Commentary

Civilian Casualties in the Light of Halakhah and Ethics: Revisiting Rav Shaul Yisraeli’s Analysis

Qibya from Nili

 

Aryeh Klapper

General Introduction
On the night of October 12, 1953, terrorists infiltrated from Jordan and threw a hand grenade into a private house in Israel, killing a mother, her three-year-old daughter, and her infant son. Her thirteen-year-old son was severely injured. Three nights later, a retaliation raid into the Jordanian village of Qibiyeh blew up more than forty houses, and in the process killed approximately sixty residents, including women and children.

In 1954, Rav Shaul Yisraeli published a magisterial article, “The Raid on Qibiyeh by the Light of Halakhah” (Hebrew),[1] examining whether the raid was halakhically justifiable. The article has evergreen significance for a wide variety of halakhic and metahalakhic topics. As with many groundbreaking works that “uproot mountains and grind them against each other,” its practical implications may be tempered or rejected without diminishing the honor due its author or the magnitude of his achievement.

R. Yisraeli prefaced the article with a narrative about the raid. This narrative is an essential context for evaluating his arguments, regardless of its historical accuracy.

He believed that the terrorists unquestionably came from Qibiyeh, as had many previous attackers, and that they were supported and encouraged in their terrorist activities by “the majority that might as well be the entirety” of Qibiyeh’s residents. He conceded that the raid killed noncombatant men, women, and children. He assumed that the raiders knew that women and children were present and vulnerable, but does not say whether they were targeted. 

My primary concern here is to understand and evaluate the article’s halakhic position about civilian casualties in wartime. But a metahalakhic introduction is warranted.

Halakhah and Ethics
Halakhah and Jewish ethics are not the same discipline practically or intellectually. Halakhah may leave practical options open that ethics foreclose, or demand behavior that ethics proscribe, and vice versa. Halakhah may utilize formal reasoning that is meaningless in ethical terms, and ethics may assume principles never formulated in halakhic terms.

Competent Jewish ethicists nonetheless derive much of their data from halakhah, and competent poskim nonetheless account for ethics in their halakhic process. These accountings may not be explicit. A halakhic responsum may appear to move smoothly and inexorably, via formal reasoning, to its edifying conclusion. Yet, latent flaws in the formal arguments might have been exposed had they led to ethically disturbing outcomes. Conversely, legal principles that individually seem incompatible with ethical principles may combine to generate ethically satisfying outcomes.

A standard principle of modern ethics[2] is “universalizability,” meaning that rules for action must bind every rational actor in the same way. A related principle is ‘reversibility,” meaning that the rules for Side X toward Side Y must be identical to the rules of Side Y toward Side X. Thus, the rules of military ethics must be the same for all parties to a conflict, and they apply regardless of soldiers’ race, intelligence, gender, etc.

However, Dr. Noam Zohar argues convincingly in Alternatives in Jewish Bioethics[3] that universalizability cannot be an absolute requirement without dramatically limiting the scope of ethics. Questions such as “Should I violate Shabbat to help an elderly person carry a package?” have no meaning unless the actor considers themselves Jewish. Therefore, ethics that don’t take subjective identifications and motivations into account will be of limited relevance and use to real people.

I suggest that Dr. Zohar’s argument can be extended to methods as well as cases. To demand that a Jew abandon halakhic reasoning in any area of practical decision making is to totalitarianize ethics, just as it would to assume that a Muslim may not consider hurting a host’s feelings in order to avoid eating pork.

The profound challenge is to incorporate the particularistic without making ethics meaningless. One possible approach is to evaluate the outcomes of alternate systems within an ethical framework to see whether they are defensible, as opposed to whether they would be the most likely result of ethical reasoning. 

R. Yisraeli’s halakhic analysis in “The Raid on Qibiyeh” does not pay even lip service to universalizability and reversibility. He assumes that halakhah obligates, treats, and values Jews and non-Jews differently. Any ethically meaningful evaluation of his article therefore must focus on its conclusions rather than on its arguments. Which asymmetrical outcomes does he accept, which does he reject, and why?

