Lesson 49: Shoftim

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

---

With God’s help

Concepts

• Two types of scriptural decrees.
• Reasons for the commandments.
• Applications of the principle, “it extends no further than its novelty.”

Summary

This week we discuss the law of conspiring witnesses. It is customary to view this law as a “scriptural decree,” since we are dealing with a clash between two sets of witnesses, and yet we choose the second set, believe them, and determine that the first set were liars.

I argue here that this law cannot be understood as a “scriptural decree” in the simplistic sense, that is, as a law without any rationale. It cannot be that witnesses are executed for lying when, in reality, they may not have lied at all. A scriptural decree does not transform reality; it only instructs us to act in one way or another, perhaps even without a clear explanation.

My conclusion is that scriptural decrees are not necessarily reasonless, and certainly not the decree concerning conspiring witnesses. I will show that Maimonides holds that all Torah laws have reasons, and he flatly rejects positions that view them as royal decrees in the simplistic sense.

This, of course, raises the question: if there is indeed legal logic at the basis of the law, why do we need a biblical verse as its source? To answer this, I distinguish between two types of scriptural decree. Sometimes a scriptural decree expresses the fact that a law has no immediately visible rational basis. At other times, a scriptural decree expresses the fact that a law conflicts with other principles of Jewish law, even though taken on its own it is clear and reasonable. I show that these two types of scriptural decree have different implications for the rule, “it extends no further than its novelty.” In the first case, we will tend to restrict the scope of the novel law as much as possible. In the second case, there is no reason for such restriction.

On the basis of this distinction, I offer a consistent explanation of the dispute between the Tur and Maimonides, a dispute that later scholars understood, somewhat simplistically, as revolving around the question whether the law of conspiring witnesses has a rationale or not. On my reading, the debate concerns what exactly the scriptural decree innovated, and I discuss several halakhic (Jewish-legal) implications of this disagreement.

On Conspiring Witnesses¹

A Look at Scriptural Decrees

Introduction

In our Torah portion there is a passage dealing with the laws of false witnesses (Deuteronomy 19:15-21):

A single witness shall not rise against a man for any iniquity or for any sin, in any sin that he commits; by the mouth of two witnesses, or by the mouth of three witnesses, shall a matter be established. If an unjust witness rises against a man to testify perversely against him, then the two men who have the dispute shall stand before the Lord, before the priests and the judges who will be in those days. The judges shall inquire thoroughly? No Hebrew allowed -> The judges shall inquire thoroughly, and behold, if the witness is a false witness and has testified falsely against his brother, then you shall do to him as he plotted to do to his brother, and you shall remove the evil from your midst. Those who remain shall hear and fear, and they shall not continue to do again any such evil thing in your midst. Your eye shall not pity: life for life, eye for eye, tooth for tooth, hand for hand, foot for foot.

The context of this passage is unclear. In some of the verses it seems to speak about litigants—”the two men shall stand”—but by the end of the passage it is clear that the discussion concerns the witnesses, who are ultimately found to have spoken falsely. Rashi, following the Sages, explains it that way as well. He further adds that the “witness” under discussion is really a set of two witnesses, who are found to be conspiring witnesses. Their punishment is that we do to them what they plotted to do to the defendant. If they testified in a way that, had their testimony been accepted, would have required lashes, they themselves receive lashes. If its acceptance would have led to a death sentence, they receive that same mode of execution.

The section of conspiring witnesses is unique in several respects. Their law stands between ordinary monetary liability and a fine—there is a tannaitic dispute whether their liability to pay is compensatory payment or a fine; see Babylonian Talmud, Makkot 3a and parallels. Their testimony appears no less reasonable or admissible than that of the witnesses who refute them—and yet they are deemed false witnesses, and more.

In Babylonian Talmud, Sanhedrin 27a, there is a discussion whether the law of a conspiring witness is a novelty or not. That discussion leads us to several fundamental issues concerning scriptural decrees in general, and that will be our topic this week.

A. Is Hazamah a “Scriptural Decree”?

Hakhashah and Hazamah

In Jewish law there are two kinds of collision between sets of witnesses:

1. Hakhashah—when the two sets give opposite testimony regarding some event.
2. Hazamah—when one set testifies about an event, and another set testifies that the first set was not even present at the scene.

In a case of hakhashah, it is impossible to determine clearly who is lying, and therefore at most we can disqualify both sets from that testimony. This is the dispute between Rav Huna and Rav Hisda in Babylonian Talmud, Bava Batra 31b and parallels. In any event, there is no possibility of punishing one of the sets, and certainly not both, as liars. According to Rav Huna, we even leave both sets under their prior presumption of fitness, and allow each set, separately, to testify in future cases.

In the verses above, the Torah speaks about a situation in which false witnesses are found in court. The background difficulty is that, on its face, a set of two witnesses can never conclusively be found false. If another set appears against them, we simply have two against two. Even if the second set numbers one hundred witnesses, Jewish law tells us that their force is equal, because “two are like one hundred.” If so, how can a set of two witnesses ever be found to be false? This problem leads us to the definition of hazamah.

The Mishnah in Babylonian Talmud, Makkot 5a, addresses this question:

Witnesses do not become conspiring witnesses unless they refute them with respect to themselves. How so? If they said, “We testify about a certain man that he killed a person,” and others said to them, “How can you testify? For that victim—or that alleged killer—was with us on that day in another place,” these are not conspiring witnesses. But if they said to them, “How can you testify? You yourselves were with us on that day in another place,” then these are conspiring witnesses, and they are executed on that basis.

The Mishnah distinguishes between hazamah and hakhashah. When there is a clash regarding the content of the testimony, this is hakhashah. But when the second set testifies about the first set and says that it could not possibly have seen the events, that is not ordinary two-against-two contradiction, but hazamah. In such a case, the second set is believed and the first is established as false.

This is the answer to the question how a set of witnesses can ever be found false. In an ordinary conflict—that is, hakhashah—we do not conclusively disqualify either set. But the mechanism of hazamah is what makes it possible to identify one set as false. According to Jewish law, in such a case there is a clear preference for the refuting set, and the refuted set is established as false.

The Source for the Law of Hazamah

The Gemara there derives this distinction between the two kinds of collision from the verses quoted above:

Gemara: From where are these matters derived? Rav Adda said: Because the verse says, “And behold, the witness is a false witness; he has testified falsely”—only when the falsehood lies in the witnesses themselves. The school of Rabbi Ishmael taught: “to testify perversely against him”—only when the perversity lies in the witnesses themselves.

The derivation is based on the wording “the witness is a false witness,” which indicates that the lie must be in the person of the witness, not merely in the content of the testimony.² From here emerges the definition of hazamah given above.

Is the Law of Conspiring Witnesses a “Scriptural Decree”?

In Babylonian Talmud, Sanhedrin 27a, it is explained that beyond the punishment of “as he plotted,” conspiring witnesses are also disqualified from future testimony as wicked persons. There, Abaye and Rava dispute whether conspiring witnesses are disqualified from the time they gave testimony or only from the time of hazamah. According to Abaye, they are disqualified from the time they testified, because it has now become clear that they were already wicked then. According to Rava, they are disqualified only from the time of hazamah. The Gemara gives two explanations of Rava’s view, which is the more novel one:

1. Conspiring witnesses are a novelty, and the novelty goes no earlier than the moment it was introduced.
2. Because of loss to purchasers, who would not know of their disqualification and would use them as witnesses.

