’ view, the non-counting of reasons is not just a special case of the non-counting of negations, but has a distinct significance that depends on understanding the very concept of “reason.” We will try below to clarify that significance. But first we should sharpen Nahmanides’ position a bit further, and only then return to the dispute between them.

Clarifying Nahmanides’ Position

At first glance it may seem that Nahmanides claims these verses are not reasons for the commandments to which they ostensibly relate, but independent commands. Thus, for example, Professor Peintukh understood Nahmanides in his book Pekudei Yesharim on Sefer HaMitzvot, p. 77, in the course of his discussion of the eighth root:

Nahmanides’ approach is clear and unequivocal. According to our rabbis, the Torah does not present reasons for the commandments in the language of negation at all. The reasons for the commandments appear in a different form.

Such an understanding of Nahmanides is very difficult. First, it is extremely hard to wrench Scripture away from its plain sense. When the Torah commands the king not to multiply wives, horses, and wealth, for example, and then adds “and his heart shall not turn away,” it is very hard to believe that Nahmanides did not read the verse in its plain sense as giving the reason for these prohibitions. Second, the citation from Nahmanides on which Peintukh relies explicitly points in a different direction. Nahmanides writes there:

We have not seen our masters in the Talmud treat the Torah’s prohibitions as superfluous on the ground that they are the reason for the preceding prohibition, unless they come to add or subtract—that is, unless we learn some novel law from that reason… and certainly we do find in the Torah reasons that intensify the gravity of a transgression… but all those stated in the language of prohibition are prohibitions. And had it been only a reason, it would have said…

Nahmanides’ language explicitly says that the Sages do not treat such verses as mere reasons when they introduce no legal novelty. But if they do introduce such a novelty, then these verses most certainly are reasons. The explanation of this argument will be discussed below, but the plain meaning of his words is clear. These verses are reason-verses if they generate a legal novelty—even though they are phrased with the word “not.” Later in the same passage Nahmanides denies that these verses are only reasons, not that they are reasons at all. Perhaps that is what Peintukh meant.

Thus Nahmanides too agrees that these are indeed reasons for the commandments to which they relate. His difference from Maimonides is that such reasons are still included in the count of the commandments because they are command-verses. Nahmanides’ method is that these are command-verses which also function as reasons. By contrast, verses stated without the language of “not” are mere reasons and not commands.

This is also clear from Nahmanides’ language throughout the entire root, especially at the beginning of his remarks, where he discusses the prohibition on a king’s multiplying horses and wives and says, as part of his attack on Maimonides, that everyone agrees “and his heart shall not turn away” is the reason for the prohibition. The whole question there is only whether that reason should be counted. We will elaborate on this below, and the talmudic sugya itself will show that this is correct according to all views.

Summary

In this chapter we have seen that verses which state a reason but are not phrased as commands are, according to everyone, explanatory verses—as in the case of false oaths, where the verse explains the gravity of the punishment. The dispute between Maimonides and Nahmanides concerns only verses stated in the language of command, namely with “not.” Their dispute is not over whether these verses function as reasons for the commandments that precede them; as noted, both agree that they do. The dispute is whether such verses should be counted: whether, despite being reasons, they are still command-verses, as Nahmanides holds, or explanatory verses, as Maimonides holds. According to Maimonides, as we have seen, there is a principle that prevents us from counting reasons as commandments, and consequently he classifies such verses as explanatory. Nahmanides, by contrast, holds that even reason-verses can be command-verses.

We must now try to understand the disagreement between Maimonides and Nahmanides. That will also explain why, according to Nahmanides, it matters whether the reason written in the verse has practical legal consequences, in order for us to treat it as a reason that also functions as a command. We will do this through a discussion of the issue of expounding the reason of the verse, where these topics arise explicitly in the talmudic discussion in Bava Metzia.

C. Expounding the Reason of the Verse: Clarifying the Sources of the Dispute

Introduction

In order to understand the foundations and sources of the dispute between Maimonides and Nahmanides, we now turn to the first sugya Nahmanides cites in his glosses on this root: the sugya of multiplying wives, in the Mishnah on Sanhedrin 21a, which is also discussed in Bava Metzia 115a. As preparation, let us begin with the tannaitic dispute over whether one derives law from the reason of the verse.

The Basic Tannaitic Dispute

In our essay on the ninth root we noted that in Guide of the Perplexed, Part III, chapter 31, Maimonides flatly rejects the view that the commandments have no reasons, and he attributes it to weak-minded people. As we saw there, such a position would lead to the view that one cannot establish any hierarchy of greater or lesser severity among commandments, along with various other problems, and therefore it has no place within a halakhic outlook.¹

For this reason many biblical commentators and many works of Jewish thought have labored to identify the reasons for the various commandments. Yet even though we insist that every commandment has a reason and a purpose, the tannaim still disagreed over whether those reasons may be used in legal interpretation, as distinct from non-legal interpretation.²

In Babylonian Talmud, Bava Metzia 115a, the basic dispute between Rabbi Shimon and Rabbi Yehudah is brought as follows:

Mishnah. A widow, whether poor or rich, may not be taken as collateral, as it is said: “You shall not take a widow’s garment in pledge.”

Gemara. The rabbis taught: A widow, whether poor or rich, may not be taken as collateral—these are the words of Rabbi Yehudah. Rabbi Shimon says: A rich widow may be taken as collateral, but a poor widow may not, because you must return the collateral to her, and thereby you bring her into disrepute among her neighbors.

The Torah forbids taking a widow’s property as collateral. The mishnah explains that this law applies to every widow. As background, the Gemara there records a tannaitic dispute over which widow is meant: Rabbi Shimon holds that it applies only to a poor widow, not to a rich one. The Gemara explains that Rabbi Shimon takes the reason for the prohibition to be the disgrace caused by repeatedly returning the collateral, which appears improper to outside observers when the lender keeps entering the widow’s house. That reason exists only with a poor widow, because only to her must the collateral be returned each night. If the debtor has other usable vessels and does not need the collateral for immediate use, there is no nightly return requirement; the pledge is simply held until payment. Rabbi Yehudah, by contrast, holds that we do not derive law from the reason of the verse. As practical halakha, we follow Rabbi Yehudah, and likewise the anonymous mishnah here.³

Thus the basic dispute among the tannaim is whether or not one derives law from the reason of the verse, and the halakha follows Rabbi Yehudah, who says that one does not.⁴ Let us again stress: this is not a dispute over whether commandments have reasons, nor over whether discussing those reasons is important. The dispute concerns only whether legal interpretation may make use of them.

Rejecting the “Slippery Slope” Explanation

One might explain the dispute by suggesting that there is a fear that people will not observe the law once they reduce it to its reasons—a kind of slippery-slope concern, as we will later see with rabbinic decrees. This would seem to fit Solomon’s attitude toward multiplying wives: he thought he could multiply wives without his heart turning away.

This consideration will be discussed below. But even here it should already be clear that such an explanation is not relevant to the dispute over whether one derives law from the reason of the verse. Such a concern could explain why one should not publicize the reasons for the law broadly, or perhaps why one should not discuss them at all. But the dispute over the reason of the verse is not about whether one should study reasons for commandments—everyone agrees that commandments have reasons, and most views also hold that it is worthwhile to study them on non-legal planes. The dispute is only whether halakha itself should be shaped by them. Here the question is what the law actually is, not what people will do with it. It is implausible to say that Rabbi Yehudah interprets the prohibition on taking a widow’s collateral as applying to every widow merely out of a slippery-slope concern that people might otherwise become lax. Rabbi Yehudah holds that this is in fact the prohibition. This is an interpretive claim about the Torah’s intent, and therefore it has nothing to do with such practical concerns.

A First Explanation of the Dispute

The common explanation is that the dispute concerns whether we can trust ourselves to identify the correct reasons. According to Rabbi Yehudah there is always a danger of error, and therefore we refrain from using reasons in legal interpretation. Rabbi Shimon, by contrast, is more optimistic about our ability to reach the truth.⁵

The Problem with This Explanation

It is easy to see that this explanation is problematic on its face. If commandments do in fact have reasons, as we assumed above, then when we approach a verse like the one forbidding the taking of collateral from a widow, two options stand before us:

1. derive the reason and infer a legal conclusion regarding a rich widow;
2. do not derive the reason, and adhere to the text as written—whether or not one wants to call that “plain sense”—and infer that the law applies to a rich widow as well.

According to the explanation just given, the conclusion regarding the reason is speculative, so using it may lead us into error, and therefore it should not be used, at least on the legal plane. But is refraining from using the reason really safer? If the reason is in fact correct, then the legal conclusion is that there is no prohibition against taking collateral from a rich widow. If so, forbidding it is itself an error—and it may cost the lender, who may himself be poor and destitute, his money, since the collateral was meant to secure repayment. So both routes involve possible error. It is not at all clear that the route that refuses to derive the reason is the optimal one from the standpoint of minimizing error. Perhaps sometimes it is, but there is no evident reason to suppose that this is always the case.

Moreover, if we had to decide which of the two interpretive routes seems more plausible, it would obviously be the one supported by the more convincing reason. True, there is always some possibility of error, and it is never certain that we have arrived at the correct interpretation. But if the reasoning itself yields that particular reason, then that very fact makes that interpretive route appear more plausible.

In other words: the danger of error exists in both directions—using the reason and not using it. We must therefore return to the question: which interpretation is more plausible, that is, closer to the truth and to the Torah’s intent? Here it seems fairly clear that the interpretation tied to the plausible reason is the more plausible one.

Thus considerations of possible error could just as well lead us to conclude that we should derive law from the reason of the verse.

A Second Explanation: An Ideological Dispute

Some have suggested that this dispute reflects the broader temperaments of the two tannaim involved, Rabbi Shimon and Rabbi Yehudah. We know that Rabbi Shimon bar Yohai is associated with the esoteric dimension, and therefore he is more drawn to the hidden reasons of halakha; as we saw above in the first explanation, he also believes more strongly in our capacity to reach them, perhaps with the aid of kabbalistic tradition. Rabbi Yehudah, by contrast, is a man of the revealed surface, and therefore relates only to what lies before us.

People also often connect this dispute with other disputes between these two tannaim in the laws of the Sabbath. According to Rabbi Yehudah, the laws are determined by the act itself. Therefore even if someone did not intend the prohibited result, or did not desire the prohibited outcome, he is still liable. What matters for Rabbi Yehudah is the objective fact that the prohibited act was performed. The person’s intentions and desires are irrelevant. Rabbi Shimon, by contrast, exempts both the person who did not intend the result and the person whose act was “not needed for its own result,” because he follows the hidden, subjective layers of the person, not the overt, objective act. Thus for him what matters are the intentions and reasons behind the act, not the revealed act itself. So too in biblical interpretation: what determines the law is, in his view, the hidden layer and not the overt one; and the reverse is true for Rabbi Yehudah.⁶

There is also an aggadic link to this dispute,⁷ namely the sugya in Babylonian Talmud, Shabbat 33b, at the opening of the famous story of Rabbi Shimon bar Yohai and his son in the cave:

Why was he called “the chief speaker in every place”? Because Rabbi Yehudah, Rabbi Yosei, and Rabbi Shimon were sitting together, and Yehudah ben Gerim sat beside them. Rabbi Yehudah began and said: How fine are the deeds of this nation! They established marketplaces, established bridges, established bathhouses. Rabbi Yosei was silent. Rabbi Shimon bar Yohai answered and said: Everything they established, they established only for their own needs. They established marketplaces to place prostitutes in them, bathhouses to pamper themselves, and bridges to collect tolls. Yehudah ben Gerim went and reported their words, and they were heard by the government. They said: Yehudah, who praised, shall be exalted; Yosei, who remained silent, shall go into exile to Tzippori; Shimon, who denounced, shall be killed.

