Sieges: Ancient Strategy, Modern Application


Ethicists have considered and written about the most ethical way in which to conduct war. Just war theory is the branch of philosophy that studies how to most ethically commit deeds that seem to be, at their core, unethical.  The protection of civilian life is the guiding principle of just war theory. These ethical codes of acceptable wartime conduct require that all attempts be made by military personnel to protect civilians from damage and death in times of war. The two philosophical principles governing this requirement are the principle of distinction and the doctrine of double effect. The principle of distinction requires the users of force to distinguish between the combatant and the noncombatant. The doctrine of double effect (DDE) permits an action with a primary positive end, despite certain harmful side effects.[i] When applied to wartime, the DDE allows for the death of civilians in the natural, violent course of war if it is in the process of targeting and “taking out” enemy forces. In modern warfare, however, these two moral guidelines and their interplay are challenging for the scholar to understand and agonizingly difficult for the soldier to apply.

 The concern for the protection of civilian life can also be found in the Jewish tradition. The Bible describes how to treat civilians in wartime, and Maimonides codifies military laws of conduct.[ii] Ranging from the eshet yefat to’ar (female captive taken in war)[iii] to the protection of fruit trees,[iv] the Halakhah records wartime legislation. Interestingly, though, a religious doctrine that is often hailed as moral seems to have no analogs to the principle of distinction and the DDE; this disconcerting quietude of  biblical, rabbinic, and medieval sources must be considered, especially in view of biblical accounts of the slaughter of civilians.[v] This textual reticence is difficult to understand philosophically, and creates a near-impossibility of determining how to act in a way that is based upon the Judaic tradition and yet also moral by a modern standard.  There is a gap between the Judaic and Classical war ethics that deserves examination.

Generally, it is difficult to compare Classical and biblical views of war ethics because of the gap of literature: While the Classical view has a healthy, developed history, the Judaic tradition has been stunted by the 2,000-year exile. This gap requires modern halakhists and ethicists to infer a Judaic just war ethic, and this inference is often shaky and inadequate. However, sieges are an excellent point of comparison between the Classical and Judaic war ethics, because they are one of the few war strategies that were used both in antiquity, as in Jericho according to the Bible,[vi] the siege of Masada, and today, as in the Taliban’s sieges of US outposts in Afghanistan.[vii] Because of its history, there is a large amount of literature and thought on the topic, including questions concerning its morality.

Classical Ethics of Sieges

  A siege is a very simple maneuver: When the enemy protects itself in a stronghold, the attacking army surrounds the enemy, thereby starving out the pent-up population. The siege will last until one side gives up or the fortifications are overcome. While modern weaponry, e.g., aerial bombing, has diminished the value of sieges as a military strategy from both the defensive and offensive perspectives, sieges are still used today.

Despite the tactical simplicity of laying a siege, this strategy is fraught with considerable moral ambiguity. More often than not, civilians are entrapped in the besieged city or stronghold, and because sieges often result in shortages and assaults upon the city, civilians are usually inadvertently killed in the process of attacking combatants.  This situation violates the principle of distinction, which mandates that civilians not be harmed alongside combatants. The question of the morality of sieges and the treatment of civilians, therefore, is troubling to many ethicists.

Michael Walzer, in his seminal book Just and Unjust Wars, describes sieges as “the oldest form of total war.”[viii] Total war is “military conflict in which the contenders are willing to make any sacrifice in lives and other resources to obtain a complete victory.”[ix] These sacrifices often include civilian lives and property which, in a non-total war situation, are considered to be innocent and illegitimate targets. Sieges are a form of total war, because, as Walzer writes, the ultimate purpose of the siege is not to starve the enemy into submission, but rather to starve the civilians, who in turn force the hand of the enemy government. Sieges are a form of total war because they intentionally target civilians.

The notion of Total War is old, but its denunciation is rather modern. For much of human history, total war was the only mode of warfare. Yet as ethics developed, total war has shifted from being seen as a given, to a tool requiring justification, to an outlawed practice. Total warfare has been banned by international law to protect civilians from the ravages of war, because, as noncombatants, they are deserving of protection. Basing his ideas off of the principle of distinction, Walzer equates sieges to total war, and denounces them. He argues that ethical conduct in war demands that the besieging army to open a path for civilian flight.

However, there are two flaws in Walzer’s analysis. First, a siege can be used not only as a means to put pressure on the army via civilian deaths, but also as a means of putting pressure directly upon the army. Although rare, if a besieged city has no civilian population, but is a fortress composed entirely of combatants, then a siege would not be a tool of total war.  For instance, if the Taliban were to besiege an American outpost in the Helmand region, there would be no ethical violation on the Taliban’s part, as there are no American civilians in that area.