Civilian Law 
R.Yisraeli first examines a set of potential justifications based on civilian law, including:
1. holding the village capitally responsible under Noahide law for not bringing the terrorists to justice

2. regarding the terrorists as agents of the village citizenry, and therefore holding the citizenry guilty of murder, even if the terrorists would have killed just as many without support from the village citizenry

3. regarding everyone in the village as rodefim (enablers of pursuers-with-intent-to-kill), and therefore liable to be killed if necessary to stop the pursuers.

R. Yisraeli rejects #1 on the grounds that the terrorists would likely have punished or even killed any villager daring to oppose them, and that non-Jews are not required to sacrifice their lives (Rambam), or perhaps even endure serious deprivation (his understanding of Ramban), rather than transgress. Furthermore, Tosafot require a formal court verdict for non-Jews to be capitally punished.

He rejects #2 on the grounds of the general halakhic principle that “there is no agency for Gentiles,” meaning that halakhah does not recognize non-Jews either as agents or as having agents.

He rejects #3 on the grounds that the raid killed children, whereas only willing, conscious, and responsible humans are liable to execution as accessories-to-pursuers. However, he finds #3 sufficient to justify the collateral deaths and even the targeting of adult noncombatants. In other words, he believes that one may kill a deliberate rodeif (enabler) if killing the enabler will save the rodeif’s intended victims.

A possible explanation for R. Yisraeli beginning with non-war arguments is that the civilian law section of the essay may establish principles that are crucial to his later analysis of the laws of war. Here are two examples:

1) In his discussion of agency, R. Yisraeli raises the possibility that assassination-by-agent is prohibited as an exception to the general rule that non-Jews are excluded from halakhic agency. This exception is midrashically derived from Genesis 9:5.

However, while Rambam includes the prohibition against hiring assassins in his outline of the prohibition of bloodshedding (Laws of Murder and Preserving Life 2:2), he omits it in his outline of the Noahide prohibition of bloodshedding (Laws of Kings 9:4).

The prohibition against suicide is also derived from Genesis 9:5. This prohibition, as well, is included in Laws of Murder and omitted in Laws of Kings. R. Yisraeli concludes that Rambam holds that the prohibitions against suicide and hiring assassins do not bind non-Jews[4].

He then takes this conclusion two steps farther. First, he argues that if suicide is not prohibited to non-Jews, they may license others to kill them. Killing a non-Jew who has consented to be killed is therefore not a violation of bloodshedding.

Second, he contends that such consent can be given in advance, conditionally, and reciprocally. Thus X and Y may legally contract to duel to the death, and the surviving party will not be liable so long as they obeyed all agreed-upon rules. This analysis applies to soldiers as well.

2) In his discussion of rodeif, R. Yisraeli notes an apparent contradiction in Rambam’s Mishneh Torah. Hilkhot Yesodei Ha-Torah 5:4 states that although a Jew is obligated to die rather than violate avodah zarah (idolatry), gilui arayot (forbidden sexual relations), or shefikhut damim (murder), a Jew who violates those prohibitions under threat of death is exempt from punishment. However, 5:6 states that a Jew who heals themselves from a fatal condition by violating one of those prohibitions is given “the punishment appropriate to their sin.” Why do these cases have different outcomes?

R. Yisraeli explains that in 5:4 the violation is considered coerced (“the action is not done at all with his intent”), whereas in 5:6 it is considered a willing choice to commit the sin, even though the motivation is to save one’s own or someone else’s life. Therefore, in 5:4, the only culpable violation is failure to fulfill the command to sanctify God’s Name, which is not a capital crime. In 5:6, however, a capital prohibition has been deliberately flouted.

If violations performed to save a life are not considered coerced, why may the deathly ill violate prohibitions other than “the big three” to save their lives? R. Yisraeli answers that the verse “and live by them” (Vayikra 18:5), which entails an obligation to save life, overrides many prohibitions.

However, he contends, “and live by them” is not by itself sufficient to override all prohibitions other than avodah zarah, gilui arayot, and shefikhut damim. He derives from a number of Rishonim that lifesaving does not justify imposing costs on third parties, and therefore does not justify violating prohibitions such as theft. It follows that “and live by them” cannot permit killing someone else to save one’s own or another’s life.  Therefore, “and live by them” is not sufficient to permit killing a rodeif.  