Here too Rabbi Shimon, the man of the hidden, sees deeds in light of the motives and intentions behind them. Rabbi Yehudah, by contrast, again looks at the revealed layer—the act itself—while ignoring the motives and intentions.

Let us stress that these connections presuppose a conception of the dispute different from the error-based explanation presented above. Here it seems that both tannaim may agree that we can identify the correct reason and are not chiefly worried about interpretive error. Even so, purposive considerations—what contemporary legal theory would call teleological interpretation—may sometimes lead to different results than interpretation that adheres to the text itself, often called textualism or, not always accurately, positivism.

According to this approach, the dispute is not about the fear of interpretive error but about a more ideological question: which plane is more important, or which plane is determinative—the reason, or the text?

It is important to notice that this assumes that even if we have identified the correct reason, purposive interpretation may still not yield the same results as textual interpretation. That creates a dilemma as to which of the two planes legal interpretation should follow, and on that question Rabbi Yehudah and Rabbi Shimon disagree. We will examine this more closely below.

The Problem with the Second Explanation

With regard to civil law, this is easy enough to understand. Statutes are drafted by human beings, and therefore the text is not always precise and does not always fully express the law’s original aims or the legislature’s intent. Moreover, sometimes the interpreter is not trying to recover the intent of the original legislator, but rather what he or she regards as the proper purpose of the law today. In such cases it is obvious that large differences can arise between purposive and textual interpretation, and the question is which of the two should govern.

But with regard to halakha the matter is much more problematic. The author of the text is God Himself. If our interpretation has indeed hit upon the correct reason, then that reason ought to fit all the details exactly. Therefore, if there is a difference between textual and purposive interpretation, that is a sign that one side or the other is in error. If the reason is correct, then its legal consequences are necessarily binding, and the text must not yet have been interpreted correctly. In the halakhic context we are trying to arrive at God’s intent as expressed in the Torah and as understood by us. If the reason is correct, then it reflects the intent of the Giver of the Torah, and we ought to use it to uncover that intent.

The conclusion is that in the Torah context it is hard to accept a position that does not assume some danger of error in identifying reasons, yet still allows that textual and purposive interpretation may diverge.

At this point we seem stuck. As we have seen, fear of error is not a satisfying explanation of the dispute. On the other hand, as we have just explained, neither is the question of which plane governs—the text or the reasons. So what does explain the dispute? The real difficulty is with Rabbi Yehudah, who holds that one does not derive law from the reason of the verse, for doing so would seem more logical in light of all the considerations raised so far.

When the Torah Itself States the Reason: The Case of a King’s Wives

The dispute we have been considering so far concerns commandments for which the Torah itself does not state a reason. In those cases the tannaim disagreed as to whether one may derive law from the reason of the verse. But the present root deals with a different kind of case: commandments for which the Torah itself writes the reason. Admittedly, that is not exactly the same as the question whether we derive law from reasons in general, but it is closely connected. What would Rabbi Yehudah and Rabbi Shimon say in such a case? Would they derive law from the reason here or not?

On this point the Gemara in Bava Metzia goes on to cite the Mishnah in Sanhedrin 21a:

“He shall not multiply wives for himself”—only up to eighteen wives. Rabbi Yehudah says: he may multiply wives, provided that they do not turn his heart away. Rabbi Shimon says: even one wife who turns his heart away he may not marry. If so, why does the verse say, “He shall not multiply wives for himself”? To forbid even wives like Abigail.

At first glance, the positions here seem to have reversed themselves completely. Rabbi Yehudah now derives law from the reason of the verse, and therefore he holds that a king may multiply wives if they are women who do not turn his heart away—in other words, righteous women. He derives the prohibition from its reason, namely that multiplying wives may turn the king’s heart away, and infers that if these wives would not do so, the prohibition does not apply. Rabbi Shimon, by contrast, now seems not to derive law from the reason of the verse, since he forbids multiplying wives even when they would not turn the heart away. True, he also forbids even one wife who would turn the heart away, even if the number remains below eighteen.

The Gemara in Bava Metzia then explains this by appealing to the fact that in the case of the king’s wives the reason is stated explicitly in the Torah:

In general Rabbi Yehudah does not derive law from the reason of the verse, but here it is different, because Scripture itself explains it: “He shall not multiply wives for himself, and his heart shall not turn away.” Why may he not multiply wives for himself? Because his heart might turn away.
And Rabbi Shimon says: since in general we do derive law from the reason of the verse, Scripture could have written only “He shall not multiply,” and we would already know the reason—why may he not multiply? Because his heart might turn away. Why, then, did Scripture write “and his heart shall not turn away”? To teach that even one wife, if she turns his heart away, he may not marry.

So Rabbi Yehudah, who generally does not derive law from the reason of the verse, does so here because the reason is stated explicitly. This implies that Rabbi Yehudah refuses to derive law from the reason only when the reason is not written; when it is written, he certainly does derive from it. Rabbi Shimon, who usually does derive law from reasons, sees the clause “and his heart shall not turn away” as superfluous, since he would have known the reason anyway even without the verse, just as in all the commandments whose reasons are not stated but are nevertheless derived. He therefore concludes that the second half of the verse is not a mere reason, but an additional command. Thus, according to Rabbi Shimon, the verse contains two commands:

1. not to multiply wives, even if they would not turn the king’s heart away. This is learned from the first half of the verse.
2. not to marry wives who would turn the king’s heart away, even if they are not numerous. This is learned from the second half, which he therefore treats not as a reason but as an independent command.

It is worth noting that this sugya is precisely what underlies the common view that the one who does not derive law from the reason of the verse does so out of fear of error. For we see that in a place where there is no such fear—where the Torah itself states the reason—Rabbi Yehudah agrees that the reason should be used.

Returning to the Legal Status of Reason-Verses

So far we have presented the sugya as it stands. Nahmanides, at the beginning of his glosses on the fifth root, cites this sugya as proof for his own position:

This is like the dispute over “He shall not multiply wives for himself, and his heart shall not turn away.” This second prohibition is certainly the reason for the earlier restriction on multiplying wives, and nevertheless the sages of Israel disputed about it. Rabbi Yehudah derives: why “he shall not multiply”? Because “his heart shall not turn away,” and thus teaches that he may multiply wives provided they are not of the sort that turn his heart away—that is, women proven to be fit and righteous. Rabbi Shimon expounds that prohibition independently. He says: even one wife who turns his heart away he may not marry. If so, why was “He shall not multiply” stated? To forbid even wives like Abigail. Thus he makes it an independent prohibition on account of its superfluity.

Nahmanides begins with the claim that it is clear that, at least according to Rabbi Yehudah, “and his heart shall not turn away” functions as the reason for the prohibition on multiplying wives, exactly as we explained above. He then explains that the tannaim dispute what to do with such a reason. According to both, although it is a reason, it still has normative force; they disagree only as to what that force is. According to Rabbi Yehudah, the reason-verse introduces a legal definition: one may multiply wives provided they do not turn the king’s heart away. Thus it is not a mere reason, but a verse that carries legal content and therefore belongs to the normative sphere. According to Rabbi Shimon, the second half is actually an additional command, on top of the one in the first half, since in his view the reason was already known even without Scripture stating it.

At first glance this would imply that according to all the tannaim—and certainly according to Rabbi Yehudah, whose view is accepted in practice because the anonymous mishnah regarding the widow follows him—a reason-verse is never a mere reason. Nahmanides’ conclusion is that when the Torah writes a reason, it is never a mere reason; it is either a command or a verse that introduces a legal detail into the command under discussion. This is exactly what we saw above in his method.

This appears to be a decisive proof against Maimonides, who treats such verses as reason-verses belonging outside the normative plane—they neither introduce a legal detail nor add an additional command. But a closer examination of Maimonides’ reading of this sugya casts new light on it, and through it, on our entire issue.

Maimonides’ Method: The First Tanna

In his Commentary on the Mishnah to Sanhedrin there, Maimonides writes:

The halakha is neither like Rabbi Shimon nor like Rabbi Yehudah.

This is very puzzling. How can Maimonides rule not in accordance with either of the two tannaim stated in the mishnah? Which view, then, does he adopt in practice? More fundamentally, what is the practical law in the case of multiplying wives?

He answers this in Laws of Kings 3:2:

He shall not multiply wives for himself. By oral tradition they learned that he may take up to eighteen wives; wives and concubines together may not exceed eighteen. If he adds even one more and has relations with her, he incurs lashes; he must divorce her and may marry another in place of the one he divorced.

Here Maimonides indeed presents a view different from both tannaim: there is a prohibition on multiplying wives, with no distinction based on the character or righteousness of the women. In other words, he really does treat the second half of the verse as a reason-verse and does not derive law from it. The law is determined by the textual interpretation of the first, legal half of the verse—“He shall not multiply wives for himself”—while ignoring the second half in legal interpretation, that is, treating it as a mere statement of reason.

What supports Maimonides here? In the dispute between Rabbi Yehudah and Rabbi Shimon over whether one derives law from the reason of the verse, the accepted halakha is like Rabbi Yehudah, and Maimonides himself rules that way. One would therefore expect him to rule like Rabbi Yehudah here as well.

But if we return to the mishnah in Sanhedrin, we find that Maimonides read it in a completely different way:

“He shall not multiply wives for himself”—only up to eighteen. Rabbi Yehudah says: he may multiply wives, provided they do not turn his heart away. Rabbi Shimon says: even one wife who turns his heart away he may not marry. If so, why does the verse say, “He shall not multiply wives for himself”? To forbid even wives like Abigail.

At first glance, the opening sentence looks like a heading, followed immediately by two tannaitic interpretations of the verse. Nahmanides certainly read the mishnah that way.⁸ Maimonides, however, reads the opening sentence as a first opinion, distinct from those of Rabbi Yehudah and Rabbi Shimon. According to him, there is here an anonymous first tanna who states that a king is forbidden to multiply wives, without distinction between righteous and wicked women. Rabbi Yehudah then offers his own interpretation, and Rabbi Shimon a third one. Following the standard rules of halakhic decision, Maimonides rules like the anonymous first tanna, the anonymous mishnah. That is how he arrives at the ruling cited above, both in his commentary and in his legal code.

If so, Nahmanides’ objections to Maimonides, which at first seemed so powerful, fall away on their own. There is indeed a tannaitic opinion that does not derive law from the reason of the verse even when the reason is written explicitly in the verse. It treats such language as explanatory and legally irrelevant, and Maimonides rules in accordance with it. Indeed, that is precisely the method Maimonides follows in the present root as well. He sees reason-verses as belonging not to the legal plane but to the plane of explanation. Thus, although Nahmanides saw this sugya as an objection to Maimonides, in fact Maimonides’ source for the present root is exactly that mishnah in Sanhedrin.

Analyzing the Tannaitic Views

We thus learn that the medieval authorities divide along the lines of an earlier tannaitic dispute. Nahmanides rules like Rabbi Yehudah and therefore concludes that there are no mere reason-verses; such clauses are always either commands or legal innovations. Maimonides, by contrast, rules like the anonymous first tanna—whose existence Nahmanides does not recognize, and therefore he reads the mishnah differently—and this first tanna holds that the second half of the verse is indeed a reason-verse. So we must understand the tannaitic dispute in order to understand the medieval one.

The difference between Rabbi Yehudah and Rabbi Shimon somewhat resembles the two understandings we suggested in the first chapter for the second, normative component of Maimonides’ argument. There we saw one possibility according to which reason-verses are not command-verses at all, and therefore cannot be included in the count of the commandments by their very nature. We also saw a second possibility, according to which they are indeed command-verses but have no practical legal consequence, and therefore there is no reason to count them in addition to the original prohibition. On that reading, the problem is not their non-normative character but their duplication of the first half of the verse.