 The second flaw in Walzer’s argument is his assumption that the death of civilians is the true intent of the besiegers. If, however, a fortress or city is strategically located, the intent of the attackers is not to destroy the opposing army, but rather to replace the opposing army with their own forces.  The death of combatants and noncombatants alike is an unintended, though welcome, byproduct. Walzer’s depiction of a siege is reminiscent of those sieges that resulted in bloody massacres, such as Masada, Troy, etc. Walzer, however, fails to acknowledge the possibility of a besieging army being justified by the DDE.

The DDE states that “it is sometimes permissible to bring about a harm as a merely foreseen side effect of an action aimed at some good end, even though it may have been impermissible to bring about the same harms as a means to that end.”[x] It is a philosophical principle that governs the ethics of medicine, business, and war. The DDE has its roots in medieval Christian theology, and is still used today. When applying the DDE to war ethics, four preconditions must be fulfilled in order to justify the death of noncombatants:

1) The act is good in itself or at least indifferent, which means, for our purposes, that it is a legitimate act of war.

2) The direct effect is morally acceptable – the destruction of military supplies, for example, or the killing of enemy soldiers.

3) The intention of the actor is good, that is, he aims only at the acceptable effect; the evil effect is not one of his ends, nor is it a means to his ends.

4) The good effect is sufficiently good to compensate for allowing the evil effect…[xi]

 Walzer argues that besieging areas that contain large civilian populations is forbidden, as it fails, at a minimum, to satisfy the third condition, and, more often, the fourth condition. While he succeeds in proving this assumption in specific cases (the British blockade of Germany, for example[xii]), he fails to prove it in general. There can be an instance, though it may be rare, when besieging a city that contains both combatants and noncombatants can be ethical by virtue of the DDE.

Thus, despite the questionable morality of besieging a city, and the possibility of it being an aspect of the abhorred total war, sieges are not outlawed by classical jus in bello ethics, because sieges can, in fact, fulfill the DDE and the principles of just war.  Nonetheless, such a justification for sieges is generally not used. Walzer observes that, far more often, leaders justify sieges by claiming that the noncombatants had consented to the defense of the city by combatants, and so brought the siege upon themselves. Walzer rejects this notion: “The siege itself is an act of coercion, a violation of the status quo, and I cannot see how the commanders of the besieging army can escape responsibility for its effects. [The commander] has no right to wage total war, even if civilians and soldiers within the city are politically united in refusing surrender.”[xiii] As such, writes Walzer, the besieger has the moral obligation to allow noncombatants to escape the city.

Although Walzer is considered an expert on just war theory, his opinion on sieges is not universally accepted. Richard A. Gabriel thinks that sieges actually place the onus of noncombatant deaths upon the defenders of the city. While normal conditions of war protect civilians, sieges are abnormal, and the treatment of civilians therefore changes. He notes that sieges inevitably involve the citizens, from the perspectives of both the besieger and the besieged.  He writes, “Civilians are involved not only in the tactics of the besieger but in the tactics of the defender, who, if he is to be successful, must marshal every available civilian to keep the city running and defensible.”[xiv] When combatants insert themselves into civilian areas, the defenders of the city “change such areas, which then become fair game; the civilian population with them also become subject to military action.”[xv] Gabriel asserts that the protection of civilians is incumbent upon the defensive force, not the offensive.

In any event, it is clear that the Classical view of sieges is wholly dependent upon the principle of distinction, which requires a differentiation between combatants and noncombatants, and the DDE, which requires that the death of noncombatants be justified. It is the interpretation and application of these principles that cause the different conclusions. Yet the Judaic view of sieges is entirely different, not only because of the reticence in Judaic texts on such philosophical principles, but also because the Judaic tradition has entirely different principles governing sieges.

Sieges in Judaic Sources

 While the Classical just war ethic is rich in sources and nuance, the Judaic just war ethic has been stunted by virtue of the Diaspora and the army-less situation of the Jewish People.  To construct a Jewish just war theory, many modern scholars attempt to extrapolate preconceived Classical just war ethics from the ethos of Halakhah and biblical stories.