In this law, the ruling follows Abaye, and therefore conspiring witnesses are disqualified retroactively, from the time they testified.

At first glance, the dispute between Abaye and Rava is about whether the disqualification of conspiring witnesses is a novelty or not. According to Abaye, it seems not to be a novelty but a straightforward rule. According to Rava, however, the matter appears to depend on the two formulations just cited in the Gemara’s explanation of his view.

What is the novelty in the law of conspiring witnesses? Clearly the issue is not the special punishment imposed on them. That is a novelty according to everyone, but it is not the subject of the passage. This passage deals with their disqualification from testimony, not with the punishment imposed on them. The reason given there for the novelty is:

A conspiring witness is a novelty: why do you see fit to rely on these? Rely on those!

In other words, the novelty is that hazamah is not treated like ordinary two-against-two contradiction, but rather the later witnesses are deemed more credible than the first.

Rashi there links the concept of “novelty” with that of “scriptural decree”:

It is a novelty—that two are disqualified because two others say, “You were with us.” For why do you see fit to rely on these? Rely on those! Rather, it is a scriptural decree. Therefore, it goes no earlier than the time of its novelty, namely, from the time they were refuted.

All this is Rava’s view, in the first formulation. In practical law we rule like Abaye, and on his view it would seem, at first glance, that there is no novelty here at all.

Among the medieval authorities, however, we find in several places a rational explanation for the law of conspiring witnesses and for why hazamah differs from hakhashah. The first to suggest such a direction was Nissim Gaon, who wrote in his Sefer ha-Mafte’ah as follows; see Teshuvot ha-Geonim, ed. Asaf, no. 702, pp. 195-196:

The difference between hakhashah and hazamah is that in hakhashah the contradiction between the two sets of witnesses falls upon the substance of the testimony… and the rule there is to invalidate both testimonies, since we do not know which is true. But in hazamah… the rule is to execute… and the denial by Reuven and Shimon is of no help at all, because the testimony is directed against them and they have become like litigants. When testimony is given against them, their own contradiction is of no value against the witnesses testifying against them. In contradiction between witnesses, however, the conflict falls only on the substance of the testimony; there they remain in the category of witnesses, and therefore both testimonies are invalidated.

Nahmanides, too, in his commentary to the verses above, writes the following (Deuteronomy 19:18):

“The judges shall inquire thoroughly, and behold, the witness is a false witness”—Scripture does not explain how one knows he is a false witness. For when the matter rests on two witnesses testifying about an event, even if one hundred come and contradict them, it is not thereby established that they testified falsely. Nor can we say that the murdered man came walking in on his own feet, for in that case the verse would not say, “The judges shall inquire thoroughly.” Therefore the reliable tradition came and explained that hazamah is when they say, “But on such-and-such a day you were with us” (Babylonian Talmud, Makkot 5a). The reason is that this testimony concerns the witnesses themselves, and they are not believed about themselves to say, “We did not do this,” for the others could equally say about them that they committed murder or desecrated the Sabbath.

Nahmanides explains the difference between hakhashah and hazamah by saying that in hazamah the second set testifies about the persons of the first set, whereas the first set is then effectively testifying about itself. Such a case is like one in which the second set disqualifies the first as robbers; in that case, according to all opinions, the second set is believed, as is explicit in Babylonian Talmud, Sanhedrin 27a. There it is explained that in such a case, even Rava agrees that there is no novelty, because the first set is not believed about itself, since it is a party to the case.

The same appears in the Tur, Hoshen Mishpat 38:

What is the difference between contradiction and hazamah? Contradiction is not in the bodies of the witnesses, but merely that they contradict them—for example, these say, “So-and-so borrowed from so-and-so,” and those say, “We know that he did not borrow, for we were with him the entire day and saw that he did not borrow.” Hazamah, however, concerns the witnesses themselves, for the second set says, “At the time when you say he borrowed, you were with us.” For this reason the latter are believed, since they testify about the bodies of the witnesses; it is as though they testified about them that they committed murder or desecrated the Sabbath, and the first witnesses are not believed about themselves to say, “We did not do such-and-such.” Even if the first set numbered one hundred, the latter are believed against them to have them executed—whether the one hundred testified all at once or two after two, and the refuting witnesses refuted each set one after another.

Sefer ha-Chinukh, commandment 524, also cites this reason in the name of one of the sages as “some measure of logic in the matter.”

In contrast to all these, Maimonides writes as follows in Mishneh Torah, Laws of Testimony 18:3:

That the Torah believed the later witnesses against the first witnesses is a scriptural decree. Even if the first witnesses were one hundred, and two came and refuted them, saying, “We testify that all one hundred of you were with us on that day in such-and-such a place,” they are punished on the basis of these two, because two are like one hundred and one hundred are like two. Likewise, when two sets of witnesses contradict one another, we do not follow the majority; rather, we reject both.

According to Maimonides, then, it seems that even Abaye agrees with Rava that the credibility of the second set is a scriptural decree and does not stem from legal logic. On his view, the dispute between Abaye and Rava concerns only whether, as a result of this novelty, the disqualification should take effect only from that point onward or also retroactively.

The Lechem Mishneh on this ruling of Maimonides cites the above words of the Tur, who disagrees with Maimonides, and explains that they follow their respective approaches concerning a case in which hazamah is accompanied by hakhashah. There are additional indications that Maimonides and the Tur consistently follow their own views in this matter; for example, see Maimonides there, Laws of Testimony 18:5, and Shiurei Rabbi Shmuel by Rabbi Shmuel Rozovsky, Makkot, no. 229.

A Comment by Later Scholars

Several later scholars—notably Shiurei Rabbi Shmuel there, and the supplementary essays there, no. 3, and Zikhron Shmuel no. 40—noted that the rationale of Nahmanides and the Tur is very difficult. As long as the testimony of the first set is not about themselves, the fact that it has implications for them does not turn them into litigants. If testimony that they were present in a certain place were considered inadmissible because it is self-testimony by litigants, then in every testimony in the world we would have to disqualify the witnesses, since embedded within their words is always the assertion that they themselves were present at the scene. That disqualification would not require a second set to come and refute the first. The testimony should already be invalid as testimony by litigants.

Rabbi Shmuel Rozovsky therefore wrote that it is obvious that Nahmanides and the Tur did not mean to offer a genuine rationale. Obviously, according to everyone, this is a scriptural decree and contains no legal logic. Those authorities meant only this: once Scripture innovated that the second set is believed, the legal characterization of the matter is that the testimony of the first set is treated like testimony of interested parties, and is therefore invalidated.