One may therefore say that the difference between those formulations depends on whether these clauses are really reason-verses or command-verses. Only afterward does the further question arise why they are not counted—whether because they are not commands, or because they are duplicative. In this sense, the tannaim in the mishnah may be disputing precisely that point: Rabbi Yehudah and Rabbi Shimon think the second half is a command, while the anonymous first tanna thinks it is not. Let us now spell this out more carefully.

Further reflection on Rabbi Yehudah’s view shows that not only does he treat the second half of the verse as a command; he actually thinks that one should count specifically the second half as the commandment, and only it. He does regard the second half as normative, but he also sees it as overlapping with the first. The question then becomes: which of the two should be counted among the commandments? His answer is that one should count specifically the “reason” clause.

Why does he choose the second half? Because the first half is ambiguous and could lead us to false conclusions—in our case, that multiplying wives is always forbidden. It is therefore preferable to count the second half, which explicitly forbids only the kind of wives that turn the king’s heart away. But that still does not exhaust Rabbi Yehudah’s view. His deeper question is: what is the point of a prohibition on multiplying wives at all? According to Rabbi Yehudah, there is no such prohibition in halakha. The proof is that the king may multiply wives who do not turn his heart away. It follows that the actual prohibition is not against multiplying wives, but against turning the heart away. Multiplying wives is only a practical example of a situation that may lead to such turning away. Therefore Rabbi Yehudah thinks the right thing to count is specifically the prohibition on the turning away of the heart. That is the real legal prohibition.

The matter goes so far that one is led to ask why the Torah wrote the first half of the verse at all—if there is, in truth, no legal prohibition of multiplying wives as such. It seems that, according to Rabbi Yehudah, it serves only as an example meant to illustrate the abstract prohibition of turning the heart away. If so, Rabbi Yehudah adopts the second option above: these words are indeed a command.

Rabbi Shimon, by contrast, sees the plain meaning of the words as a reason. But in his view, if that were all, the second half would be superfluous. Therefore he concludes that it must in fact function as an additional command warning against the turning away of the heart. According to him there are thus two commands, and he counts the second in addition to the first.

The anonymous first tanna, however, sees the second half as the reason for the command and therefore assigns it no legal significance. He counts only the first half as the commandment.

We thus learn that the three-way tannaitic dispute covers all the interpretive possibilities. According to the anonymous first tanna, the prohibition is the command-verse, “He shall not multiply wives for himself,” and its continuation is merely a reason. According to Rabbi Yehudah, the “reason” is itself the prohibition, “and his heart shall not turn away,” while the first half is only an example. According to Rabbi Shimon, there are two prohibitions: “He shall not multiply wives for himself,” which forbids multiplying wives of any kind, and the “reason,” “and his heart shall not turn away,” which forbids any act that may lead to such turning away.

The Reason for the Prohibition According to Rabbi Shimon

As Nahmanides rightly emphasizes, Rabbi Shimon too does not dispute that “and his heart shall not turn away” is the reason for the prohibition on multiplying wives. Nahmanides writes:

This second prohibition is certainly the reason for the earlier restriction on multiplying wives, and nevertheless the sages of Israel disputed about it…

In Rabbi Shimon’s view, the clause is indeed a reason—but more than that, it also hints at an additional prohibition, because the Torah does not write reasons for nothing. Indeed, Rabbi Shimon’s whole problem with the verse arises from the fact that he understands the turning away of the heart to be the reason for the prohibition. Precisely because the reason is stated explicitly, it seems superfluous to him. If that were not the actual reason, there would be no superfluity.

Practical Halakha and Summary

As noted, Maimonides rules like the anonymous first tanna, while Nahmanides apparently does not recognize that such a first tanna is present and therefore rules like Rabbi Yehudah. How can Maimonides rule against Rabbi Yehudah, given that the halakha elsewhere on the issue of deriving law from the reason of the verse follows Rabbi Yehudah? As we saw, the dispute here is a consequence of the broader one.

To understand this, we must notice that the anonymous first tanna also does not derive law from the reason of the verse. On this matter he agrees with Rabbi Yehudah; the only difference is that he extends the rule even to cases where the reason is explicitly stated. Rabbi Yehudah, by contrast, limits the rule to cases where the reason is not explicitly written in the Torah. Thus, although in the case of the widow’s collateral the halakha has been decided that one does not derive law from the reason of the verse, that does not necessarily mean the halakha is specifically like Rabbi Yehudah. The anonymous first tanna also agrees with him there; in that sugya there is no practical difference between them. The halakha has indeed been decided that one does not derive law from the reason of the verse—but when the reason is explicitly written, there is still a dispute whether it should be used. It is here that Maimonides reveals that he is really following the anonymous first tanna rather than Rabbi Yehudah, though in the case of the widow their views coincide.

To summarize: according to the anonymous first tanna—and Maimonides, who rules like him—reason-verses are explanatory verses with no legal significance. According to Rabbi Yehudah—and Nahmanides, who follows him—the reason-verses define the actual prohibition, rather than the formal command-verses that appear to state it.

D. Purposive Interpretation in Rabbinic Prohibitions

Introduction

In the previous chapter we saw that there is a dispute whether Torah law should be interpreted purposively—so that the command or prohibition is really the reason, “and his heart shall not turn away,” rather than the formal command, “he shall not multiply wives for himself”—or textually, so that the command itself defines the legal scope. In practice, we adopt the textual approach, and therefore we view the prohibition as a prohibition on multiplying wives, not as a prohibition on the turning or diversion of the heart.

In this chapter we will examine a parallel phenomenon that exists with respect to rabbinic law. This discussion may in turn illuminate the issue of deriving law from the reason of the verse even on the biblical level.⁹

The Decree Against Reading by Lamplight on the Sabbath

It is well known that the Sages decreed that one should not read by lamplight on the Sabbath, lest one tilt the lamp. Thus the Mishnah in Babylonian Talmud, Shabbat 11a, states:

Mishnah. A tailor may not go out with his needle shortly before nightfall, lest he forget and go out with it; nor may a scribe go out with his pen. One may not inspect his garments, and one may not read by lamplight.

In a tannaitic source elsewhere, a baraita on Shabbat 12b, this enactment is stated together with its reason: “lest he tilt it.” That is, the concern is that if the lamp’s flame weakens, he will adjust the wick or flame.

The talmudic discussion there contains a striking exchange about this decree:

One may not read by lamplight. Rabbah said: even if the lamp is two stories high, or even as high as two plow-handles, or even ten houses one above another.

At first glance this suggests that the decree is being interpreted textually, for at a height of two stories there is obviously no concern that the reader will actually come to tilt the lamp.

On the other hand, we also find that when another person is present to watch and prevent tilting, reading is permitted:

The implication is that one person may not read, but two may. Yet was it not taught: not one and not two? Rabbi Elazar said: there is no contradiction; here it is one subject, there it is two subjects. Rav Huna said: with a bonfire, even ten people are forbidden.

At this point the Gemara brings the well-known story of Rabbi Ishmael:

Rava said: if he is an important person, it is permitted. An objection was raised: One may not read by lamplight lest he tilt it. Rabbi Ishmael ben Elisha said: I will read and I will not tilt. Once he read and sought to tilt it. He said: How great are the words of the Sages, who said one may not read by lamplight! Rabbi Natan says: He read and actually tilted, and wrote in his notebook: I, Ishmael ben Elisha, read by lamplight and tilted the lamp on the Sabbath; when the Temple is rebuilt, I will bring a fat sin-offering.

Already here a major difficulty arises. Even if Rabbi Ishmael thought he would not come to tilt the lamp, the prohibition remains in force. Is one permitted to violate rabbinic decrees merely because one personally is not concerned about the reason for them? Would anyone allow himself to eat chicken with milk—after the decree had been accepted, though before that there were certainly places that did not follow it, such as the region of Rabbi Yosei the Galilean—just because he is not worried that it will lead him to eat meat with milk? Such an approach would pull the rug out from beneath the entire institution of rabbinic decrees and enactments.

In fact, what we see here is a conception that does derive law from the reason of the decree. Rabbi Ishmael inferred a legal conclusion from the reason for the decree, and therefore allowed himself to read in a case where he thought the feared outcome would not occur.

What changed after the event? Rabbi Ishmael almost did come to tilt—or according to Rabbi Natan, he actually did. He then checked himself and recognized the greatness of the Sages who made the decree. But note carefully: even here Rabbi Ishmael does not retract his basic principle that if the concern is truly absent, the decree should not apply. What changed for him is only the factual concern that such self-assessment may be unreliable. Had he been certain that he would not tilt, it still seems that he would have maintained that there was no need to obey the decree.

The Permission for an Important Person

We should recall that the Gemara introduces Rabbi Ishmael’s story in connection with the question whether an important person may read by lamplight. From Rabbi Ishmael’s experience the Gemara initially proves that even for an important person it is forbidden. To this the Gemara immediately responds:

Rabbi Abba said: Rabbi Ishmael ben Elisha was different, because in matters of Torah he regarded himself as an ordinary person.

That is, Rabbi Ishmael treated himself as if he were not important, and therefore it was forbidden for him. The simple assumption of the Gemara, then, is that the decree was not enacted for an important person, because he can be trusted to guard himself. Whether this is a general principle about rabbinic decrees or a unique feature of this case is unclear.

The commentators there discuss how Rabbi Ishmael could have come to tilt at all, given the principle that the Holy One does not bring stumbling through the righteous—or even through their animals. Their answers differ. Tosafot explain:

Rabbi Natan says: he read and tilted. Here, and in Babylonian Talmud, Makkot 5b regarding Yehudah ben Tabai who executed a false witness, and in Yevamot 96b regarding the Torah scroll that was torn in their anger, the Gemara does not ask: “If the Holy One does not bring a mishap through the animals of the righteous, all the more so through the righteous themselves!” Rabbenu Tam says that this question is asked only regarding the eating of forbidden foods, because that is especially disgraceful for a righteous person.

The Ritva there cites Rabbi Yitzhak, who disagrees:

He read and tilted. Rabbi Yitzhak explained that the mishap came through him because he was effectively acting deliberately, in the sense that he set out to transgress the words of the Sages. Were it not for that, the Holy One does not bring stumbling through the righteous. Therefore there is no need here for Rabbenu Tam’s explanation.

At first glance, this is a dispute over whether what Rabbi Ishmael did was legitimate at all. According to Rabbi Yitzhak and the Ritva, his action was a transgression: it was forbidden for him to read by lamplight, and he nonetheless relied on himself when he should not have. According to Tosafot, there was no transgression here, because Rabbi Ishmael really was an important person. See also the Sefat Emet there.

Perhaps the disagreement hinges on whether this reflects a broad principle, namely that when the reason for a decree does not apply, the decree does not apply either. Tosafot seem to understand the sugya this way. Rabbi Yitzhak and the Ritva, by contrast, seem to hold that one is not permitted to violate a rabbinic decree simply because one personally is not concerned about the outcome. Only where the enactors themselves made an exception for an important person was there an exception, and Rabbi Ishmael did not consider himself to belong to that category.

Commentators have already noted that Maimonides omits the permission for an important person to read by lamplight. It is difficult to draw conclusions from this alone. One could explain it by saying that he simply does not rule in accordance with that permission; or perhaps he is skeptical about such assessments of reality, as Rabbi Ishmael ultimately became; or perhaps he thinks that in our times no one qualifies as an “important person” in this context.