Why then is killing a rodeif permitted? R. Yisraeli answers that a deliberate rodeif forfeits their life, so death is not legally an imposed cost.

However, Rambam applies the law of rodeif in two cases that cannot be considered “deliberate.” First, in Laws of Murder and Preserving Life 1:9, he declares that a fetus may be aborted to save the mother’s life because it is a rodeif. Second, in Laws of Wounders and Damagers  8:16, he rules that one may throw another person’s luggage overboard to prevent a ship from sinking and causing those on board to be drowned. These cases seem to demonstrate that the status of rodeif is independent of intentionality.

R. Yisraeli responds that when X is pursuing Y directly, then one is permitted, and even obligated, to kill X to save Y, even if X has no malicious intent; all the more so if X is inanimate luggage. But if X is indirectly threatening Y, then if X is a person, even if killing X would save Y, X is considered a rodeif only if acting deliberately. Thus, one cannot kill an emerged infant to save the mother, even though one could have aborted it as a fetus moments earlier.

Therefore, even in the context of war, the law of rodeif would not permit targeting infants to save one’s soldiers, even when the infants are being used as human shields by the enemy. It certainly would not allow killing infants in the course of a retaliatory raid intended to deter acts of terrorism.

However, there remains an avenue of halakhic justification for the raid via rodeif, one that brings us back to this essay’s introduction. Our discussion of rodeif so far has made no distinctions between Jews and non-Jews. R. Yisraeli is open to such distinctions.

We noted above that Rambam permits aborting a fetus to save the life of its mother, but forbids killing an infant under otherwise identical circumstances. One might argue that the difference between the cases is the homer ha-aveirah (the severity of the sin), and define that as follows: killing an infant is a violation of retzihah (the prohibition of murdering a “fellow”), derived from Exodus 14:14, whereas killing a fetus is a violation of shefikhut damim (bloodshedding), derived from Genesis 9:6.

The prohibition of killing Gentiles is also derived from Genesis 9:6. It would follow that the rules for killing an infant Gentile as a rodeif are the same as those for killing a fetal Jew, meaning that one can kill even an indirect and unintentional infant gentile rodeif. That would suffice to justify the killing of infants in Qibiyeh.

One might object on the grounds that for a Jew to kill an unintentional Gentile rodeif, whether infant or adult, is a violation of the moral principle “Who says that your blood is redder than his?!”  (Sanhedrin 74a). But R. Yisraeli does not treat that as a purely rhetorical question. He is in principle willing to say that Jewish blood is redder.

However, he argues that because a hiyuv hatzalah (duty to save a life) is not sufficient to impose costs on someone else, the moral sevara (logic) of “who says your blood is redder than his” is unnecessary to forbid killing one human being to save another. The sevara is necessary only when from the perspective of halakhah there are no imposed costs. (Conversely, the sevara can apply even in cases where all prohibitions against killing are overridden.)

R. Yisraeli therefore does not justify the Qibiyeh raid using this approach. Instead, he argues that the principle of not imposing costs on a third party to save a life is derived from the verse “and your brother shall live with you” (Vayikra 36:36). Rabbi Akiva derives from that verse that “your life comes first” (Bava Metzia 62a), and R. Yisraeli understands this to mean that the obligation to save a life does not create an obligation to incur costs greater than those obligatory to fulfill any other positive commandment. R. Yisraeli then argues that this verse creates a universal principle, not just one among Jews, and that therefore one may not take a Gentile life to save a Jewish life.

I don’t know what drove R. Yisraeli to understand R. Akiva’s statement as universal, while, at the same time, limiting “whose blood is redder” to intra-Jewish comparisons. He does not provide any evidence for either position.

Bottom line: R. Yisraeli does not find a civilian-law justification for the raid. He rules that the mitzvah of saving a Jewish life cannot justify taking the life of a non-Jew who is not classified as a willing and responsible pursuer-enabler. This certainly excludes taking the lives of infants. However, he does think that the enabling-a-pursuer argument is sufficient to justify the killing of noncombatants who willingly support the pursuer and whose killing would save the pursued.