Indeed, in Nahmanides himself, in the very next verse (Deuteronomy 19:19), one can see that he views the law of conspiring witnesses as a “decree of the Ruler”:

“As he plotted”—and not “as he did.” From here they said: if they already caused the defendant to be executed, they themselves are not executed. That is Rashi’s formulation from the words of our Rabbis (Babylonian Talmud, Makkot 5b). The reason for this is that the judgment of refuted witnesses is by decree of the Ruler, for they are two against two. Thus when two come and testify about Reuven that he committed murder, and two others come and refute them, Scripture commanded that they be executed, because by the merit of Reuven, who was clean and righteous, this happened. Had he been wicked and deserving of death, God would not have saved him from the court, as it says, “I will not justify the wicked” (Exodus 23:7). But if Reuven was executed, we assume that what the first witnesses testified was true, for he died because of his sin, and had he been righteous, the Lord would not have abandoned him into their hands, as the verse says, “The Lord will not leave him in his hand, nor let him be condemned when he is judged” (Psalms 37:33). Furthermore, the Lord would not allow the righteous judges who stand before Him to shed innocent blood, for judgment belongs to God, and in the midst of judges He judges. All this is a great praise of the judges of Israel, and a promise that the Holy One, blessed be He, stands with them and agrees with them in the matter of judgment. This is the meaning of “and the two men who have the dispute shall stand before the Lord” (verse 17), for they stand before the Lord when they come before the priests and the judges, and He will guide them in the path of truth. I have already mentioned this in the section of “These are the ordinances” (Exodus 21:6).

Nahmanides continues there to explain the law of conspiring witnesses on a metaphysical plane, and in light of that he also explains the rule that if they already caused the execution, they themselves are not executed—”as he plotted, and not as he did.” It seems, then, that he has no purely legal explanation for the credibility of the second set of witnesses.

How does that fit with the explanation he himself gave on the previous verse? According to Rabbi Shmuel Rozovsky, these are one and the same: the law is a scriptural decree, and the legal structure of the law is that we reject the credibility of the first witnesses as though they were parties to the case. There is no actual self-interest here, but the scriptural decree is that we view them as parties to the case.

Is There Really No Legal Logic?

Rabbi Shmuel Rozovsky’s claim is not compelling. It is true that every testimony contains, implicitly, the claim that the witnesses were present at the event. But it is obvious that with respect to that point they are not interested parties in their own testimony. As long as the subject is the event itself, the fact that a witness is, in some sense, speaking about himself changes nothing. But when the second set comes to refute them, that very same testimony of the first set takes on a different significance. The subject of the discussion is now not only the event itself, but also the fitness of the first witnesses. In such a situation, the testimony of the first witnesses about themselves can no longer be accepted, because now they are parties to the matter under discussion.

Beyond that, one can suggest other rationales for the credibility of the second set. For example, Penei Yehoshua on Makkot 5a offers the following reasoning: if the second set wanted falsely to save the defendant, it could have disqualified the first set as robbers, or at least contradicted them directly. Beyond that, Nahmanides himself offers another rationale, as we saw above—a more metaphysical one. The relation between these two rationales in Nahmanides is not entirely clear.

Likewise, in Derashot ha-Ran, the eleventh sermon (p. 197 in the Feldman edition; the point is also cited by Abravanel on Deuteronomy 19:14, who adds other reasons for the credibility of the later set), a rationale is given for the law of conspiring witnesses. He writes that the first witnesses relied on the assumption that no one saw them, and therefore their credibility is weaker. The refuting witnesses, by contrast, must fear that if they are lying, someone may discover that they were not in the place they claim to have been, or that the first set was not with them there—for the first witnesses themselves will certainly exert themselves to prove that. Therefore the later witnesses have greater credibility.

B. The Underlying Problem

Can One Execute on the Basis of a Scriptural Decree?

The picture that emerges up to this point—at least according to Rabbi Shmuel Rozovsky—is that in principle hazamah is really just another case of two against two, like ordinary contradiction. From the standpoint of legal reasoning, there remains a doubt whether in actual reality either set is speaking truth, exactly as in hakhashah. But there is a scriptural decree that despite all that, the second set is believed and the first is treated as parties testifying about themselves; therefore they are not believed. In a case of hazamah in capital testimony, we even execute the conspiring witnesses because of their plot to kill an innocent defendant through false testimony.

But we must notice what this means. In actual reality, it is by no means necessary that the defendant is innocent or that the first witnesses are liars. From the standpoint of reality itself, the first set is just as credible as the second; only a scriptural decree tells us to believe the second and execute the first. Can it be that we would take people who, in reality, spoke truth, declare them liars by force of a scriptural decree, and as a result execute them for that offense? As we shall now see, that is utterly inconceivable, and on two different levels.

The Moral Aspect

We can focus the question morally: can it be that the Torah requires us to kill people who did nothing wrong, purely by force of a scriptural decree?

It is important to remember that conspiring witnesses do not require prior warning; see Babylonian Talmud, Makkot 4b. They both come in innocence to testify truthfully, and Jewish law even requires a person to testify if he possesses relevant information. These two witnesses come, fulfill their duty, and testify. Then suddenly, after hazamah—which from the standpoint of reality proves no real falsehood on their part, since the actual situation remains, logically, two against two—we execute them. All this without any warning and without any proven guilt.

From another angle: what were these witnesses supposed to do in order not to become offenders? They had an obligation to come and testify to what they knew, and our assumption is that they really did know what they reported—for even after hazamah we still regard them, at least from a standpoint of doubt, as possibly truthful. And yet the Torah decrees that we must kill them.

At first glance one might compare this to Sabbath violators. There too the Torah decrees that we execute people for acts that ordinary human reasoning would not usually regard as deserving death. They trapped a deer, or separated refuse from food on the Sabbath, and for that they are stoned. In what way is our case of scriptural decree different from other life-and-death decrees? Why do we assume specifically here that such a Torah instruction would be morally very troubling? More generally, practical Jewish law does not usually derive law from the reason of the verse; see Babylonian Talmud, Sanhedrin 21a and parallels. Why, then, should we enter into the reasons of the Torah here?

This argument, however, is mistaken, because the two cases are fundamentally different. In Sabbath desecration, the person knows at the time of the act that the Torah forbids it, and also knows that the Torah punishes it by stoning, even if the reason for that is perhaps unknown to him and to us, and is not legally decisive. Without that knowledge, we really do not punish him. He must undergo the process of warning and acceptance of warning, both regarding the prohibition and regarding the punishment, and only if that process has been completed do we stone him. In our case, by contrast, those witnesses came innocently to testify. It is entirely possible that they testified truthfully. They committed no offense, and the Torah did not warn them against what they were doing—for if they were in fact testifying truthfully, they really did nothing wrong. Moreover, no one warned them—neither regarding the offense nor regarding the punishment. And yet we kill them, just like that.

If so, the conception of conspiring witnesses as a simple scriptural decree is utterly Kafkaesque. The Torah sends people and requires them to testify. After they fulfill their duty, it declares truthful people to be liars by force of a scriptural decree. It then disqualifies them from testimony and sentences them to death because of their supposed falsehood. At no stage in this Kafkaesque process do they have any possibility of defending themselves or avoiding the punishment. They are caught in an impossible trap.

The Legal-Halakhic Aspect

So far we have discussed the moral aspect. But the more serious problem is the legal, exegetical, and halakhic one. The verses explicitly say that the conspiring witnesses are punished for the sin of falsehood. They tried to incriminate an innocent person. But according to the explanation that this is a scriptural decree without legal logic, they did not lie, and he is not necessarily innocent. The question here is not moral but interpretive: according to those who take the “scriptural decree” approach, how can we interpret the biblical verses that instruct us to punish the conspiring witnesses for lying?

If the Torah had wanted simply to decree that we execute them, as in the case of Sabbath violators, it should have told us to kill witnesses who were refuted. I am not dealing now with the moral problem. But that is not all the Torah says. It tells us to execute these witnesses because they are liars. Why should the Torah need to “lie” to us and classify them as liars, who die because of their lying, while in fact they are not such? If the Torah wanted to kill them just because, it should simply have said so, without any such reason.