While there is no direct mention of a principle similar to the distinction doctrine in biblical or talmudic sources, R. J. David Bleich indirectly derives it from the principle of rodef [pursuer with an intent to kill]. In a case where an individual is being pursued by a potential murderer, and the only way to halt the murderer is to indirectly kill an innocent bystander, no action may be taken, for, Bleich explains, “Since the law of pursuit is designed to preserve the life of the innocent victim, it is only logical that it is forbidden to cause the death of a bystander in the process since to do so would only entail the loss of another innocent life.”[xvi] As such, writes Bleich, the lives of civilians must be protected in wartime.[xvii]   Yet this theory is derived, so the link between the biblical and halakhic sources and the principle of distinction remains weak.

Bleich also attempts to find a source for the DDE, though in less protracted form. According to Ha’amek Davar to Genesis 9:5, homicide is not punishable during wartime, regardless of the victim’s status as a civilian or military actor. Though this would seem to allow for the intentional targeting of civilians, Bleich re-interprets and limits the Ha’amek Davar. It is “logical to assume that military action leading to civilian casualties may be regarded as legitimate… only [emphasis added] when the loss of civilian life is incidental to military purposes, but not when wantonly undertaken as an end in itself.”[xviii] Thus, there is the preexisting, though stretched, notion of the DDE in the Judaic tradition. While these inferred ethics are less nuanced than the Classical versions, that is to be expected, as they are only inferences. These modern ideas, such as the principle of distinction and the DDE, came into existence long after the Judaic war ethic stopped developing. Yet it is not these philosophical principles that truly govern the Judaic approach to sieges.

 By virtue of their ancient use, sieges and their morality are mentioned explicitly in the Judaic tradition. Numbers 31:7 records, “And they warred against Midian, as the LORD commanded Moses; and they slew every male.”[xix] Sifrei derives from this verse that the Israelite army “gave them a fourth side to enable them to flee:” “And they went to war on Midian, and surrounded [the city] from four sides. R. Natan says, ‘he gave them a fourth side to enable them to flee.’”[xx]  In line with the biblical text that notes that the Israelites killed every male, Sifrei makes no distinction between combatant and noncombatant.  Sifrei’s qualification here, though, is odd, as the Bible records the Israelites’ treatment of the Midianites; they killed all the males and leaders, took the women and children captive, claimed the property, and burnt the cities.[xxi] Why, then, does Sifrei record the gesture of leaving an escape route, similar to Walzer’s advice, when the treatment of the Midianites is more akin to total warfare?

 Based on Sifrei, Maimonides establishes siege law. Maimonides records in Mishneh Torah that when besieging a city, one side must be left open to allow people to flee.  He writes, “When a siege is placed around a city to conquer it, it should not be surrounded on all four sides, only on three. A place should be left for the inhabitants to flee and for all those who desire, to escape with their lives, as it is written (Numbers 31:7): ‘And they besieged Midian as God commanded Moses.’”[xxii] This law, however, applies only to a milhemet reshut, or sanctioned optional war, not to a milhemet mitsvah, or mandatory war. What is interesting is that Maimonides does not differentiate between combatants and noncombatants; his law applies to both, with no principle of distinction determining who is permitted to flee.

When discussing milhemet reshut, Maimonides records that it is forbidden to kill women and children: “If they do not make peace, or if they make peace, but refuse to accept the seven mitsvot, war should be waged against them.  All males [past majority] should be killed. Their money and their children should be taken as spoil, but neither women nor children should be killed, as is stated (Deuteronomy 20:14): ‘But the women and the children…take as spoil.’” “The children” refer to males below the age of majority.”[xxiii] Yet for a milhemet mitsvah, defined by Maimonides here as a defensive war or a war against the seven Canaanite nations or Amalek,[xxiv] one is obligated to leave no one alive. Thus, milhemet mitsvah follows the rules of total war, for which all weapons, tactics, and objectives are fair game, including the intentional targeting of civilians, while milhemet reshut is similar to limited war, for which the weapons, tactics, and objectives are limited, including the intentional targeting of civilians. Therefore, according to Maimonides, the principle of distinction depends upon the type of war: While the law is merciless for a milhemet mitsvah, it is merciful above and beyond the Classical tradition’s formulation in a time of milhemet reshut.