The Legal Significance

What emerges from the sugya in Shabbat, and from Rabbi Ishmael’s conduct, is an approach that interprets rabbinic decrees purposively. According to that approach, someone who is not concerned about the feared result is not bound by the decree. It is not clear, however, whether this is a unique case or a general approach.

Now in the work Chazut Kashot, by Rabbi Issachar Dov Alterecht, at the end of tractate Beitzah—drawing on his own Avnei Choshen, Hoshen Mishpat 34:3, subparagraph 9—we find that he indeed understood the sugya this way and even drew a legal conclusion from it.

The Mishnah in Beitzah 36b rules that one may not hold court on the Sabbath. The Gemara explains that this is a decree lest one come to write. The Rif there, on 20b in the pagination of the Rif, cites the Jerusalem Talmud:

Jerusalem Talmud: In all these cases, whether they acted under compulsion, inadvertently, deliberately, or by mistake, what they did is valid on the Sabbath, and certainly on a Festival. From this we learn that one who transgressed and transferred ownership of land or movable property on the Sabbath, the transfer is valid.

That is, if judges did judge on the Sabbath or a Festival, then after the fact their judgment stands.

Rabbi Akiva Eger asks in his comments there: if they transgressed a rabbinic prohibition and judged on the Sabbath, then they should become disqualified as judges and witnesses under rabbinic law, and their ruling should not stand.

The author of Chazut Kashot answers that the sugya in Shabbat proves that Rabbi Ishmael himself held that even an ordinary person may read by lamplight if he is not concerned that he will come to tilt it. From this he derives a general principle: in any rabbinic decree there is no prohibition for one who is not concerned about the feared outcome, though one must be very cautious in making such factual assessments, if that is even possible.¹⁰ It follows that a court that held session on the Sabbath because it did not fear that it would come to write is not regarded as consisting of wrongdoers.

Thus the author of Chazut Kashot understands the sugya in Shabbat as teaching a general principle: whoever does not fear the result of the decree is not prohibited from acting against it. This is pure purposive interpretation. Yet such an approach would seem to contradict everything generally accepted about rabbinic enactments. Even in this very sugya we find that reading by lamplight is forbidden even when the lamp is two stories high, although it is obvious that in such a case one will not come to tilt it. How can that be reconciled with Rabbi Ishmael’s approach?

When the Reason Is Explicitly Included in the Formulation of the Decree

Several commentators have written that, in general, one must follow the wording of the decree; but where we find that the reason for the decree is written explicitly into its formulation, there is room for purposive interpretation. In the present case, the baraita itself says, “lest he tilt it.” If so, this decree is indeed open to purposive interpretation. It may be that only in such cases can one draw the broad conclusion proposed by the author of Chazut Kashot. In such cases the very enactors hint to us that the decree should be read purposively. When they want us to relate only to the wording of the enactment—that is, to interpret textually—they omit the reason for the decree.

This distinction closely parallels what we saw in the biblical context. There too, Rabbi Yehudah—and Nahmanides following him—do not derive law from the reason of the verse in general, that is, they interpret the Torah textually rather than purposively. But when the reason is explicitly written, they too agree that the law may be interpreted purposively.

As we recall, Maimonides—following the anonymous first tanna—disagrees with this even in the biblical context. In his view, one should not interpret purposively even where the reason is explicitly stated in the verse. It is interesting that with respect to the decree about reading by lamplight, Maimonides likewise omits the purposive reading—the permission for an important person—and records instead the rule that reading is forbidden even when the lamp is very high. Here too he does not derive law from the reason for the decree, even though the reason is explicitly stated in its formulation.

Summary and Conclusions

We have found approaches that strongly incline toward purposive interpretation of rabbinic decrees. According to these approaches, the decree is really its reason and not the formal command itself. Some apply this only where the reason appears in the actual wording of the enactment; in that case the purposive interpretation is also textual, since the text includes the reason. And from the flow of the Gemara in Shabbat it appears that even those who reject this approach do not reject it in principle, but only for technical reasons—above all, the fear that a person will misjudge the facts, as happened with Rabbi Ishmael.

A more radical position is attributed to the Vilna Gaon: rabbinic decrees have hidden reasons, and therefore even where the overt reason no longer applies there is no room to be lenient. This view rejects purposive interpretation of rabbinic decrees, but even here the rejection is for technical reasons: it is not clear that we have identified the true reason. As we saw, that is also the usual explanation of the view that does not derive law from the reason of the verse on the biblical level, though we have already pointed out its difficulties.

It is hard to find a conception that rejects purposive interpretation in a principled way, as if decrees bind in and of themselves. Without a reason, there is no basis for prohibition. It is only considerations of formal uniformity or technical difficulties that produce a formalist approach.

E. Why Do We Not Expound the Reason of the Verse?

Introduction

In chapter C we saw that in the tannaitic dispute over deriving law from the reason of the verse, the practical halakha is that one does not do so. In other words, we adhere to textual interpretation and do not adopt purposive interpretation.

However, even within the camp of tannaim who do not derive law from the reason of the verse—the anonymous first tanna and Rabbi Yehudah—there is an internal dispute, reflected also in the rulings of the medieval authorities, over whether one does derive law from the reason when it is written explicitly in the Torah. Here Maimonides and Nahmanides disagree: according to Maimonides, who rules like the anonymous first tanna, one does not derive law from the reason even then; according to Nahmanides, who follows Rabbi Yehudah, one does. Their disagreement in the present root, in its interpretive component, is a direct consequence of that earlier dispute.

We also saw that according to Rabbi Yehudah and Nahmanides, when the Torah explicitly states the reason, the contours of the prohibition are determined by that reason rather than by textual interpretation. In practice, the actual prohibition is the reason that appears in the second half of the verse, not the formal command in the first half. The prohibition is not multiplying wives, but turning the heart away. According to Maimonides, by contrast, that may indeed be the reason, but the prohibition is still determined textually, without the reason affecting the legal interpretation.

In this chapter we will examine how this dispute connects to the rationale for adhering to textual interpretation.

Two Reasons for Not Deriving Law from the Reason of the Verse

Above we saw two possible explanations of why one does not derive law from the reason of the verse:

1. fear of error;
2. Rabbi Yehudah’s ideological preference for the revealed, objective command before us over the hidden, subjective reason behind it.

In effect, Rabbi Yehudah’s approach is a form of positivism, rejecting purposive interpretation. It treats the reasons of the law as belonging not to the legal plane but to a different one, somewhat like Maimonides’ view that reason-verses do not belong to the normative sphere and therefore should not be treated as commandments.

We can now see that both of these explanations are problematic:

1. Fear of error. In Maimonides’ case this certainly will not work, because he refuses to derive law from the reason even where the reason is written explicitly in the Torah. In such cases there is no danger of interpretive error, since the reason is clearly stated. But even in Nahmanides’ case this explanation is very difficult, as we explained earlier, because refusing to use purposive interpretation may lead to even more serious legal errors—and may in fact be the more error-prone path.
   In the rabbinic sphere, by contrast, we did see approaches—indeed the plain sense of the sugya in Shabbat 12b points this way—that treat textual interpretation as the result of formal uniformity or fear of error. But these are considerations within rabbinic law. On the biblical level, where the discussion over the reason of the verse takes place, most views leave little room for such considerations.¹¹ There we are trying to determine the most correct interpretation of what the Torah says, and logic would seem to require using reasons—unless there is fear of error, which we have already rejected.

In places where refusing to derive the reason leads to stringency, however, there may indeed be room for an error-based consideration of the sort we saw in rabbinic law. That is probably how most commentators understand the principle that one does not derive law from the reason of the verse.

2. An ideological preference for the revealed over the hidden. This too is hard to accept. If there is truly no fear of error and no formal rule of non-differentiation, then we do not even find such a position in rabbinic law. It makes no sense to say that the Torah’s reason for forbidding a king to multiply wives is that his heart might turn away, yet to assume at the same time that the Torah forbids the act even where his heart would not turn away. Why would the Torah want that? If it truly prohibited the act even when that reason is absent, that itself would show that the real reason is different. But here the reason is explicitly written in the Torah, and so such a claim is especially hard to maintain. An ideology of “revealedness” might perhaps tip the scales where two interpretations are otherwise plausible, but it cannot itself create a plausible alternative against the logic of the text.

One might suggest here, somewhat like the approach attributed to the Vilna Gaon regarding rabbinic decrees, that perhaps the Torah has additional hidden reasons, and therefore we should not follow even a reason that is explicitly written.¹² But once again we must remember the argument raised above: refusing to use a written reason can itself lead to interpretive errors, and sometimes more serious ones. The fear of hidden reasons in a case where the Torah states a reason is no different from the ordinary fear of error in a case where the Torah does not state one. In both cases it is not plausible to let that fear neutralize our interpretation, since such neutralization may itself generate cruder errors.

Still, where the danger of error in one direction is much graver than in the other, this sort of reasoning may have some force. For example, if the danger of violating the prohibition against taking collateral from a widow were much more severe than the danger of needlessly preventing a lender from securing repayment from someone against whom such collateral would in fact be permitted, then there might be grounds to forbid collateral from all widows and not follow the reason. Something similar is true of rabbinic decrees, where one usually does not follow the reason of the decree because that typically leads to stringency, and therefore the opposite error is not as serious.

Is Fear of Error a Necessary Explanation of Nahmanides’ View?

Even so, Nahmanides’ position—that when the Torah itself states the reason we do derive law from it—seems at first to imply that the refusal to derive law from the reason of the verse is only due to fear of error. Therefore, when the reason appears explicitly in the Torah, that fear disappears and the reason is used. As noted, at least unless one adopts the Vilna Gaon’s approach, that explanation does not seem possible for Maimonides, since he does not derive law from reasons even when they are stated.

Indeed, this is the usual explanation given for the halakhic principle that one does not derive law from the reason of the verse; see throughout the entry “Ta’ama DeKra” in Encyclopedia Talmudit. But in light of the difficulties we have just raised, it is important to note that even Nahmanides need not be understood this way. He too may not regard the refusal to derive law from the reason of the verse as merely the product of fear of error. Another possibility is that the consideration operates on the interpretive plane: when the Torah itself writes the reason, that is a sign that it wants us to use it, that is, to make legal use of it. When the Torah does not explicitly write the reason, it thereby hints that it does not want us to use it, even if there is no great fear of error. Thus, even according to Nahmanides, it may be that we do not derive law from the reason of the verse because that is the Torah’s own directive, and not necessarily because of fear of error.

That explanation is possible, but it too requires clarification. Why indeed should the Torah not want us to use reasons if there is no fear of error? What sense is there in applying a prohibition where its reason does not exist? The matter thus requires explanation both for Rabbi Yehudah and Nahmanides and for the anonymous first tanna and Maimonides.

Two Problems in Maimonides’ View

Two points are clearer in Nahmanides’ system. First, there may be an explanation there for why one does not derive law from the reason of the verse—namely fear of error, even if that explanation is problematic. Second, on the interpretive plane it is clear why the Torah would choose to state the reasons for certain commandments.¹³ It is to introduce some legal novelty, or to teach the seriousness of the prohibition. According to our suggestion above, it may also be meant to hint to us that in such cases one should derive law from the reason of the verse—and that itself yields the legal innovation that Nahmanides finds in reason-verses.

For Maimonides, however, both problems remain, and in full force:

1. The interpretive problem: Why does the Torah write the reasons for the commandments specifically in these places? Maimonides cannot say that it is to tell us that here we should derive law from the reason, because in his view even here we do not.
2. The substantive problem: Why not follow the reasons? If, according to the anonymous first tanna, it is clear that the reason for the prohibition on multiplying wives is the turning away of the heart—since the Torah explicitly says so—then the real prohibition should be the turning away of the heart, not the multiplying of wives. Why then forbid even multiplying righteous wives when that would not bring about the feared result?