Put differently: nowhere else in the Torah do we find a scriptural decree that changes reality, that is, that calls black white. Certainly we do not find a scriptural decree that changes reality and then establishes laws—and certainly punishments—on the basis of that virtual reality. Jewish law does not punish black for being white. At most, it determines that black is liable to punishment by scriptural decree.

It therefore stands to reason that if the Torah defines the conspiring witnesses as liars, that is because in its eyes they really are liars. If so, there must necessarily be legal logic here; it cannot be merely a bare scriptural decree.

The Principled Difficulty and the Difficulty in the Position of the Above Interpreters

Up to this point I have described an essential difficulty with the scriptural-decree approach to conspiring witnesses. It cannot be that the Torah decrees that we treat the refuted witnesses as liars even though, in reality, they are not, and at the same time imposes the death penalty on them because of that falsehood.

In itself, however, this difficulty need not trouble us, because as we have seen there are several rational explanations for the credibility of the second set. The real problem lies in the approach of later scholars, such as Rabbi Shmuel Rozovsky, who argued that the rationale of the Tur and Nahmanides is unsound, and therefore concluded that according to everyone this law is a pure scriptural decree. Their assumption is that the view of the Tur and his camp is the more problematic one. Yet, as we have seen, it is specifically the position of Maimonides, if understood simplistically, that is far more problematic—indeed, virtually impossible, for the reasons explained above.

A Possible Solution: A Divine Guarantee

One might have suggested that God guarantees that whenever a case of conspiring witnesses arises, the second set will always be telling the truth and the first set will always be the liars. In other words, there may be no natural legal logic for assuming that the first set is false, and in that sense there is a scriptural decree here. But the Torah guarantees us that in practice the first set will always in fact be the true liars. Divine providence will ensure that no distorted judgment emerges from our hands. This is similar to the metaphysical consideration offered by Nahmanides above, according to which God assists the judges in bringing justice to light and prevents them from erring in judgment. Something of this sort is also found in the midrash in Babylonian Talmud, Makkot 10b, about the inadvertent killer who goes into exile: God causes him and his victim to come to the same inn, where the killer deserved exile and the one killed had deserved death.³

But this too is not a real solution. First, how would God prevent the free choice of wicked people who decide to appear as two witnesses and refute a truthful set? Human beings have free choice even to sin. Second, arguments of this kind can be made when the procedure itself is sensible, but we are troubled by special cases in which an error might emerge and innocent people might be killed. There God can tell us that if we acted correctly, no mistake will issue from our hands. But the procedure itself must have some internal logic, and the divine guarantee merely fills in the gaps. Here, by contrast, the procedure itself is irrational. We could just as well have left the matter in the state of two against two, as we do in hakhashah, and let God guarantee that no harm would come from that. The Torah decrees that we must kill the conspiring witnesses, but according to this explanation there is no reason for that at all. There is only an after-the-fact guarantee that no error will result.

By the same logic, the Torah could have commanded us to fire into the air on the main street every hour on the hour, while guaranteeing that no one who does not deserve it would be injured. Clearly that would make no sense, because the basic command itself would be senseless. The divine guarantee can only come at stage two, after there is already some basic logic to the law, in order to prevent side problems in certain cases. Only if there were some basic logic to such a command to shoot could the Torah then say that it guarantees that no innocent person will be harmed.

C. Is a Scriptural Decree Necessarily an Unintelligible Law?

The Solution: A Scriptural Decree with Reason and Legal Logic

The required conclusion is therefore the opposite of what those later scholars suggested. Not only do the Tur and Nahmanides certainly hold that the foundation of the law is legal logic; Maimonides, who at first glance speaks of a scriptural decree, must also agree that there is legal logic here. According to all views, then, it is clear in actual reality that the first set are the liars and the second set are certainly the credible ones. As we have seen, this cannot be a scriptural decree detached from actual reality.

Two Principled Approaches to the Reasons for Commandments and Laws

An important background to our question appears in Maimonides, Guide of the Perplexed, III:31, where he writes:

Among people there are those for whom it is difficult to assign a reason to any commandment, and who find it preferable that no commandment or prohibition should have any intelligible purpose at all. What brings them to this is a sickness in their souls, one that they cannot articulate and do not know how to express. Their thought is this: if these laws are useful in this world and we were commanded because of that utility, then they would seem to come from the thought and reflection of an intelligent human being. But if a matter has no intelligible purpose at all and brings no benefit, then it is surely from God, for human thought would not produce such a thing. It is as though these weak-minded people think that man is more perfect than his Maker. For man says and does only what serves some end, whereas God—according to them—would command us to do what brings us no benefit and forbid us to do what causes no harm. Far be it from Him, far be it. The matter is the exact opposite: the entire intention is to benefit us, as Scripture says, “for our good always, to preserve us alive, as it is this day” (Deuteronomy 6:24), and it says, “and they shall hear all these statutes and say, ‘Surely this great nation is a wise and understanding people'” (Deuteronomy 4:6). It is thus clear that even the statutes will show all nations that they are wisdom and understanding. If there were something with no known reason, producing no benefit and preventing no harm, why would one call the person who believes in it or practices it wise, understanding, and of great stature, such that the nations would marvel at it? Rather, the matter is unquestionably as we have said: every one of the 613 commandments is either to impart a true belief, remove a false belief, establish a proper social order, remove injustice, train us in good character traits, or warn us away from bad ones. All are dependent on three things: beliefs, character traits, and the practices of political and social governance. We need not count verbal formulations separately, because the verbal statements that the Torah urges us to say or warns us against are themselves included either in social practice, or in teaching true belief, or in teaching character traits. Therefore these three categories suffice in giving a reason for every commandment.

As emerges from Maimonides’ words, there are those who wish to see all the laws of the Torah as scriptural decrees—in other words, as things without reason, or whose reason transcends our understanding. In their eyes, this is the glory of the Giver of the Torah: His ways are higher than our ways. By contrast, there are those who aspire specifically to ground all Torah laws, as far as possible, upon the foundations of reason and human uprightness. In their view, precisely this is the glory of the Giver of the Torah: that His judgments are righteous and straight. Maimonides himself holds that all the commandments have good reasons, whether in the realm of beliefs, character traits, or social-political order.

These two approaches described by Maimonides also have further implications, touching the very meaning of the term “scriptural decree.” Some will say that a scriptural decree is by definition a law without a reason—or at least a law whose reason cannot be grasped by us. Such people likely belong to the first school. Those of the second school, by contrast, can argue that scriptural decrees do exist, but they too have reasons, and we may even understand them, and probably ought to try to do so. According to this second approach, the question arises: why was it necessary for the Torah to write this law in a verse, if one could have derived it through reason? Presumably, because reason alone would not have sufficed to determine the law—whether because the reasoning is not conclusive, or because the quantitative threshold is unclear, or because there are opposing reasons, and so on.

Down to our own day there remains a dispute over how, and to what extent, scriptural decrees should be treated. Some see their very existence as testimony to the loftiness of the Creator’s ways, and therefore view all Torah laws in this manner. Others seek to minimize their number as much as possible, and to make them more intelligible. This disagreement expresses itself in styles of learning, and even in attitudes toward the service of God in general.⁴

Returning to Conspiring Witnesses

As we have seen, Maimonides in the Guide holds that every law has reasons touching beliefs, character traits, or social-political order. It is therefore highly likely that, in his view as well, conspiring witnesses are not simply executed for no reason. Presumably, in his approach this scriptural decree—and perhaps all such decrees—must have a clear humanly intelligible reason. This is all the more so because we execute them specifically as false witnesses, as explained above.