This reading of Maimonides and the presumed mercy of Halakhah falls apart when other texts and the interpretation of Maimonides are examined. While Maimonides codifies in Mishneh Torah that one must leave open a side when sieging a city, he does not include this precept in his Sefer ha-Mitsvot.  Based on this, Meshekh Hokhmah explains that the assertion of Maimonides is only advice, a military tactic that can be discarded by a military leader at will. The rationale for this advice, writes Meshekh Hokhmah, is that the swiftest way to conquer a city is to let the combatants flee, allowing the besiegers to take the city. R. Yitzchak Blau explains the rationale of the Meshekh Hokhmah: “If the enemy feels that it has no escape route, it will redouble its fighting efforts. If it has an escape route, soldiers will run and the rest will lose fighting spirit. As this merely reflects a wartime strategy rather than a religious ideal, it does not merit being counted as a separate mitsva.”[xxv] Thus, this interpretation of Maimonides transforms a moral, humanitarian halakhah into tactical military advice. If a general thinks that allowing combatants to flee will undermine military objectives, there is then no need to allow for flight. Women and children, the medieval equivalent of noncombatants, are still protected, but on an equal level to the principle of distinction. Overall, then, the Maimonidean perspective on sieges during milhemet reshut is similar to the Classical tradition.

Sifrei to Numbers 31:7, the source from which Maimonides derives his military tactic, is, however, explained in more than one manner. Naḥmanides and Sefer ha-Hinnukh also contemplate the Sifrei, and arrive at different conclusions from Maimonides’.  Naḥmanides writes that, in every milhemet reshut, the besieging army must leave one side of the city open for all people within the city, fighter and civilian, to flee. The rationale is both to inculcate the Jewish army with the attribute of mercy and, as a war tactic, in order to prevent the enemy from fighting back.[xxvi] Sefer ha-Hinnukh similarly limits the commandment explicitly to a context of milhemet reshut.[xxvii]

In contrast to Maimonides who applies Sifrei only to a case of milhemet mitsvah, these medieval halakhists believe that Sifrei is only speaking of a milhemet reshut, a sanctioned war. This alternate interpretation fits with Maimonides’ idea of equating milhemet mitsvah to total war and milkhemet reshut to limited war.  The proposed escape path is open to all individuals, not limited to noncombatants.  The reason that individuals are allowed to flee in a sanctioned war is out of humanitarian concern for the people in the city; the Jewish besiegers are commanded to have compassion for their enemy. The tactical advantage that the Israelites may gain in this strategy is a welcome byproduct, but not the goal of the exercise. Even if the tactical advantage would cease to exist, writes Naḥmanides, the Jews are still required to leave open a side to allow for combatants and noncombatants to flee. The Naḥmanidean perspective is far more humane than the Classical one.

The ambiguity of the Halakhah’s and the Bible’s views of morality is also discussed in modern times. The 1982 Israeli siege of Beirut prompted a slew of writings and debates upon the halakhic propriety of this military action.

Two Traditions in Action:  Classical War Ethic Versus Judaic War Ethic

In the 1982 Lebanon War, the IDF besieged Beirut in order to capture PLO terrorists. For thirty-three days, the IDF closed the city and controlled access to food, water, and fuel.[xxviii] Between 4,000 and 5,000 civilians died from military action during the siege. Yet the IDF did not wholly surround the city.  Similar to Walzer’s formulation that the besiegers must allow for civilian flight, “[t]hroughout the siege, the IDF kept open two major escape routes from the city to Syrian positions. Of the 500,000 people trapped in West Beirut, about 100,000 took advantage of the Israeli escape routes and did leave.”[xxix] Only civilians were allowed to flee. This method adheres to the jus in bello principles propounded by Walzer and goes far beyond Gabriel’s more lax formulation, but disregards  Maimonides’ and Naḥmanides’ requirement to allow combatants to flee as well.

Interestingly, Walzer labels Maimonides’ formulation of the war strategy as

hopelessly naive. How is it possible to “surround” a city on three sides? Such a sentence, it might be said, could only appear in the literature of a people who had neither a state nor an army of their own. It is an argument offered not from any military perspective, but from a refugee perspective.[xxx]

Walzer views Maimonides’ war law as antiquated and irrelevant. The IDF did not follow Maimonides’ (or Naḥmanides’) demand that combatants be permitted to flee the besieged city; they did not allow PLO combatants to flee. Only noncombatants were permitted to use the two escape routes.[xxxi] Despite the prompting of many religious figures to act in a more halakhic manner, the IDF continued to act in accordance with jus in bello rules. The IDF’s actions were more akin to Classical than Judaic ethics.[xxxii]


Although the Judaic war ethic has a gap of 2,000 years in its development, its growth has been healthy. The IDF and Israeli ethicists have an abundance of material and literature from which to derive a war ethic. The sources of the Judaic ethic developed prior to the Classical tradition, and the early interpretation of these sources occurred simultaneously with the development of the Classical tradition.