Finally, R. Yisraeli turns to the halakhot of war. 

Milhemet Reshut
Can an act of war be legitimate if it foreseeably leads to the deaths of noncombatant adults, innocent adults, and/or infants? And if so, can it be legitimate, still, if any of these were targeted?

R. Yisraeli approaches this issue from a conventional Maimonidean platform. According to Laws of Kings and their Wars 5:1, channeling the Mishnah (Sotah 9:7), halakhah recognizes two categories of legitimate war, mitzvah (obligatory) and reshut (permissible but not obligatory).

The prima facie problem with classifying any contemporary Israeli act as milhemet reshut is that a king must obtain the consent of the Sanhedrin in order to launch a milhemet reshut, and there is no Sanhedrin nowadays. R. Yisraeli does not see that argument as dispositive. His primary ground is that the Sanhedrin’s permission is necessary for a coercive draft, as in a monarchy, but that a democratically elected government may go to war without such permission because the soldiers are considered to be volunteers.

However, R. Yisraeli notes that the entire concept of milhemet reshut seems morally impossible: Why should it be permitted to kill people to gain political ends, especially if halakhah forbids imposing costs on others even to save a life? He answers that going to war constitutes a waiver of one’s right to life, conditional on the other side fighting in accordance with the rules accepted by both sides.[5]

There is much room to discuss how one determines consent, who has consented, and which rules were consented to. One might argue that if the permission to kill in war depends upon consent, then since it is impossible for infants to give such consent, infants cannot be targeted. Since R. Yisraeli does not make this argument, he probably thinks that national consent to war binds even infants.

Nonetheless, R. Yisraeli is clear that a milhemet reshut cannot justify killing except within the generally agreed rules of war. It seems to me that he holds this way even with regard to enemies who themselves do not obey the rules of war, so long as they acknowledge their disobedience as transgressive. Therefore, R. Yisraeli concludes that the Qibiyeh raid can be justified as an act of milhemet reshut only if it conformed to international norms of war. Rather than adjudicate that issue, he turns back to milhemet mitzvah.

Milhemet Mitzvah
Rambam includes wars with the purpose of ezrat Yisrael mi-yad tzar (aiding Israel from the hands of an attacker) in the mitzvah category. Does ezrat Yisrael mi-yad tzar include preemptive wars? Rambam, in his Commentary on the Mishnah, classifies at least some preemptive attacks as reshut rather than mitzvah.

The Mishnah (Sotah 9:7) records a dispute between an anonymous position (which the Talmud identifies as “the Sages”) and Rabbi Yehudah regarding the Torah’s exemptions from the battle line. The Sages say that the exemptions apply in a milhemet reshut but not in a milhemet mitzvah; R. Yehudah says that they apply in a milhemet mitzvah, (laudable, but not obligatory,[6] war) but not in a milhemet hovah (obligatory war). In Sotah 45b, Rabbi Yohanan contends that this dispute is purely semantic. Rava, however, states that there is a practical difference – whether the concept of oseik be-mitzvah patur min ha-mitzvah (one who is engaged in performing a mitzvah is exempt from other mitzvot) applies. However, he also narrows the dispute to one case: a war intended “to diminish the idolaters, lest they attack us.”

R. Yisraeli notes that there are many other practical halakhic differences between milhemet reshut and milhemet mitzvah. For example, as noted above, a king requires the consent of the Sanhedrin to draft for a milhemet reshut. Why, then, does Rava suggest that oseik be-mitzvah is the only practical difference between calling a war “to diminish the idolaters, lest they attack us” a mitzvah as opposed to a reshut?

R. Yisraeli’s answer sends us to Yerushalmi Sotah. Yerushalmi Sotah also first cites R. Yohanan, saying that the dispute between the Sages and R. Yehudah is purely semantic. Rav Hisda then greatly broadens the dispute. He understands the Sages to be thinking in terms of classical categories: “the wars of David (mitzvah) and the wars of Joshua (hovah),” whereas R. Yehudah is thinking in practical categories: “us attacking them (reshut) and them attacking us (mitzvah).”(Note: R. Hisda seems to reverse the attribution of the terms reshut and hovah in our mishnah; this is addressed by many commentaries, and R. Yisraeli’s explanation can most likely be made compatible with any of them.)