Why, then, does Maimonides speak of a scriptural decree? We should recall that Nahmanides too, in his commentary, speaks of the “decree of the Ruler” immediately after offering a rationale. Rabbi Shmuel Rozovsky took this to mean that the entire foundation is a scriptural decree. But according to our approach, the picture is exactly the opposite: the foundation is the reasoning, and the scriptural decree merely reveals it. That is also how Maimonides may be understood.

What Is a “Scriptural Decree”?

In Encyclopedia Talmudit, s.v. “Gezerat Hakatuv,” the opening definition is: “a rule of the Torah, as opposed to reason.” And that is indeed the accepted understanding of the term. Some laws stem from logic or from a conception of reality; others stem from a decree of the Torah that runs counter to, or at least stands apart from, human reasoning and actual reality.

But what has emerged from our discussion so far is a different understanding:⁵ the verse is what reveals the reasoning and the reality. Once the verse has taught us the law of conspiring witnesses, we now know that they really are liars.⁵

Evidence from Meiri

The Gemara at the beginning of the chapter on the stubborn and rebellious son derives that the law applies only to sons and not to daughters. Both the Babylonian and Jerusalem Talmuds say there that this is a scriptural decree. Yet several medieval authorities nevertheless offered a reason for the law. For example, Maimonides writes in Mishneh Torah, Laws of Rebels 7, and the same appears in Sefer ha-Chinukh:

It is a scriptural decree that only a stubborn and rebellious son is stoned; but a daughter is not judged by this law, because it is not her way to be drawn after eating and drinking as a man is. As it says, “a stubborn and rebellious son” (Deuteronomy 21:18)—and not a daughter, nor a person of indeterminate sex, nor one with both male and female sexual characteristics.

To be sure, this is not necessarily proof for our point, because it is Maimonides’ way—and certainly the way of Sefer ha-Chinukh—to give reasons for commandments that do not purport to be the full and true root of the law.

But Meiri, in his discussion of the stubborn and rebellious son—at the opening of that chapter, on the first Mishnah, s.v. “And this is how he expounds”—writes as follows:

That is, a daughter is not judged by the law of the stubborn and rebellious child at all, because the Torah was concerned only with one whose way is to be drawn after his desires and sink into them, and this is not true of a daughter but of a son. Although all these matters are scriptural decrees, they all follow this principle, even though at first glance some of the details might seem to point in the opposite direction…

Meiri says that this is a scriptural decree, and yet adds that it has a reason—namely, that the Torah is concerned only with one whose nature is to be drawn after his desires. Unlike Maimonides and Sefer ha-Chinukh, Meiri immediately goes on to ask: if the law has such a rational basis, why do the Talmuds call it a scriptural decree? From the very fact that he asks this question, it is clear that in his view this reason is not merely a “reason for the verse,” but the true and full rationale underlying the law that a daughter is not judged as a stubborn and rebellious child. That is why he asks why the Talmuds nonetheless refer to it as a scriptural decree.

What is Meiri’s answer? First, he says that all these laws follow from this rationale—”they all follow this principle”—meaning that this is a true and correct rationale, the basis of all the laws of the stubborn and rebellious child. In the continuation of his remarks, he apparently seeks to explain why this is nonetheless called a scriptural decree. At first glance, the opposite might have seemed reasonable. One might have reasoned, for example, that one should be more stringent specifically with a daughter, or specifically with one who steals from others rather than from his parents; so he cites from the Jerusalem Talmud. In other words, there is indeed a rationale here, but other and opposite rationales could also have been proposed. Therefore the Torah had to write that this particular rationale—the concern with one who is naturally drawn after his desires—is the decisive one, and the basis of all these laws.

The conclusion is that a verse is indeed needed here, in order to reject the competing rationales. But in the end we understand that the true reason for the scriptural decree is that the Torah is concerned with one whose way is to be drawn after his desires. This is thus a scriptural decree without which we could not have decided clearly in favor of that rationale, but once the verse is given we understand its reason fully, and everything follows from that rationale. That is exactly our point.

A Note on the Ran’s Derashot

It is interesting that the author of Derashot ha-Ran, whose rationale for the credibility of the refuting witnesses we cited above, himself asks there how the Gemara can say that the credibility of the refuting witnesses is a “novelty” if there is legal logic behind it.⁶ He answers that the credibility of the second set is indeed greater than that of the first, by virtue of the reasoning we cited above. However, that difference is not enough for us to decide on that basis, and certainly not in capital cases. Therefore a verse in the Torah is needed to instruct us to rely decisively on that rationale of the superiority of the second set, even to the point of executing on that basis. That is the meaning of the scriptural decree in the law of the credibility of refuting witnesses.

This is precisely the principle we saw in Meiri, and it is also how we explained Maimonides above regarding the credibility of the refuting witnesses. Below, however, we will see that the author of Derashot ha-Ran probably means something slightly different.

Several Types of Scriptural Decree

Encyclopedia Talmudit, s.v. “Gezerat Hakatuv,” notes that there are several types of scriptural decrees. There are decrees that run against human reason and logic. There are decrees that run against other principles found elsewhere in the Torah itself. And there are decrees that involve an internal contradiction within the very law in which the innovation appears—a kind of paradox, as in the red heifer, which purifies the impure and defiles the pure, especially since the impurity falls specifically on the person engaged in the purification.⁷

Up to this point, we have understood the scriptural decree concerning the credibility of the refuting witnesses as belonging to the first type, that is, a decree against human logic. But from the Ran’s words it emerges that he sees it as a decree of the second type: a contradiction to other Torah principles. This is what he writes:

This does not contradict what the Sages said, that conspiring witnesses are a novelty—”why do you see fit to rely on these? Rely on those.” What they mean is that it is not the way of the Torah to rely on what appears likely from the facts as to whether something is true or false, but only on the testimony of two witnesses. For the Torah believed even two utterly lowly men, so long as they are not disqualified, just as much as Moses and Samuel, whom we know would not alter anything, even in casual conversation.

Thus, if here we rely more on the refuting witnesses than on the refuted ones because the words of the refuting witnesses appear more likely to be true, that is a novelty and a departure from the Torah’s general method elsewhere. Nevertheless, the facts do indeed appear as we have written…

The Ran determines that ordinarily the Torah does not allow us to rely on reasoned assessments of reality, certainly not in capital law. In capital cases, circumstantial judgments do not suffice. See Maimonides, Mishneh Torah, Laws of Sanhedrin 20 and 24, on the contrast between monetary and capital law in this respect.

If so, the reasoning behind the credibility of the refuting witnesses may be excellent and complete. We may have an excellent reason why the later witnesses should be believed. The problem is that in capital law one ordinarily cannot rely on reasoning alone, however good it may be. If so, the problem is not that the rule fails to fit our logic or that the rationale is unintelligible. The problem is that it does not fit the Torah’s ordinary rules of evidence. According to the Ran, this is a scriptural decree of the second type.