A Possible Explanation of Maimonides: Error in Application

We are thus driven to the conclusion that the instruction not to derive law from the reason of the verse rests on the assumption that there is no need to do so. The Torah is written in an optimal way, and therefore textual interpretation, one that adheres to the wording of Scripture, ought to lead us to the correct conclusions even without reference to purposes.

In other words: differences between a good textual interpretation and a good purposive interpretation are the result either of an imperfectly drafted text or of imperfect interpreters. The Torah assumes that neither problem exists here. Therefore textual interpretation should bring us exactly where the Torah wants us to arrive.

Where do problems arise regarding the reason of the verse? In those places where textual and purposive interpretation seem to lead to different conclusions, as in the case of the widow’s collateral and similar examples. In those places we should infer that the error lies in the purposive interpretation, because it is our product, not in the textual interpretation, because the text is divine. This means that the text itself was written on the assumption that it would be interpreted textually and not purposively.

Thus, when Scripture says, “You shall not take a widow’s garment in pledge,” without distinguishing between rich and poor, that is a sign that the law applies to every widow. What, then, of the reason—namely the bad reputation that repeated visits may cause? Here one must say either that this is not in fact the reason, or that we are not applying it correctly.

To sharpen the point, let us look again at a case where the reason is explicitly stated by the Torah itself: the prohibition on a king’s multiplying wives. Here there is no fear of error about identifying the reason; it is obvious that this is the correct reason. Yet a difference still arises between the result of textual interpretation and the result of purposive interpretation. How can this be? Here it is clear that what has gone wrong is not the identification of the reason but its application.

In this case it is easy to see what the anonymous first tanna would regard as the mistake. Rabbi Yehudah, who derives law from the reason and therefore permits the king to multiply wives like Abigail, assumes that the turning away of the heart results only from multiplying wicked wives. But the anonymous first tanna replies that the Torah tells us to adhere to the text itself, and the text does not distinguish among kinds of wives. Therefore, in halakha, it is forbidden to multiply wives of any kind, even wives like Abigail. And what of the reason? It is still correct, because Scripture itself states it—but Rabbi Yehudah is applying it incorrectly. Even many wives like Abigail may divert the king’s heart away from God and from his people. Excessive preoccupation with women, however righteous, can itself turn the heart away.¹⁴

Thus the reason is correct, and there is no fear of error in identifying it, yet one must adhere to the text and deviate from what one might have inferred from the reason. The reason is that a correct reason is being applied incorrectly: the reason of “turning the heart away” does not lead to the conclusion that many wives like Abigail are permitted.

For the same reason, the prohibition on taking collateral from a widow may indeed be grounded in the concern that one thereby brings her into disrepute among her neighbors, yet that concern may somehow apply even to a rich widow, since after all the lender still has contact with her even if he need not return the collateral nightly.¹⁵ There too, the refusal to derive law from the reason of the verse may reflect concern about a mistake in application rather than concern about the reason itself.

Returning to the Different Cast of Mind of the Tannaim

We noted above that some explain the dispute over whether one derives law from the reason of the verse by broader differences in the temperaments of Rabbi Yehudah and Rabbi Shimon. Rabbi Yehudah is ideologically inclined to focus on the revealed layer; the hidden layer interests him less. Rabbi Shimon, by contrast, is drawn specifically to the hidden layer. We stressed that this contrast is not rooted in ordinary fear of error but is essentially ideological. What is the basis of this difference?

Perhaps we can now revisit that point. Rabbi Yehudah holds that the wording of the text is perfect, and therefore there is no need to consult reasons. He thus represents a positivistic stance opposed to purposive interpretation. If the wording is God’s and the interpretation is ours, then interpretation is more vulnerable than the text itself.

From this he concludes that the two planes ought in principle to correspond exactly. Therefore there is no point in turning to the plane of reasons at all, since it is unnecessary. And even if a gap seems to arise between textual and purposive interpretation, it still does not pay to rely on purposes and reasons, for such a gap is itself a sign that we have not interpreted or at least not applied the reason properly. Even when there is a discrepancy, it is therefore unwise to ground halakhic interpretation in the reasons rather than in the text.

Rabbi Shimon, by contrast, holds that part of textual interpretation itself is based on reasons. The formulation of the text was crafted on that assumption. The drafter took into account the fact that we would not be prisoners of the text, but would also use reasons—purposive interpretation. Therefore, when a difference arises between textual and purposive interpretation, it is the text that must be deemed imprecise, because its author relied on our use of purposes as interpretive tools in order to reach the correct legal conclusion.

So at the heart of the dispute there really are two ideological approaches, and not simply differing degrees of fear of error. As we saw above, these two outlooks characterize Rabbi Shimon and Rabbi Yehudah more generally, and they surface here as well.

Maimonides in Sefer HaMitzvot on a King’s Multiplying Wives

In negative commandment 365, Maimonides discusses the prohibition against the king’s multiplying horses, wives, and wealth. He writes:

The 365th prohibition is that the king was warned not to multiply special wealth for himself, as it says: “And silver and gold he shall not greatly multiply for himself.” The limit is that it be no more than what he needs to spend on his chariotry and his personal servants. But to gather wealth for the benefit of all Israel is permitted.

God, exalted be He, already explained in the verse the reason for these three commandments—I mean, that he shall not multiply horses for himself, nor multiply wives for himself, nor greatly multiply silver and gold for himself. And because their reasons and causes were known, the deviation from religion that they were meant to prevent came about, as is already famous from the story of Solomon, peace be upon him, despite his great level of knowledge and wisdom and despite his being beloved of God. They said that there is in this a hint and an admonition to people: if they knew all the reasons for the commandments, deviation would come to them. For if this man, perfect in wisdom and lofty in rank, already imagined and supposed that this act would not necessarily lead to that transgression, what would happen with the masses, in all their weakness? They would reflect on them and say: Was this act forbidden, or that act commanded, only because of such-and-such? I will be careful to avoid the thing for which this command was given, and I will not pay attention to the command itself. Then the uprightness of religion would be destroyed. Therefore God concealed their reasons. Yet there is not one among them that lacks a reason, cause, and purpose. But most of those reasons were such that the common masses needed not to grasp or understand them. And they are all, as the prophet testified: “The precepts of the Lord are upright, rejoicing the heart.”

Maimonides writes explicitly that the reason for multiplying wives—like the reasons for multiplying horses and wealth—is the turning away of the heart, even though, as we saw, in practical halakha he does not distinguish among different kinds of wives.

Moreover, Maimonides’ wording here strongly recalls Rabbi Ishmael’s attitude: “I will read and not tilt.” The king too may claim that he is a great and important person and therefore he will not fail. Indeed, Solomon famously said, “I will multiply and my heart will not turn away.” To this the Torah responds: multiplying wives may turn away even the heart of a great man—and even if the wives are righteous.

At the end of this passage Maimonides broadens the point to the question of the reasons for the commandments in general, and there one could hear a concern about error or a formal policy of uniformity. But that is not the only possible reading of his words. It may equally well mean that “the uprightness of religion would be destroyed” if one turned to purposive interpretation, because the real prohibition is what emerges from the text rather than from purposive interpretation, as we suggested above. In other words, the problem is not that the reasons are wrong, but that they may be wrongly applied. Otherwise, why would one not use reasons that are certainly correct, as in the very case under discussion where the Torah itself states the reason?[^\16]

To conclude, let us recall what we said above: one cannot explain the principle that one does not derive law from the reason of the verse by saying that people may come to sin—a slippery-slope concern, as with Solomon or Rabbi Ishmael. Such a consideration may explain why one should not reveal the reasons to the general public. But the legal question whether conclusions should be drawn from those reasons concerns the content of the law itself, not what different people may do with it. Therefore that consideration is irrelevant to the dispute over deriving law from the reason of the verse.¹⁶

Solving the Interpretive Difficulty

This also allows us to solve the interpretive problem. We asked: if one does not derive law from the reason of the verse, then why does the Torah write the reason specifically in these cases?

Perhaps the answer is that in these cases there is an especially strong danger of misunderstanding the reasons. Even though we do not use them in legal interpretation, the Torah still wants us to know them. If so, these really are pure reason-verses, and they have no role in the normative sphere.

Returning to Nahmanides’ Method

Above we raised doubts about the possibility that Rabbi Yehudah—and Nahmanides, who rules like him—refuses to derive law from the reason of the verse solely out of fear of error, even though that is the standard explanation. It is possible that Nahmanides too agrees with Maimonides that the refusal to derive law from unstated reasons stems from the fact that it is usually unnecessary. In other words, even according to Nahmanides, where there is a discrepancy between what the reason suggests and what the text itself yields, that may indicate a mistake in applying the reason.

Why then, according to Nahmanides, does one derive law from the reason when the Torah states it? Because there the Torah itself instructs us to do so. There the Torah assumes that textual interpretation alone will not lead us to the correct conclusion, and therefore it adds the reason in order that we use it to shape the law in practice. For that reason Nahmanides maintains that reason-verses always come to introduce a legal novelty or to function as an independent command.

At this point Nahmanides departs from Maimonides’ path—or, in tannaitic terms, Rabbi Yehudah departs from the anonymous first tanna. He may agree with Maimonides’ explanation of why one does not derive law from unstated reasons, but interpretively he cannot accept verses framed as commandments that nevertheless have no function on the normative plane. Therefore he understands the writing of the reason in such cases as an operative instruction to apply purposive interpretation to them. From this he concludes that the appearance of such a verse almost always has legal consequences, as we saw above in his method.

In practice, according to Maimonides there is no fear of error in identifying the reason itself, but sometimes there may be error in applying it. According to Nahmanides, by contrast, there is no such concern once the reason is explicitly stated, and therefore legal conclusions may be drawn from reasons written in the Torah; that is why he treats them as command-verses rather than as pure reason-verses. When the reason is not written in the Torah, Nahmanides too refuses to derive law from it. There the explanation may perhaps be fear of error in identifying the reason itself, or perhaps the same explanation Maimonides gives—namely that there is no benefit in using it, because the result should in principle be the same.

Rabbi Abraham, Son of Maimonides

We have reached the conclusion that the principle that one does not derive law from the reason of the verse, at least according to Maimonides, is not based on fear of error about the reason itself, but on the assumption that textual and purposive interpretation ought to coincide.

In fact, something like this appears in a responsum of Rabbi Abraham ben HaRambam, in Maaseh Nissim, responsum 5, printed in the Frankel edition of Sefer HaMitzvot, p. 552, s.v. “And what seems right.” The discussion there concerns Maimonides’ eleventh root, where he lays down the principles by which we classify commandments and decide whether two elements count as one commandment or two—for example the blue and white strands in ritual fringes, or the head-tefillin and arm-tefillin, or the four species of Sukkot. Rabbi Daniel objected there: why are the blue and white threads counted as one commandment because the Torah states one common purpose for them—“that you may remember all My commandments”—whereas the two kinds of tefillin are not counted as one commandment, even though the Torah also states a common purpose for them: “that the Torah of the Lord may be in your mouth”?

Rabbi Abraham responds:

The thought that because he said that the blue and white threads are one commandment since they have one purpose, it therefore follows that the two commandments of tefillin are also one commandment because they have one purpose—this is an error. For on that reasoning every set of commandments that has one purpose would count as one commandment, and then the 613 commandments would shrink to only a few, since the purpose of all forbidden foods is “Sanctify yourselves and be holy.” And likewise all the 613 would become one or two commandments, for the purpose of them all is love and fear, as the Torah says: “If you do not observe to do… to fear the Lord…” and again, “If you diligently keep… to love the Lord…” This is nothing at all. The two commandments of tefillin are distinct even though they have one purpose, while the white and blue threads are joined even though they have one purpose. You can see that they explicitly said concerning tefillin that they are two commandments, while concerning the white and blue they said that together they are one commandment.