R. Yisraeli asks: Since we have already established that there is no obligation to risk one’s life to save someone else’s, why should a war to save someone else be obligatory? He answers that both of R. Hisda’s categories within R. Yehuda are wars intended to save lives. However, the obligation to save life can make the war a mitzvah but not a hovah, since one need not risk one’s life to save another. R. Yehudah nonetheless thinks that one engaged in such a war is exempt from other mitzvot, since the mitzvah of lifesaving, even when engaged in voluntarily, pushes aside other mitzvot, while the Sages disagree.

Why, then, does it matter which side attacks? Here, R. Yisraeli makes an astounding leap. He asserts that “when they attack us” transforms the war from a war of self-defense, which entails and fulfills the mitzvah of lifesaving, into a war of revenge, which is a mitzvah in and of itself, and does indeed obligate risking one’s life.

Where does the idea of a “war of revenge” come from? The obvious source is Bamidbar 31:2, where Hashem tells Moshe: “Avenge the vengeance of Benei Yisrael from the Midianites.”[7] 31:7 reports that Benei Yisrael marched against Midian as commanded by Moshe, and that they killed all of the males. Nonetheless, when the army returns in 31:14, Moshe is angry with them for having left non-virgin girls alive.

What justifies Moshe’s anger, when he had given no specific instructions? R. Yisraeli argues that Benei Yisrael applied the regular rules of milhemet mitzvah, whereas they should have known via reason that the rules for a war of vengeance are different and what those differences are.

One might expect R. Yisraeli to jump straight from there to justifying the Qibiyeh raid. But that would not be sufficient, as infant girls were killed. He may also have had moral qualms about justifying the targeting of any children.

Rather, R. Yisraeli makes the following argument:

The war against Midian in its details was exceptional, because those details were related specifically to the Midianesses’ seduction of Jewish men into idolatry at Ba’al Peor. The closest comparison is to the Ir Ha-Nidahat, the Law of the Idolatrous City among Jews. However, from the Midian war one can derive the principle that a war of vengeance is a milhemet mitzvah.

This suffices to make the Qibiyeh raid a milhemet mitzvah act. But it is not sufficient to justify the deaths of children in the raid. So R. Yisraeli goes one step further – but it is vital to note that he goes only one step further.

The cases of Ir Ha-Nidahat and Midian teach us that, under some circumstances, minors can be punished for the sins of their parents. While this principle is never applied by human courts, Jewish or Noahide, except in the narrow idolatry-related contexts above, these cases teach us that sometimes that is the ruling of the Heavenly Court.

If so, R. Yisraeli argues, the incidental death of children as collateral damage from ordinary acts of war aimed at military targets is not necessarily an injustice; it may instead represent the judgment of the Heavenly Court. Therefore, in a war of vengeance, one may engage in military activities without specific attention to whether the collateral damage will involve infants and children as well as adults. However, one still may not target children and certainly not infants, even if that targeting serves a military purpose, and certainly not in the name of revenge.

Evaluation
The details of the war against Midian – its jus in bello – make it a dangerous precedent for contemporary military ethics, as R. Yisraeli concedes. He nonetheless argues that we can derive a hitherto unrecognized category of milhemet mitzvah from its jus ad bellum (grounds for war). I find two arguments against this compelling.

a. In Devarim 2:9, God tells Moshe, “Do not attack Moab.” Bava Kamma 38a asks: Why is this instruction necessary? Isn’t the default not going to war? It answers that Moshe thought that he could derive via reason that Moab had behaved at least as badly toward the Jews as Midian had, and therefore deserved to be attacked. But his perspective was too narrow. Moab, as it was, indeed deserved to be attacked and destroyed. But Moab could not be attacked, lest that prevent the birth of extraordinary descendants.

I have argued publicly that this passage disproves R. Yisraeli. How can the war against Midian be used to justify any subsequent war, let alone create a category of obligatory war, if even Moshe Rabbeinu would have misapplied the precedent!?  