This explanation is somewhat different from what I suggested above. The earlier direction, in line with Meiri, saw the role of the verse as strengthening the reasoning because doubts and objections could be raised against it, or because opposing rationales could be offered. Now, by contrast, we are not relying on doubts about the rationale itself, but on the fact that the Torah generally is simply unwilling to rely on reasoning, however good, in capital cases. The verse is needed to instruct us to rely on a rationale that may be very strong in itself, and to override the principle that in capital cases one does not rely on reason alone. But the verse does not come to strengthen the rationale itself.

In these two types of scriptural decree, then, the verse plays a different role. In the first type, the verse reveals that the rationale is correct and decisive, and that one may rely on it. In the second type, the verse reveals that in the specific context under discussion it is permissible to rely on reasoning at all. More generally, in this type of scriptural decree the Torah instructs us that we may depart from the ordinary legal framework it establishes elsewhere. In any event, in both cases the fact that something is a scriptural decree does not negate the existence of legal logic, and as we have seen, sometimes there is even complete and compelling logic.

Conclusion Regarding Conspiring Witnesses

The conclusion, then, is that according to all views there is legal logic at the basis of the rule that we believe the refuting witnesses and not the conspiring witnesses. A verse is indeed needed, either because the reasoning is weak and not decisive, or because there is a general rule that in capital law we do not follow reasoning. But once the verse is there, it is clear that there is some logic and reason why the second set is believed more than the first. The dispute among the medieval authorities, which later scholars tend to understand as if it concerned whether there is a scriptural decree here or a rationale, is not really of that kind. Its details will be clarified below.

D. Explaining the Dispute Between the Tur and Maimonides: “It Extends No Further than Its Novelty”

“It Extends No Further than Its Novelty”⁸

We now come to the implications of all this. The main consequence of classifying a law as a “scriptural decree” is that one does not handle it in terms of ordinary human logic.⁹ In the case of a scriptural decree, we do not ask on the logical plane what it would make sense to do. Rather, we ask on the exegetical plane what the Torah commands us to do. The contours of the law are determined by what is written in the verses, not by our own logic.

One of the clearest expressions of such an approach is the rule: “it extends no further than its novelty.” That is, we tend to restrict the scope of a scriptural decree as much as possible. Wherever there is no necessity, we interpret in accordance with ordinary human logic. Only where we have no choice do we say that the Torah innovated that we must depart from that logic.

In the case of conspiring witnesses, Rava says that the disqualification from testimony, which is a scriptural decree, applies only from the time of hazamah. In other words, we restrict the disqualification in a way that runs contrary to ordinary human logic. If the witnesses were indeed actual liars, ordinary logic would require that the disqualification begin from the time they testified, because that is when the falsehood was committed, and from then on they should be classified as wicked. But because their classification as wicked is by scriptural decree and not by logic, we apply it only to the minimal period possible from an exegetical standpoint.¹⁰

But our entire discussion thus far suggests that even a law that is a scriptural decree may still have logic behind it. The conspiring witnesses are indeed liars, and there are good reasons for saying so. If that is the case, why minimize the scope of the law? If the law is intelligible and rests on an actual rationale, it would seem that we should apply it everywhere that the underlying rationale is relevant. In the case of conspiring witnesses, that would mean disqualifying them from the moment they testified.

And indeed, as we have seen, according to Maimonides and the Ran at least, even Abaye holds that conspiring witnesses are a novelty, yet he still applies the disqualification retroactively. Presumably, he does so for exactly this reason. It is a novelty, but one with legal logic behind it, and therefore there is no reason to restrict the law.

But the difficulty remains in full force with respect to Rava. Rava concludes from the assumption that this is a novel law—a scriptural decree—that one must limit its scope. Is he unaware of the logic underlying this rule? Does Rava really think that we execute the witnesses without legal logic? Put differently, all the difficulties raised until now remain valid at least with respect to Rava’s view.

Furthermore, Abaye as well—and certainly the decisors who rule in practice like him—do not deny the general rule that one should narrow the scope of a law learned from a scriptural decree. Presumably all would agree to the rule “it extends no further than its novelty,” which appears explicitly or implicitly in many places in rabbinic literature. Must we therefore say that in every place where that rule is applied, there is no rationale at all behind the scriptural decree? As noted, that certainly cannot be so in Rava’s view, since he speaks about the execution of conspiring witnesses, and all the earlier difficulties would immediately reappear.

Applying the Principle to the Two Types of Scriptural Decree

We saw above that one must distinguish between two types of scriptural decree. With regard to a scriptural decree that stands against ordinary reasoning, the role of the verse is to sharpen and reinforce the rationale and to reject competing rationales. In the end, the rationale remains weak on its own terms, and therefore even after the scriptural decree is introduced, there is still a genuine novelty here. In such a situation, it is indeed reasonable to restrict the scope of the novel law to the minimum possible.

By contrast, in a scriptural decree whose basic problem is that it conflicts with other legal principles, while the rationale itself is perfectly intelligible—as in the Ran’s understanding of conspiring witnesses—there is no necessity to minimize the scope of the law.

One could also suggest the exact opposite: specifically in the second case one should minimize the deviation from the ordinary principles of Jewish law as much as possible, whereas when the verse teaches us a rationale, there is no reason to narrow the law’s application, for if the rationale is valid, it should be valid everywhere equally.

In any event, although it is difficult to formulate universal rules on this issue, it is quite clear that different considerations may lead us in some cases to minimize the law’s scope, and in other cases to apply it wherever the rationale we attribute to it remains valid. This is what we saw in Meiri, who appears fairly confident in the rationale behind the law of “a son and not a daughter,” and therefore says that everything follows from that rationale. One would therefore not expect Meiri to minimize the scope of that law.

Applying This to Conspiring Witnesses

Let us now return to conspiring witnesses. We saw that Rava narrows the scope of their disqualification as much as possible, and it seems that he understands this as a novel scriptural decree whose rationale remains doubtful, even though it cannot be utterly devoid of rationale, since we execute on its basis, for all the reasons explained above. In such a case, one should minimize the innovation as much as possible.

Abaye, by contrast, also agrees that we are dealing with a novel scriptural decree—at least according to Maimonides—but in his view that is not a reason to restrict the scope of the disqualification of conspiring witnesses. He apparently places full confidence in the rationale behind this law.

Different Applications of the Principle of Limiting the Scope of a Scriptural Decree

We have seen that the principle “it extends no further than its novelty” can be applied in different ways to different scriptural decrees. Let us now consider an even more surprising situation: there are cases where a certain decisor minimizes the scope of a law as much as possible in some contexts, and yet in other contexts that same decisor applies that very same law more broadly than the bare minimum.

To examine this with respect to conspiring witnesses, we must return to the remarks of the Lechem Mishneh cited above concerning the dispute between the Tur and Maimonides. As noted, at first glance it seems that the Tur and Maimonides disagree on whether, according to Abaye, the disqualification of conspiring witnesses is a novelty—Maimonides—or whether there is logic behind it—the Tur. That is also how the Lechem Mishneh understood their dispute. He connects this to another dispute between them, concerning a case in which there is both hazamah and hakhashah between the two sets. Maimonides writes as follows in Mishneh Torah, Laws of Testimony 18:2:

How so? Witnesses came and said, “We saw this man kill a person,” or, “He borrowed one maneh from so-and-so on such-and-such a day in such-and-such a place.” After they testified and were examined, two others came and said, “On that day and in that place we were with you and with them the entire day, and nothing of the sort happened: this one did not kill that one, and this one did not borrow from that one.” This is contradiction. Likewise, if they said to them, “How can you testify this way, when the killer, or the one killed, or the borrower, or the lender, was with us on that day in another province?” this is contradictory testimony, for it is as if they said, “This one did not kill that one, and this one did not borrow from that one,” since they were with us and this event did not occur. The same applies in all similar cases. But if they said to them, “We do not know whether this one killed that one on that day in Jerusalem, as you say, or did not kill him. We testify only that you yourselves were with us on that day in Babylonia,” then these are conspiring witnesses and are executed or must pay, since the witnesses who refuted them paid no attention to the substance of the testimony itself, whether it was true or false.