Rabbi Abraham clearly holds that although these commandments may share a purpose, shared purpose is not the right criterion for counting commandments. The determinant, in his view, is the legal structure of the act—how the commandment is performed—not its purpose. This is exactly parallel to what we are arguing here: the reason does not determine the legal contours of the commandment, but not necessarily because of fear of error.

He supports this by noting that if we classified commandments by their purpose, all the commandments would collapse into one or two, namely love and fear of God, since that is the purpose of them all. Notice that he does not deny this purpose, nor suggest that it is a mistaken way of seeing the commandments. On the contrary, he accepts that this is indeed their purpose, and nevertheless insists that they must be counted as distinct commandments.

For exactly the same reason, Maimonides argues here that one must follow the legal definition of the commandment and not its reason, and that is why one does not derive law from the reason of the verse. It is not because the reason may be wrong, but because the reason is not what defines the commandment. If there is a gap between the legal definition and the reason, that is an indication that the reason has been misapplied, as we saw in the examples above.

An Additional Mechanism

We have seen that according to Maimonides the refusal to derive law from the reason of the verse is not due to fear of error. One can suggest an additional mechanism that also leads to that conclusion, at least in some cases.¹⁷

It is well known that one may distinguish between commandments of action and commandments of result. Sometimes the Torah imposes a duty to perform an action, rather than directly commanding the state of affairs that the action is meant to produce. As we explained in the previous essay on the sixth root, this may happen for two different reasons:

1. In some cases, as we saw there, the Torah truly wants the action itself, not merely the resulting state. We gave examples such as the father’s obligation to circumcise his son and the commandment of tending the lamps.
2. But sometimes the Torah’s focus on the action rather than the result does not mean that the result is not what it really wants. It may indeed want the desired state of affairs, but that result is not fully in our hands; all that is in our control is the action and the effort directed toward achieving it. In such cases the Torah commands the action, even though what it truly desires is the result.

In other words: had the Torah commanded the result itself, refraining from the action would not count as a direct transgression but only as placing oneself in a situation where one can no longer achieve the result; that is not what the Torah wanted. It therefore defined the action itself as the commandment, even though its desire is the result.

This second possibility opens another path for understanding why one does not derive law from the reason of the verse. Let us briefly mention two examples. In our essay on Parashat Bereshit, from 2007, we used the commandment to be fruitful and multiply as an example. There the action and effort are in our hands, but the result is in the hands of Heaven. A person does not control whether children will actually be born, and if so whether male or female. In such cases the Torah tells us to make the effort, and that is the legal definition of the commandment, while the reason for the commandment is the desired result—the birth of children.

This is an example of a mechanism in which there is no fear of error about the reason for the commandment, yet the commandment is still legally defined as a commandment of action rather than of result. Here too we do not derive legal conclusions from the reason—the birth of children—because the halakhic definition of the commandment is one of action, not of result. So here we have a mechanism that explains why the Torah follows the legal structure rather than the reason of the commandment, not because there is fear that the reason is mistaken.

Another example concerns prohibitions of eating and of benefit. As is well known, the Torah forbids deriving benefit from certain things, such as leaven on Passover, while other things, such as pork, are forbidden only for eating. At first glance, prohibitions of eating are prohibitions of action and not of result, since the benefit is the result of the eating. Yet it is clear that even prohibitions of eating require some form of benefit in order to incur liability. Rabbi Yohanan and Reish Lakish disagreed over whether what is required is benefit in the throat or benefit in the intestines, but according to all views some benefit is required; the only question is which kind.

It is possible that the prohibited benefit is really digestive benefit—the result of the action—but because that is not directly in our control, the Torah forbids the act of eating, which is in our control, and that is what is termed “benefit in the throat.”

This too yields interesting legal consequences. They stem from the fact that the legal definition of the prohibition is benefit in the throat, even though its reason is the desire to prevent digestive benefit. See the sugya in Ketubot 30a, regarding someone who shoved food into another’s mouth with a reed, and the early authorities there. Thus here too the reason is known, yet we still do not derive legal conclusions from it, because the legal definition concerns the act rather than the result, and it is the legal definition that determines the relevant interpretive tools.

F. Semantics and Syntax: Soundness and Completeness

Introduction

As we saw above, at least according to Maimonides, the explanation for not deriving law from the reason of the verse is that there is no need to do so. This view assumes that when formulation is unconstrained—that is, when it is perfect and optimal—the result of textual interpretation will be identical to the result of purposive interpretation. Such a perfect situation exists in two cases:

1. a divine text—the Torah;
2. a mathematical text, which is absolutely precise in its formulation and does not compromise that precision for the sake of elegance, convenience, or anything of the sort.

In every other form of discourse there will be some gap between the text as such and the meanings it seeks to express, either because of unintentional imprecision or because of considerations of convenience and elegance. In everyday language, meaning and message matter more than exact wording, and therefore exact formulation is not always preserved.

For this reason, in legal interpretation of civil law—which deals with humanly drafted statutes in ordinary language—it is reasonable to employ purposive methods. There the wording is not mathematical, and one might even say that the subject matter itself does not allow such precision. True, some legal theorists still advocate restricting interpretation as far as possible to textual tools. But there the motive is fear of error, or fear that judges and interpreters will abuse interpretive methods to impose their personal views under the guise of interpretation. In the legal context this is always a second-best solution, because purely formal textual interpretation will never fully capture the legislator’s intent. The reason is that the legislator is a human being using language that is a social convention, not a mathematical-theoretical construction and certainly not a divine one. On the other hand, such an approach also places a heavy burden on the legislator: to come as close as possible to a maximally precise and optimal formulation, something very difficult in law, unlike mathematics, which is defined from the outset in those terms.

Thus, in a perfectly precise text, one assumes a perfect correspondence between the textual plane and the purposive plane—the plane of reasons. Leibniz, and after him several twentieth-century analytic philosophers, aspired to develop an everyday language that would be equally precise, a formal language, but they did not succeed. There is a major debate over whether such a project is even possible, but that is beyond our concern.

In mathematics, this correspondence between formulation—syntax, the form—and meaning—semantics, the content—is known in logic as the theorem of soundness and completeness. That theorem both follows from, and is the highest expression of, the optimal precision of mathematical language.

In light of what we said above, we may now say that in the world of biblical interpretation—at least according to the anonymous first tanna and Maimonides—this correspondence is expressed by the rule that one does not derive law from the reason of the verse. As noted, this approach assumes a perfect parallelism between the reasons and the wording. According to the anonymous first tanna, the rule both presupposes and expresses the precision of biblical language.

In this chapter we will examine the meaning of that correspondence and try to clarify it a bit further.

Parallel Planes of Explanation

The phenomenon we have been describing is the parallelism of planes of discourse. The same event or fact may be described on several different planes, each in its own terms, and in our case they may even lead to the same results. This phenomenon is discussed at length in the fourth gate of the book Enosh KaChatzir. Here we will mention only a few examples to illustrate the point.

The Four Classic Levels of Interpretation

It is customary to speak of four planes of biblical interpretation: straightforward meaning, allusion, homiletical interpretation, and esoteric interpretation. Each explains the text on a different plane, sometimes even within a completely different conceptual world, and yet they are all supposed to converge in some sense. For example, it is well known that the Ari used to rule in halakha like Maimonides, even though one could hardly imagine a decisor farther from the world of hidden, esoteric meaning. The reason is that he assumed that if Maimonides reaches a halakhic ruling through reliable reasoning on the revealed plane, then the resulting practical halakha will also be correct on the hidden plane. It is also known of certain decisors, such as the author of Divrei Chayim of Sanz and others, that they would rule only after examining the issue in both the revealed and hidden dimensions.

By contrast, the Magen Avraham, in the laws of tefillin in Orach Chayim 32, writes that whenever there is a contradiction in halakha between the revealed and hidden dimensions, one must follow the revealed. Mystics, of course, would not agree. The Vilna Gaon, however, is known for rejecting the very existence of such contradictions and even forcing the plain sense to conform to the esoteric one, as in the sugya of sleeping directions and the like.

The meaning of these approaches is that one can arrive at the same practical halakhic conclusions from very different kinds of reasoning, even from entirely different conceptual systems. There is a correspondence between the systems, and therefore they yield the same results.

Our present case is another instance of such correspondence: between the textual plane and the plane of meaning and purpose, though both still belong to the revealed dimension, as syntax and semantics.

Natural Law and Theology

The famous story is told of Newton sitting under a tree, when an apple fell on his head and he asked himself what caused it to fall. From there he inferred the existence of the law of gravity.

Newton, as is well known, was a deeply religious Christian, and presumably believed in divine providence over the world. One might therefore have thought he would be satisfied with a theological explanation: the apple fell on his head as punishment for some sin he had committed. Yet he was not satisfied with that, and also asked the scientific question. Thus he related to the falling apple on two planes: the theological, where the explanation is punishment for sin, and the physical, where the explanation lies in the weakening of the fruit’s attachment to the branch and the force of gravity. These two explanatory planes arrive at the same result, even though there appears to be no connection between them. They are different conceptual worlds, based on different principles, and yet they are both expected to fit the same world of phenomena.

Similarly, the Torah tells us that if we do the will of God He will give rain in its time and produce in the fields. We are also committed to a scientific worldview according to which rain is the product of climatic conditions and crop yield the product of soil conditions, climate, and so forth. How do these two planes fit together? Here too there is a scientific explanation and a theological explanation, and both are meant to arrive at the same result.

Psychology and Philosophy

Suppose someone becomes religious. His secular friends from the past explain this in psychological terms—a mental crisis, life events, or the like. His new religious friends explain it on the plane of reason and truth: he discovered what is true. When someone leaves religion, the pattern reverses: his former religious friends explain it in terms of crises and desires, while his new secular friends explain it in terms of reason and intellectual awakening.

So who is right, the philosophers or the psychologists? The answer is: both. Every step a person takes can be described both on the philosophical plane and on the psychological plane, except where an irresistible impulse overrides the philosophical dimension.

The Chinese Room

The Chinese room example was proposed by the British philosopher John Searle, in his book Minds, Brains and Science, to describe the relation between body and mind. Imagine an English speaker sitting in a closed room with two openings, one for input and one for output. Through the input opening come questions in Chinese, which he does not understand, and through the output opening he is supposed to produce answers in Chinese. For this purpose he is given barrels full of Chinese symbols from which he must assemble the responses.

Whenever he produces a wrong or irrelevant answer, he receives an electric shock. If we wait long enough—say, an infinite amount of time—he will begin to learn how to assemble the relevant answers for each type of question. In such a situation, a Chinese speaker could converse with him and think he was speaking with a fluent Chinese speaker, even though he is doing nothing more than manipulating Chinese symbols and understands none of their meaning.

This is really a version of the Turing test, which was proposed as a criterion for when a computer has become the equivalent of a normal human being. Turing suggested placing a person and a computer in separate closed rooms and having a third person converse with both. If the third person cannot distinguish which is the human and which the computer, then the computer has reached the level of a thinking human being.

The main problem with this test is that it fails to distinguish between semantics and syntax. Even if the human examiner cannot distinguish the computer from the person, this does not mean that the computer is thinking in any meaningful sense. The computer understands nothing and does not think, because thinking is engagement with meaning. A computer computes, but does not think. It deals in forms and manipulations of symbols, not in meaning. When a computer reaches an optimal level, it may fully imitate human discourse, yet a deep difference remains. The syntax may be identical, but the semantics will always distinguish computer from human—perhaps not outwardly, but inwardly.