R. Yisraeli might reply that God accepted Moshe’s logic and merely stepped in to prevent this specific application. Regarding Moab (and Ammon), God had humanly-unknowable reasons for protecting them; the model of Midian applies whenever God doesn’t step in.

I reply that we do not live in an age of prophecy, let alone in the time of Moshe Rabbeinu. How are we to know if God has unknowable reasons in a case facing us?[8]

R .Yisraeli might reply that God would have built any further exceptions into the law.

b. The war against Midian is directly commanded by God in Bamidbar 25:16-18 and again in Bamidbar 31:1-2. Nonetheless, there is halakhic consensus that the war against Midian was a hora’at sha’ah (a onetime ruling). There is no halakhic imperative for Jews to fight Midianites nowadays. War against Midian is absent from the various lists of the 613 mitzvot.

R. Yisraeli argues, however, that only the application to Midian was a hora’at sha’ah. However, war with any nation (perhaps other than Moab, Ammon, and Edom) with regard to whom Israel has a similar jus ad bellum is a mitzvah, just as the war with Midian was.

However, I’m not aware of a rabbinic text or medieval authority who explicitly describes the war with Midian as a milhemet mitzvah.[9]

Moreover, Sifri (Matot 157) cites Rabbi Natan as deriving from the war with Midian that besieging Jews must leave the fourth side open, and by halakhic consensus that requirement applies only to a milhemet reshut. Here, for example, is Ramban, Forgotten Positive commands, 5: “This is not a mitzvah just for the hour in Midian; rather, it is a mitzvah for all future generations in every milhemet ha-reshut.”[10]  

So the evidence indicates that the war with Midian was a milhemet reshut. How can that be, since it was directly and explicitly commanded by God?[11]

My answer begins with R. Yisraeli’s suggestion that there are (at least) three halakhic categories of war. The third, which includes a preemptive attack, fulfills a mitzvah, but is classified as a milhemet reshut, except for the purposes of oseik be-mitzvah patur min ha-mitzvah. As I’ve argued in a previous essay,[12] Hazon Ish made a similar claim about a war that began with the Jews attacking but in which they are now on the defensive.

I therefore suggest that the war with Midian similarly remained a milhemet reshut even though it fulfilled a direct Divine command, and cannot serve as a paradigm for a previously unknown type of milhemet mitzvah.[13]

R. Yisraeli might respond that one can use the jus in bello of as a precedent for a milhemet reshut while using its jus ad bellum as a paradigm for a milhemet mitzvah. But this seems quite complicated, unjustified, and unnecessary.

I find many of R. Yisraeli’s other leaps and readings unpersuasive. Yet, the essay overall is certainly a work of genius. Moreover, it is a work of humanist genius in its commitment – which I don’t see as incidental – to human autonomy, even to the point of permitting suicide, but also to the point of valorizing heroic risk. 

At the same time, there are moments in the argument that seem perched on a tottering edifice over a vast abyss. R. Yisraeli’s understanding of “what have you seen that makes your blood redder” allows Jewishness to be a valid answer, which leads easily to devaluing or dehumanizing non-Jews; his category of “war of vengeance” can in principle legitimate genocide. In each case, he finds a way – I think not accidentally or fortuitously, but with benevolence and humanistic aforethought – to prevent the great fall. The argument of “what have you seen” is needed only when no prohibition is involved in killing, and when would that be? Our wars of vengeance are “not at the severity” of the original.

R. Yisraeli’s article makes brilliant, essential, and indelible contributions to the development of military halakhah and halakhic military ethics. I applaud the halakhic dexterity which enables him to land on his moral feet time and again. Yet many who follow in his footsteps are likely to stumble.[14] Therefore, contemporary poskim and ethicists must engage with his arguments and conclusions deeply, critically, and cautiously, as we seek to build a compelling and practical Torah framework for waging war.



[1] R. Shaul Yisraeli, “The Raid on Qibiyeh by the Light of Halakhah,” Hatorah ve-HaMedinah, vols. 5-6 (5713-5714): 70.