Maimonides repeatedly emphasizes that hazamah must appear alone, without any accompanying hakhashah. A case that combines hazamah and hakhashah is, in his view, simply hakhashah and not hazamah. By contrast, as the Lechem Mishneh notes there, the Tur quite clearly disagrees with Maimonides and treats such a case as full-fledged hazamah.

As noted, the Lechem Mishneh explains that each follows his own broader approach. The Tur understands the rule that the later witnesses are believed against the earlier ones as a rational legal principle. It is therefore obvious that he will say the same even in a case of hazamah combined with hakhashah. After all, the rationale that the first set are parties to the case when they speak about themselves clearly applies there as well. Maimonides, by contrast, sees the credibility of the second set as a scriptural decree, and therefore rules: “it extends no further than its novelty.” In his view, the credibility of the later witnesses exists only in the pure case of hazamah, because that is certainly what the Torah addressed, and not in any other case. He limits the scope of the law to the minimum, as is appropriate for a scriptural decree.

On the other hand, Maimonides clearly rules in the basic dispute like Abaye and not like Rava. According to his view, even though the law is a special novelty by scriptural decree, the disqualification applies retroactively and not only from the time of hazamah, as Rava held. So here Maimonides does not implement the principle “it extends no earlier than the time of its novelty.” The Tur, by contrast, is again consistent with his own position, since he rules like Abaye because he does not regard the law as a novelty, and therefore it applies retroactively.

Another implication of these approaches appears in the discussion of hazamah not in the presence of the witnesses.¹¹ In Babylonian Talmud, Ketubbot 20b, it is stated that if the second set refutes the first not in their presence, this is not hazamah but hakhashah. The commentators disagree in understanding this law. From Rashi there it appears that the case does not thereby revert to ordinary two against two. Even when the hazamah was not in their presence, although the punishment of “as he plotted” is not carried out against the first set, their testimony is still certainly void, and not merely left in a state of doubt like ordinary two against two. The same view appears in Responsa of Rabbi Isaac bar Sheshet, no. 266, cited in Beit Yosef, Hoshen Mishpat 38. Rabbi Isaac bar Sheshet adds the further novel claim that the first witnesses are also disqualified in such a case, not merely that their testimony is invalid and not accepted. On this point, however, many disagree; see Shakh, Hoshen Mishpat 38:2, and Noda BiYehudah, first edition, Even ha-Ezer 72.

The common denominator of these views is that even when the hazamah was not in their presence, we have not returned to an ordinary case of two against two. Presumably, this accords with the Tur—and indeed, in Rabbi Isaac bar Sheshet the Tur‘s reasoning appears explicitly—that there is legal logic behind the credibility of the later witnesses. Since that rationale also exists where the hazamah was not in their presence—the first set is still testifying about itself and is therefore like a litigant—the later witnesses should be believed there as well.

But in Maimonides, Mishneh Torah, Laws of Testimony 18:5, it is explicitly implied that when they are refuted not in their presence, this is exactly an ordinary case of hakhashah, and their testimony is invalid only because of doubt. This is his language there:

Witnesses are refuted only in their presence, and contradicted not in their presence. Witnesses who were refuted not in their presence are treated as contradicted. Therefore, if the witnesses who refuted them died before refuting them in their presence, there is no testimony here, for these contradicted those.

It is explicit in Maimonides’ language that when the hazamah is done not in the presence of the refuted set, the case is an ordinary contradiction. Seemingly, Maimonides is again following his own view: hazamah is a novelty, and it extends no further than its novelty. Therefore, in any situation in which that innovation was not expressly stated, we are dealing only with ordinary hakhashah.

What does the Tur hold in this matter? If one examines his wording in Hoshen Mishpat 38, it appears that he takes exactly the same position as Maimonides, namely that such a case is ordinary hakhashah. If so, it would seem that here the Tur departs from his own general line and does invoke the principle “it extends no further than its novelty.”

The Meaning of This: Mixed Approaches

The upshot of our discussion is that there are tensions, both in the Tur and in Maimonides, regarding whether one adopts a method of narrowing the law by applying the principle “it extends no further than its novelty” to the scriptural decree of conspiring witnesses. Let us summarize the points raised in the previous subsection in a table. The exceptional case in each position is marked with question marks:

Issue	Maimonides	Tur
Is there legal logic here?	No	Yes
Hazamah together with hakhashah	Hakhashah	Hazamah
Hazamah not in their presence	Hakhashah	Hakhashah ?
Does disqualification apply retroactively?	Yes ?	Yes

The Tur generally adopts an approach that fits the assumption that there is legal logic behind the credibility of the later witnesses, and therefore he does not apply the principle “it extends no further than its novelty.” The exceptional case is hazamah not in their presence. Maimonides, by contrast, generally adopts an approach that fits the assumption that there is no rationale and that this is a scriptural decree, and therefore he does apply the principle that the innovation goes no further than its novelty. His exceptional case is that the law of hazamah applies retroactively.

What the two approaches share is that in both the Tur and Maimonides we see a mixed application of the principle “it extends no further than its novelty.” We noted above that different considerations may arise with respect to scriptural decrees—whether to narrow their scope or not. But here we see that with respect to one and the same scriptural decree, there are aspects in which the law is narrowed and at the same time aspects in which it is not.

How can such a situation arise? If the scriptural decree expresses an intelligible rationale, then one might have expected us to ignore the principle “it extends no further than its novelty” in all respects. By contrast, if there is no clear rationale here, or if the law conflicts with other legal principles, then one might have expected us to narrow the law’s scope in all respects. How can such mixed attitudes be logically possible? Does the same scriptural decree contain more intelligible and less intelligible rationales, each of which governs a different aspect?¹²

Beyond that, the very existence of disputes over how to treat scriptural decrees shows that the decisors’ own reasoning plays a role in shaping the law even in sugyot whose foundation is a scriptural decree. That itself hints at the presence of rational dimensions even in the world of “scriptural decrees.”

In the following subsections, I will try to offer possible explanations of both positions. My purpose is only to make the ear receptive to the existence and possibility of mixed approaches of this kind. These examples highlight the potential richness of possible approaches to laws that are, on the face of it, “unintelligible,” that is, scriptural decrees, and how the decisor’s own reasoning has a place even in halakhic determination in such cases.

A Possible Explanation of Maimonides’ Position

As noted, Maimonides holds that the credibility of the later witnesses is a scriptural decree. Yet, as we have seen, it cannot be that in his view there is no rationale at all behind this decree, for if that were so, it would be impossible to execute the conspiring witnesses. The question is what status that rationale has, and how it is weighed and affects the application of the principle “it extends no further than its novelty.”