It is worth noting that this is exactly the state of affairs we described above. Maimonides manipulates principles and concepts of the revealed plane and arrives at a halakhic answer. The Ari assumes that the same result would also emerge if one worked through the meanings of the hidden plane, because there is a correspondence between the planes. The same result can be reached by both routes.

The same holds of the relation between a text and the meanings embedded in it. One can interpret the text by textual methods and arrive at a certain halakhic answer. Maimonides’ assumption is that for a skilled interpreter and a precise language, the two routes will lead to the same conclusions. Therefore there is no need to derive law from the reason of the verse. And if a difference nevertheless appears, then the error lies in the method of meanings, because ideally textual interpretation depends less on the interpreter than on the author—namely God.

In the next chapter we will point out that this picture is somewhat naïve. Even textual interpretation is interpreter-dependent.

Semantics and Syntax in Mathematics and Logic

We saw above that a computer can imitate human activity without any understanding. This means that operations of the sort a computer performs—precise and mechanical mathematical operations—can be carried out on two levels, and the two are equivalent: they can be performed as formal manipulations without understanding, or in terms of meaning and understanding.

Without entering into detail, let us simply note that every logical relation can be defined in two ways: by its meaning and by its truth table. These are two equivalent modes, even though the conceptual worlds they involve are very different. The same is true of testing logical arguments: one may do so in both ways, and the results will always be equivalent because of soundness and completeness.

Returning to the Reason of the Verse

In light of everything said so far, we can now return to our subject. As we have seen, Maimonides holds that a textual formulation will lead us to the same conclusions as semantic analysis—the analysis of reasons. The revealed consideration parallels the hidden one, and therefore no difference ought to arise between them. The assumption is that interpretive work on the textual plane will bring us to the same conclusions as attention to reasons, just as the Ari assumed regarding the rulings of Maimonides and their fit with the hidden dimension.

G. And Yet, Purposive Interpretation

Introduction

In this chapter we must qualify the picture we have drawn so far. The talmudic decision in the tannaitic dispute might suggest a total disregard for the meanings and reasons of halakha. In halakhic interpretation, it might seem, we proceed mechanically and formally, ignoring reasons altogether. We ask “what?” and not “why?” But every serious student knows this is at best a naïve illusion. As we shall now see, reasons do in fact play an important role in interpretation, if only because they cannot be neutralized completely. There is no interpretation without recourse to reasons.

Legal Definition and Reason

In the yeshiva world it is common to distinguish between the legal definition of a rule and the reason for it. The legal definition consists of the principles governing the law’s application. The reason is the cause or purpose of the law.

At first glance there ought to be complete correspondence between the two planes: the legal definition should reflect the reason exactly. Every detail of the law should express the reason that gives rise to it. As we argued above, it is not plausible that a correct reason—where there is no fear of error—would fail to reflect the legal definition.

Textual interpretation, then, is a possible tool for reaching the legal definition while bypassing the reason. As we saw at the beginning of the previous chapter, if the text is optimal—that is, if it presents the legal definition with precision—then one can extract the precise legal definition from it without recourse to the reason. The need to turn to the reason, that is, to purposive interpretation, arises only when the wording of the text in question is not fully precise, and therefore does not perfectly express the will and aim behind it. In the biblical text, that should not happen.

This idea is reflected in the common Brisker approach, which insists that one should ask only “what?” and not “why?” There the principle is applied not to Scripture but to the Talmud, yet it still assumes a correspondence between the “what” and the “why.” It assumes one can ask “what?” without recourse to the question “why?” That assumption is even more problematic there, since the Talmud is a human text that passed through many stages of transmission, and therefore it is clear a priori that the assumption is not really valid.

What about Scripture? Very often, when we find non-textual interpretation in halakhic sources, the question is asked: how can they be deriving law from the reason of the verse here? The standard answer is: this is not the reason, but the legal definition. Yet that answer too assumes that one can speak about the legal definition of a law without recourse to its reason and still reach the correct result.

Usually there ought indeed to be a fit between the legal definition and the reason, and therefore the distinction has no practical consequence. But there are cases in which a gap arises, as in the case of taking collateral from a widow. There the legal definition is that one may not take collateral from a rich widow any more than from a poor one, while the reason is that one thereby brings her into disrepute among her neighbors—assuming Rabbi Yehudah agrees that this is the reason, as suggested above. So there is a gap between legal definition and reason. The same is true of the prohibition on multiplying wives. There too we saw that the legal definition is: do not exceed eighteen wives, of any kind whatever. Yet the reason is the turning away of the heart, which seems to point in a different direction.

As we concluded in earlier chapters, such discrepancies most likely result from a mistaken application of the reason, even if the reason itself is correct. Why does this mistaken application occur? Because the interpreter enters the picture. He cannot simply rely on the perfection of the text and assume full correspondence between the wording and its meanings, since he himself is involved in the interpretive process. Pure textual interpretation simply does not exist. Every interpretation always involves the interpreter’s reasoning and sense of purpose. One cannot reach the legal definition from the text in total purity.

The conclusion can be formulated in two ways: there is no legal definition without a reason, and the distinction between purposive and textual methods is not sharp. Therefore the Brisker approach too is, to that extent, a naïve illusion. One cannot ask “what?” without in some measure asking “why?” There are, of course, different degrees and forms of this, but contrary to the positivist illusion, there is no such thing as purely textual interpretation wholly detached from the interpreter.

We know countless examples of the use of reasons in biblical interpretation, including halakhic interpretation. How does that fit with the rule that one does not derive law from the reason of the verse? Before returning to that question directly, let us examine two examples.

First Example: The Exemption of Tooth and Foot in the Public Domain

At the beginning of tractate Bava Kamma, the mishnah presents the four primary categories of torts. The mishnah and gemara there analyze each category and its distinctive rules. Pit is exempt with respect to vessels and humans. Fire is exempt for hidden items. Horn pays only half damages until the animal is established as habitual. Tooth and Foot are exempt for damage done in the public domain.

The Rif, at the beginning of Bava Kamma, casually gives a reason for the exemption of Tooth and Foot in the public domain:

Tooth and Foot are exempt in the public domain because that is their ordinary way.

Other early authorities follow him and explain that the exemption exists because the damage occurs in the ordinary course of the animal’s movement. Since the owner has permission to walk his animal in the public domain, the injured party must protect his property from being trampled.

The Rosh, at the beginning of Bava Kamma, cites the Rif’s view and comments—see also Yam Shel Shlomo, Bava Kamma 1:3–4:

I am puzzled why he needed to explain the reason they are exempt because that is their ordinary way. The verse itself says, “and it consumed in another’s field,” from which we derive: and not in the public domain.

He asks why the Rif gives a reason when there is already a textual source in Scripture. Behind this question lies, apparently, the principle that one does not derive law from the reason of the verse. The reasons do not operate on the legal plane. The textual source is a substitute for the reason, and one can derive the legal definition from it alone. There is no need to appeal to reasons.

Immediately afterward, however, the Rosh writes:

Perhaps he came to explain the reason of the verse: why did the Torah exempt it in the public domain? Because it is its normal way to walk there, and it is impossible for the owners to follow after it constantly. But Horn is liable in the public domain even though it too goes there, because once the animal is known to gore, the owner should have guarded it. And so too the law is that this is a penalty so that he will guard his ox. A practical consequence of this reason is that if a long beam lay partly in the public domain and partly in private property, and the animal stepped on it in the public domain and thereby broke vessels in private property, since it is normal for it to walk and tread there, it is exempt. Rabbi Yitzhak bar Shmuel did not explain it this way later in connection with the statement “an ox proves it” regarding ownership.

He thus derives a legal consequence from the reason: a case in which the damaged property is located on private property, but the animal was acting in its ordinary way while still in the public domain.

How does that answer his original question? If one does not derive law from the reason of the verse, how can the Rif infer such a legal conclusion from the law of Tooth and Foot? The standard answer is that this concerns the legal definition of the law, not its reason. But that is really only a cover for the deeper issue. How did the Rif arrive at that legal definition? Clearly by reasoning that the exemption of Tooth and Foot in the public domain is based on the fact that this is their ordinary way of moving there. We thus have another example of derivation guided by reason—purposive interpretation.

Second Example: Defining the Primary Categories of Damage

In the sugya there at the beginning of Bava Kamma, the Gemara discusses the secondary forms of each primary tort category. Within that discussion it cites a baraita defining the derivatives of Horn:

What are the derivatives of Horn? Goring, biting, crouching, and kicking.

The Gemara then questions this list:

Biting—is that not a derivative of Tooth?
No. Tooth is where the animal derives pleasure from the damage; here it does not derive pleasure.
And crouching and kicking—are they not derivatives of Foot?
No. Foot is where the damage is common; these are not common.

The Gemara’s initial assumption was that biting is a derivative of Tooth because it is done with the teeth. It rejects that by saying that damage classified as Tooth is not everything done with the tooth, but only damage that involves the animal’s pleasure, as in eating. Therefore biting, which gives no such pleasure, is not a derivative of Tooth but of Horn, whose defining feature is intent to damage. The Gemara then asks why crouching and kicking are not derivatives of Foot, since they are done with the feet. It answers that Foot refers to ordinary, common damage, whereas crouching and kicking are not ordinary.

The Gemara’s initial assumption was thus that the primary categories should be interpreted on a purely textual level: Tooth means anything done with a tooth, Foot anything done with a foot, and so on. Its conclusion is that they must be interpreted conceptually, according to the defining rationale of each category as relevant to the laws of damage. Once again we see interpretation guided by reason and logic rather than by text alone, though admittedly this is not quite purposive interpretation in the strict sense, since the purpose of the law is not what is at issue.

The Difficulty

These two examples are only two out of hundreds or thousands. We see that interpretation is not carried out in a purely textual manner. If it were, we would reach absurd results. “If a man’s ox gores his fellow’s ox” would apply only to an ox that gores, not to a dog that bites, and we would have to exempt the dog’s owner from payment.

Such phenomena prove the point we made above: purely textual interpretation is impossible. The interpreter’s own reasoning is always involved. And yet it is still not clear how this fits with the rule that one does not derive law from the reason of the verse. It seems that the Sages are constantly doing just that.

A First Possibility: Different Levels of Reasons

The most natural explanation is that reasons divide into different levels. There are reasons that stand very close to textual interpretation, and these we do use. In fact, interpretation without them is impossible. As noted, there is no such thing as pure textual interpretation. Therefore, interpretation that incorporates reasons on this more immediate level is still counted as textual, that is, as not constituting the derivation of law from the reason of the verse in the technical sense. Perhaps these are what the yeshiva world tends to call the “legal definition” as opposed to the “reason.”

Where exactly is the line? That is far from clear. Even so, there does seem to be some line that one may not cross. Let us suggest two indicators that may help determine when a reason may be used and when not.¹⁸

A Clear and Agreed-Upon Reason

The Tosafot HaRosh on Bava Metzia 90a writes that where the reason is clear and simply reflects the plain sense of the text, one certainly may derive law from the reason of the verse. The explanation need not be that he thinks the reason is usually excluded only because of fear of error. Rather, such a reason lies so close to the plain meaning that using it still counts as a kind of textual interpretation.

When Textual Interpretation Leaves Several Options Open

Sometimes textual interpretation leaves us with several legitimate possibilities. In such a case there is certainly room to follow the reason of the commandment. Perhaps this can also explain the Rif’s view about the exemption of Tooth and Foot.