[2] Rabbi Francis Nataf suggests that I am overweighting the influence of Kantian ethics on contemporary ethics, and he may well be correct. If so, I will say in my own voice that universalizability and reversibility are rationally compelling starting points, but not absolute conclusions. Rabbi David Fried concurs with R. Nataf’s critique and adds, in his own voice, that law needs to be universalizable, but ethics need not be.

[3] Noam J. Zohar, Alternatives in Jewish Bioethics, (Albany: State University of New York Press, 1997),  1-14.

[4] As Rav Yisraeli notes, his argument and conclusion about non-Jews and suicide are largely taken from Minhat Hinukh 34.

[5] R. David Fried asks: “If we initiated the conflict, as with most cases of milhemet reshut, how can the other side be said to have consented?” I’m not certain how R. Yisraeli would answer. Possibly he thinks soldiers consent to the rules even though they’re forced to play the game. More likely, a general consent to the rules of war is expressed by the state on behalf of all of its citizens, which applies even to wars of defense against aggression. However, this rationale might not apply to wars that halakhah prohibits.

[6] Note that this is different from how milhemet mitzvah was translated above. This is consistent with the dual usage of the word mitzvah that occurs throughout the Talmud. When contrasted with reshut, mitzvah should be translated as “obligatory,” while when contrasted with hovah, it should be translated as “laudable, but not obligatory.”

[7] All translations are my own.

[8] R. David Fried asks why this principle doesn’t prevent all halakhic executions. My answer is that halakhah permits executions only for offenses specifically commanded by God; one cannot execute as the result of a kal va-homer (Sanhedrin 73a and others. See Encylopedia Talmudit s.v. “Ein onshin min ha-din”). I suggest that God’s reply to Moshe is best understood as an invocation of that principle.

[9] However, Amar Yosef to Rambam (Laws of Kings 7:4) describes it as a milhemet mitzvah on the basis of the biblical command, and Shabbat Shel Mi to Shabbat 64a cites Maharit as describing  it as a milhemet hovah.  

[10] The disagreement that impel’s Ramban is only about whether to count this as one of the 613 mitzvot. However, Rambam (Laws of Kings 4:7) is not explicit about the limitation.

[11] This question is asked by Minhat Hinukh (227:1). He notes that Ramban, cited above, clearly classifies it as a milhemet reshut, and leaves the question unanswered.

[12] https://moderntoraleadership.wordpress.com/2023/10/12/on-the-halakhic-status-of-wars-of-self-defense/

[13] On 16 Tammuz 5720, six years after publishing “The Raid on Kibiyeh,” Rav Yisraeli, in a shiur at Yeshivat Mercaz HaRav, addressed how the war with Midian can be called a milhemet reshut with regard to the law of sieges while at the same time being a milhemet mitzvah. He argued that wars which are purely collective obligations may be called reshut with regard to individuals. However, that shiur makes no mention of the “war of revenge” theory, and furthermore does not directly mention that theory’s core position that the war with Midian included an individual obligation. It is therefore  irrelevant to my objection here.

See R. Shaul Yisraeli, “Establishing the Concept of Milchemet Reshut and Milchemet Mitzvah,” in Lahai RO’I’: A Memorial Volume for the Brilliant Young Avraham Yitzhak Raanan-Kuk z”l (Hebrew), (Jerusalem: Mercaz HaRav, 5721).

Reprinted as Siman 14 of Rav Yisraeli’s Amud HaYemini (Hebrew).

[14] I must also note that R. Yisraeli’s moral and halakhic universe cannot be constructed out of this article alone. Rav Yitzchak Roness, writing in Tzohar 29, notes that R. Yisraeli’s essay on Entebbe constructs another original type of milhemet mitzvah, and in that context declares that kiddush Hashem is at the core of the overall category. R. Roness suggests that, according to R. Yisraeli, this should prevent a Jewish state from  violating international law, treaty commitments, and broadly accepted moral norms, even in the pursuit of a milhemet mitzvah, even during a “war of vengeance,” even if doing so would cost soldiers’ lives. R. Yisraeli’s son in law, Rav Avraham Yisrael Sharir, in Tzohar 30, pushed back hard at the last claim, and challenged R. Roness’ understanding of the category of hillul Hashem. I hope to treat their dialogue in the near future.