We saw that Maimonides consistently follows his own approach, and almost all the practical implications show that he treats the credibility of the later set as a scriptural decree. It therefore seems that the rationale is not very strong in his eyes, or at least not sufficiently transparent—in other words, one cannot shape the laws on the basis of the rationale we understand in this case. For that reason, Maimonides chooses to present this law as a scriptural decree, even though there must, in his view too, be some rationale at its root. Therefore he applies the principle “it extends no further than its novelty” in an almost complete way.

We still need to explain why, with respect to the beginning of the disqualification, he rules like Abaye that it applies retroactively. Put differently: according to Maimonides, how does Abaye disagree with Rava if both agree that the law of hazamah is a scriptural decree?

An initial clue is that the language Rava uses is not the ordinary rabbinic formula, “it extends no further than its novelty,” or anything close to it. Here Rava uses a slightly different formula: “it extends no earlier than the time of its novelty.” It seems that this reflects a somewhat different principle, and not the usual rule of “it extends no further than its novelty.”

The explanation is this: Abaye claims that although this is indeed a novelty, and therefore from an exegetical standpoint one should narrow its scope as much as possible—after all, scriptural decrees are handled more by interpretive tools than by logical ones—the narrowing should be applied only to different cases, that is, only where one can say that the Torah clearly addressed or did not address them.

But with respect to the time axis, narrowing or broadening is not an interpretive question. No one claims, on the level of textual interpretation, that the Torah says the disqualification it speaks of applies only from the time of hazamah onward. The Torah innovated that the first set are liars, and once that is so, there is no reason to distinguish between the time of testimony and the time of hazamah. Here the fact that there is a rationale becomes important: Maimonides too agrees that there is a rationale behind the disqualification—namely, that in reality it is clear that the first set indeed lied, and not that we execute them despite their having spoken truth, which is the naïve understanding of a “scriptural decree.” If so, there is no interpretive reason to limit the disqualification only to the period after hazamah. If in fact they lied, then from the moment they testified they are liars and therefore disqualified. That, according to Maimonides, is precisely Abaye’s argument against Rava. Hence Maimonides, who rules like Abaye, disqualifies the first set retroactively. A restriction that is not interpretive cannot play a role with regard to a scriptural decree that does have legal logic behind it.

All the other restrictions in the table concern different situations. For example, where the witnesses were refuted not in their presence, one can say that the Torah did not speak about that. Or where they were both refuted and contradicted at the same time, one can argue interpretively that the Torah did not speak about that either. But one cannot say that the Torah did not speak about the period between testimony and hazamah. That is not a different case, but a restriction within the very same case. In the same case of hazamah, Rava proposes limiting the disqualification to the time of hazamah. But, as noted, that is not an interpretive move. The Torah did not address the time axis at all, and if they are liars—and Maimonides must agree that in reality they are liars—then there is no place to distinguish between those times.

This reasoning is so clear that it is actually hard to understand why Rava himself does not accept it. Why, in his view, can one distinguish between the two times? Clearly Rava takes one further step. He makes not an interpretive restriction, but a substantive legal one. He rules that although the Torah established that the first set are liars, he nonetheless does not apply that to their disqualification except from the time of hazamah onward. This is an artificial narrowing, and it is not entirely clear how one can do it at all. Therefore Maimonides’ ruling like Abaye is quite understandable; indeed, it is Rava’s view that requires explanation.¹³

A Possible Explanation of the Tur’s Position

Let us now try to understand the position of the Tur. The point of departure is that the Tur holds that there is a good rationale at the basis of the disqualification of the first set. Therefore, even though a scriptural decree is involved, he chooses to present the law as grounded in legal logic, unlike Maimonides.

As can be seen from the table above, in most respects the Tur follows this approach and rules accordingly: the first set is not believed because it is a party to the case with respect to its own testimony. The exception is the law of hazamah not in their presence, where the Tur departs from his usual pattern and treats the matter as ordinary hakhashah. In such a case, it would seem that his rationale still exists, yet he nonetheless chooses to narrow the scope of the innovation of the first set’s disqualification and the later set’s credibility. Why does he depart from his usual line there?

To understand this, we must return to the passage in Ketubbot. Why does the Gemara there assume that hazamah must be specifically in their presence? Rashi there, s.v. “And just as,” explains:

And just as witnesses cannot be refuted except in their presence—because the others come to obligate them, whether to punishment of life or of money; and the Torah said, “the owner of the ox must stand over his ox” (Babylonian Talmud, Bava Kamma 112b).

Rashi explains that the rule requiring hazamah to be done specifically in the witnesses’ presence derives from the general law that when testimony is received against a person, that person must be present.

We can now understand the Tur‘s position very well. We saw that the basis of his rationale is that when the second set comes, the subject of the discussion becomes the first set itself, and no longer only the event; therefore they become parties to the case. Those who disagree with him apparently maintain that since the testimony of the first set was originally given concerning the event, the fact that it later has consequences for their own fitness does not make it self-testimony by litigants.

If so, precisely the Tur, who sees this as testimony about the first witnesses as litigants and as the subject of the testimony, must also import here the rule that such testimony must be given in their presence. If the hazamah is not done in their presence, then no testimony was received against them as litigants; what remains is only contradiction regarding the event itself. Therefore, according to the Tur, if the hazamah was not made in their presence, all that remains is hakhashah regarding the event. If so, this is not a formal narrowing of the scope of the scriptural decree at all. On the contrary, it follows directly from his own rationale.

We can now understand why he treats such a case as ordinary hakhashah. Not because he is narrowing the scope of the law of hazamah, but because his own rationale does not exist in that case. The conclusion is that there is no action here of restricting the law’s scope—that is, no ruling contrary to his general doctrinal position. On the contrary, this is an application and consequence of his own rationale. It is a direct continuation of the Tur‘s general conception.

By contrast, Maimonides, who rejects the Tur‘s rationale—or at least does not regard it as decisive—does not view the later testimony as testimony about the earlier witnesses. Therefore what remains is the question whether to narrow the scope of the scriptural decree or not, and here Maimonides follows his own general line and adopts a policy of minimizing the scope of scriptural decrees as much as possible.

The conclusion is that the Tur and Maimonides apparently rule the same way in this case, but each for a different reason. According to the Tur, this is an application of his rationale. According to Maimonides, it is a narrowing of the scope of the scriptural decree concerning the credibility of the later set, in line with the principle “it extends no further than its novelty.”

Concluding Note

Let me conclude by noting that it is hard to believe that one would find an approach like that of the later scholars mentioned above in an actual court that had to judge such a case in practice. Would judges really sentence a person to death while believing him innocent, and calm themselves by saying that there is a scriptural decree here? There is almost no doubt that, for one reason or another—as in the law of a case suspected of fraud—they would find a way to avoid doing so. If judges were not convinced that the first set had in fact lied, they would not execute them.

Is there perhaps a difference here between conceptual analysis in the study hall and practical judicial decision-making—in the spirit of “had we been in the Sanhedrin, no person would ever have been executed”; see the end of the first chapter of Makkot? Is this symptomatic of theoretical learning more generally? Or perhaps this is indeed how things should work—a similar distinction exists in the contemporary legal world between legal theorists and practicing jurists. The theorists of Jewish law formulate the scriptural decree and its implications, while those who actually sit in judgment apply it with sensitivity to common sense and reality. All this still requires further thought.

Footnotes