Yet no matter the interpretation of Sifrei (the original source for the Judaic ethic of sieges), the Judaic tradition is in direct contrast to the Classical. Neither Walzer nor Gabriel is bothered by the death of combatants in a besieged city; it is the death of noncombatants that bothers them, though they lay the responsibility for these deaths at the feet of different actors.  The Judaic tradition, on the other hand, does not adhere to the principle of distinction in the case of sieges, and instead holds that both combatants and noncombatants have the right to flee. It is interesting that the verses following Numbers 31:7 and the war with Midian record the slaughter and enslavement of noncombatants, as well as the pillaging and burning of cities, which would be considered war crimes in modern times; still, the Naḥmanidean interpretation of the verse shows more concern for human life than any Classical ethic.

The Maimonidean formulation, at least the standard interpretation of it, is a war strategy, not a systematic, logical ethic like the just war ethic. Meanwhile, the Naḥmanidean ethic is based upon the law’s desire to keep the soldiers merciful. Both perspectives are not based upon a strict, rational ethic in the way that jus in bello’s rationales are based on the philosophical principles of distinction and the DDE. This is not to say that one tradition is less valuable or ethical than the other, but merely that one is more emotional and the other more rational.

Yet the IDF and military ethicists have shown an unambiguous preference for the Classical war ethic. There are many possible explanations of this choice, but the tradition the IDF has chosen is clearly the secular one. Such a choice is only possible to discover via examining a war strategy that was employed both in ancient and modern times, such as sieges.

            Ariella Gottesman is a senior at SCW majoring in Political Scienc

[i] Allison McIntyre, “Doing Away with the Double Effect,” Ethics 111.2 (2001): 219-255, at p. 219.

[ii] For example, Mishneh Torah, Hilkhot Melakhim, 6.

[iii] Deuteronomy 21:10.

[iv] Deuteronomy 20:19-20.

[v] For examples, see Numbers 31:1-12 and I Sam 15:7.

[vi] Joshua 6:1-27.

[vii] For example, “Deputy Commander reflects on a year in Helmand.” Defence News (28 Feb 11), available at:

[viii] Michael Walzer, Just and Unjust Wars (New York: Basic Books, 1977), 160.

[ix] “Total war.” Encyclopædia Britannica Online Academic Edition (2011), available at:

[x] McIntyre 219.

[xi] Walzer 153.

[xii] Walzer 172- 175.

[xiii] Walzer 168.

[xiv] Richard A. Gabriel, Operation Peace for Galilee: The Israeli-PLO War in Lebanon (New York: Hill and Wang, 1984), 134.

[xv] Ibid. 135.

[xvi] J. David Bleich, “Preemptive War in Jewish Law,” Tradition 21.1 (1983): 3-41, at p. 19.

[xvii] Bleich limits this only to a defensive war. He chooses to include a defensive war among the milhemot mitsvah. In the majority of these types of wars, the killing of civilians is not only sanctioned, but required.

[xviii] Bleich 28.

[xix] Mechon-Mamre’s translation.

[xx] Author’s translation.

[xxi] Numbers 31: 9-11.

[xxii] Mishneh Torah, Hilkhot Melakhim 6:7.  Touger translation.

[xxiii] Ibid. 6:4. Translation adapted from Touger.

[xxiv] For an understanding of the differences between defensive and preemptive wars in the Judaic tradition, see Bleich.

[xxv] Yitzchak Blau, “Biblical Narratives and the Status of Enemy Civilians in Wartime,” Tradition 39.4 (2006): 8-28, at p. 17-18.

[xxvi] Ramban‘s Emendations to Sefer ha-Mitsvot, Shikehat ha-Asin (“forgotten” positive commands).

[xxvii] Mitsvah # 527.

[xxviii] “FM 3-06, Appendix A, Siege of Beirut.” Available at:

[xxix] Gabriel 137.

[xxx] Walzer 168.

[xxxi] This assumes, of course, that the 1982 Lebanon war is a milhemet reshut. For more information, see: Gil Student, “The Siege of Beirut,” Hirhurim. Available at:

[xxxii] The Siege of Beirut prompted a slew of halakhic discussion on whether the siege was halakhically permissible. While the disagreement between Chief Rabbi of Israel R. Shlomo Goren and R. Shaul Yisraeli is interesting, their argument hinges primarily on whether the Lebanon War was a milhemet mitsvah or a milhemet reshut, which is irrelevant to this discussion. For more information, see Student.