The Torah exempts the owner when the damage takes place in the public domain. But here one may ask: must the damaged property be in the public domain, or is it enough that the damaging animal be there? Usually both are true, since the damager and the damaged property are in the same location. But in exceptional cases, like the practical implication cited by the Rosh, the two interpretations diverge. In such a case the textual interpretation leaves two possible paths open, and therefore the Rif uses the reason of the law to decide between them.

H. What Is “the Reason of the Verse”? Two Kinds of Explanation

Introduction

We have seen that Maimonides and Nahmanides disagree over whether reason-verses are verses of a different type from command-verses. Maimonides holds that they are declarative and not imperative: they state facts rather than impose obligations. Therefore, according to Maimonides, they are not counted because they do not belong to the normative sphere. Nahmanides, by contrast, holds that these verses always introduce some legal novelty and therefore do belong to the normative sphere. According to Rabbi Yehudah, and following him Nahmanides, the real command is the reason-verse and not the formally imperative verse. For example, the king is not really forbidden to multiply wives; he is forbidden to turn his heart away. Multiplying wives is only a specific practical example of how that prohibition may be violated. That is what it means to derive law from the reason of the verse when the Torah itself writes the reason. According to Maimonides, one does not do so even then. In effect, this is a dispute between a positivist and a natural-law approach to halakha.

In this chapter we will try to understand that disagreement by discussing two different senses of the concept of “explanation” in science and the implications of those senses.

Reasons and Safeguards: Grounding the Unfamiliar in the Familiar

When one examines the sugyot that deal with the reason of the verse, one finds that the reasons are always safeguards.¹⁹ That is true even in places where the Torah itself states them. For example: “He shall not multiply wives for himself, and his heart shall not turn away.” The prohibition on taking collateral from a widow is explained as rooted in the concern that she will thereby be brought into disrepute among her neighbors, and so on.

When one tries to imagine what a reason for a commandment could generally look like, it is difficult to think of one that is not a safeguard. A reason is always a statement of what will happen if such an act is permitted, pointing to the problems that will result. Perhaps one might also imagine a reason that is not a safeguard but a result of the act itself; in that case, the reason treats the commandment as a means for bringing about, or preventing, some factual outcome.

If so, once we begin searching for a reason for a commandment, we have already decided that we see the commandment as a safeguard or means to something else that lies in the factual realm.²⁰ This gives rise to one explanatory mechanism in the domain of the reasons for the commandments: grounding the unfamiliar in the familiar. There is a familiar fact, and the norm is explained by saying that violating it may lead to a problematic fact, while obeying it may lead to a beneficial one.

Reason as Theoretical Generalization: Grounding the Familiar in the Unfamiliar

It seems, however, that there is another possible kind of reason for a commandment: explanation by means of a general principle. A specific act is forbidden because there is a more general prohibition against some broader class of acts.

We know such explanations from science as well. In physics, we explain the falling of a particular object to the earth by means of the general principle that all bodies with mass fall toward the earth. This is explanation of a particular case by means of a general law.

In the language of the philosopher of science Carl Hempel, this is the deductive-nomological model of explanation. According to that model, we seek an explanation for a given state or event in terms of a general law from which the particular case can be deduced.

One may ask, in both the scientific and halakhic contexts, in what sense this is really an explanation. We are explaining a particular case by means of a principle that is not necessarily clearer than the case itself, but only more general. If the explanatory mechanism in the previous subsection grounded the unfamiliar norm in the familiar fact, then the mechanism here grounds the familiar in the unfamiliar. So long as we know the particular fact but not the general law, and infer the general law from that fact, we are in effect grounding the familiar observation in the unfamiliar law.

Why should that count as explanation? It takes something known and grounds it in what was previously unknown. Do we now understand the event any better? At first glance, no. On the contrary, we have merely extended the scope of our ignorance to a whole set of cases.

The Meaning of This Mechanism: The Status of General Theories

In the book Et Asher Yeshno VeAsher Einenu, this question is discussed at length. Two questions are raised there:

1. How do we arrive at the general theory from the particular cases?
2. In what sense does it explain the particular cases?

Two principal approaches are presented there:

1. The analytic approach: the general explanation is no more than our own generalization for reasons of convenience. It contains no information about the world and does not even claim to say anything about the world outside us. According to this view there is no such thing in reality as a force of gravity. That is merely a convenient fiction we use to describe a collection of observed facts. The criterion for choosing a general theory is elegance and utility, nothing more.
2. The synthetic approach: we do have means of reaching the general law from the particular cases. In Husserlian phenomenology, the particular cases are treated as transparent reflections through which we see the general law. Therefore the general law is a claim that contains information about the world. For example, the law of gravity is not merely a law but also a force. That is, from the particular cases we infer that there is a force generated by the interaction of any two bodies with mass. For that very reason, the general law also explains the particular cases that are derived from it. If such a force exists, then it is clear why every body falls—namely, is drawn—toward the earth. So although we did not know the general law beforehand, once it becomes known it really does explain the particular phenomena we observed.

There are also explanations rooted in purpose rather than cause, and usually these are the sorts of explanations found in the reasons for the commandments. There too one may distinguish between grounding in the familiar and grounding in the unfamiliar. The general law may itself state the purpose of the event or command under explanation, and it may be known or unknown.

Two Ways of Understanding “Turning the Heart Away”

In light of these distinctions, the verse prohibiting the multiplication of wives because of the fear that the king’s heart may turn away can be understood in two ways:

1. Grounding in the familiar: it is obvious that one may not allow the king’s heart to turn away. The prohibition on multiplying wives is explained by the fact that it may lead him into such a condition. Here the fact explains the norm.
2. Grounding in the unfamiliar: from the prohibition on multiplying wives we infer a more general principle, namely that one may not bring about a state in which the king’s heart turns away. Here the general principle is itself normative, not factual.

According to the first approach, the turning away of the heart is a factual claim. From it one derives the norm that wives should not be multiplied. According to the second approach, the principle of turning away the heart is not a fact but itself a norm. It is a generalization from the particular norm about wives, and the particular norm is then deduced from that general norm.

According to the first approach, the prohibition is the command, not the reason—if only because the reason is a factual verse and not a normative one. According to the second approach, the prohibition is the “reason,” because the reason is no longer a fact but a general norm. The reason-verse itself is normative and therefore should be seen as the real command.

The same picture applies even where the reasons are not written explicitly in the verse. For example, Rabbi Shimon in the case of the widow’s collateral does derive law from the reason of the verse, because for him the reason is the command. True, it is not written explicitly in the Torah, but we understand the general principle from the command and therefore draw legal conclusions from it. Rabbi Yehudah disagrees. In his view, where the reason is not explicitly written, there is a danger that we have generalized incorrectly. But where the reason is written, there is no such danger, and therefore in that case he does derive law from it.

Maimonides and Nahmanides

According to Maimonides, the reason adds nothing to the command itself. It is fairly clear that he does not see the reason as a general norm of the second kind above. If he did, the halakha would change dramatically and broaden accordingly. For example, if the reason for the prohibition on multiplying wives were understood by Maimonides as a general norm rather than a factual outcome, then the king would be prohibited from entering any condition of a turned-away heart, and not only through multiplying wives.

It therefore seems that Maimonides sees the reason-verse as a factual rationale for the prohibition, while the prohibition itself remains the formal command. This is not because of fear of error in identifying the reason, for Maimonides does not derive law from the reason even where the Torah states it explicitly. The reason is that the reason does not belong to the normative layer, nor does it introduce anything into halakha beyond what emerges from the text. It merely explains, and no more. For that reason Maimonides need not fear that the reason is mistaken, since in his view it more or less overlaps the command and does not exceed it, as it would in the model of normative generalization.

Thus Maimonides understands the relation between reason and norm according to the first option above. Nahmanides, by contrast, understands it according to the second option, and that is why he insists that a reason-verse always has legal implications. He sees it as a norm, and therefore if it were merely duplicative there would be no point in stating it again.²¹

A Note on Nahmanides’ Objection

Above we noted that Nahmanides objects to Maimonides on the ground that Maimonides himself counts “and they shall not profane” in the case of a priest who had immersed that day but served before nightfall, even though this prohibition is learned from a verse that looks like a reason-verse and not a command-verse.

We can now understand that Maimonides does not count the fact that the service is profaned, but only the norm forbidding such a priest to serve. The profanation of the service is a fact that serves as the reason for the command, but it is not itself the prohibition.

Nahmanides, by contrast, treats “and they shall not profane” as the formulation of a broad normative principle. Instead of saying that one may not take ransom for the life of a murderer, the Torah says one may not defile the land. This is not a statement about the fact that the land is indeed defiled, but a general prohibition against a type of act that defiles the land. Nahmanides does not read it as a description of reality but as a strong normative declaration that such things ought not to be done.

Because Nahmanides does not see reason-verses as stating facts but as stating broad norms, he does not distinguish between these two planes and therefore asks why Maimonides counts them. But Maimonides remains perfectly consistent. For him, the fact is not counted, but the norm that is connected with it can certainly be counted.

The Rest of Their Disputes and Their Implications

As we saw, Nahmanides also rejects the idea that the Torah would ever state reasons that have no practical legal consequence. This follows from his view that the reason is just another, more general formulation of what is stated in the command itself. There is no point to a generalization that adds nothing. Would one accept as a scientific generalization the statement that a certain object fell to earth? That is the fact, not the generalization underlying it. A generalization has value only if it adds something beyond the particular facts that follow from it.

Maimonides, by contrast, holds that reason-verses describe problems in reality that result from violating the prohibitions in question. It is therefore entirely plausible that the Torah would state them in order to explain to us the consequences of our actions, even if they have no legal consequence. There is no repetition here, because these are factual verses while the commandments are prescriptive verses. According to Nahmanides, both kinds are prescriptive, and that is why they would be redundant.

The anonymous first tanna maintains that “and his heart shall not turn away” is indeed the reason for the prohibition on multiplying wives, yet we draw no practical conclusion from it because there is no practical difference between the reason and the prohibition as legally defined. Even so, the verse is not superfluous, because it teaches us further information about reality: multiplying wives—even righteous wives—does turn the heart away.

This difference also explains why, according to Maimonides, there is no reason to fear that the reason may be wrong. The legal definition is itself an indicator of the reason’s correctness. If there is a fit between the reason and the legal definition, that is a sign that the reason is correct. According to Nahmanides, however, the reason is a generalization of the specific prohibition into a broader principle. As with scientific generalizations, errors may creep into such generalization, since there are many ways to generalize from a given set of facts. Therefore one can worry that the reason is mistaken, and so one does not derive law from it when it is not explicit. For that very reason, where the Torah itself states the reason there is no such concern, and it may be used.

According to Maimonides, one does not derive law from the reason even where the Torah explicitly reveals it, because his assumption is that we know the reasons correctly in every case; the point is only that there is no difference between the reason and the legal definition, and therefore deriving law from the reason adds nothing, as explained above.

Footnotes

A memorial volume published in memory of a yeshiva student named Schlesinger, from Yeshivat Kol Torah, who was murdered in a terrorist attack in Jerusalem, contains an essay by the victim himself discussing whether, in rabbinic law, the enactment is the reason for the enactment or its legal definition. Unfortunately, we do not presently recall the name of that volume and it is not at hand. In the book Mevakshei Torah on the topics and laws of Festivals, there is also an essay by Rabbi Zalman Nechemiah Goldberg on a closely related issue: whether what determines the interpretation of rabbinic enactments is the wording of the enactment or its reasons and purposes. That is very close to our present subject.

In the present chapter we will not enter into all the examples. Instead, we will present two examples which, to the best of our recollection, are not discussed in those sources, and through them we will illustrate several principles that also bear on the question of the reason of the verse.