In December 1837 British military forces based in Canada learned that a private American ship, the Caroline, was ferrying arms, recruits, and supplies from Buffalo, New York, to a group of anti-British rebels on Navy Island on the Canadian side of the border. On the night of December 29, British and Canadian forces together set out to the island to destroy the ship. They did not find the Caroline berthed there, but they tracked it down in United States waters. While most of the crew slept, the troops boarded the ship, attacked the crew and passengers, and set it on fire. They then towed and released theCaroline into the current headed toward Niagara Falls, where it broke up and sank. Most on board escaped, but one man was apparently executed and several others remained unaccounted for and presumed dead.

In a letter to Secretary of State Daniel Webster, British ambassador Henry Fox defended the incursion into U.S. territory and raid on the Caroline. British forces were simply acting in self-defense, he said, and protecting themselves against “unprovoked attack”1 with preemptive force. In his eloquent reply to Fox, Webster rejected the British argument and articulated a set of demanding criteria for acting with a “necessity of self-defense”—in particular for a legitimate use of preemptive force. Preemption, Webster said, is justified only in response to an imminent threat; moreover, the force must be necessary for self-defense and can be deployed only after nonlethal measures and attempts to dissuade the adversary from acting had failed. Furthermore, a preemptive attack must be limited to dealing with the immediate threat and must discriminate between armed and unarmed, innocent and guilty. The British attack on the Caroline failed miserably by these standards:

It will be for that Government [the British] to show a necessity of self-defence, instant, overwhelming, leaving no choice of means, and no moment for deliberation. It will be for it to show, also, that the local authorities of Canada,—even supposing the necessity of the moment authorized them to enter the territories of the United States at all,—did nothing unreasonable or excessive; since the act, justified by the necessity of self-defense, must be limited by that necessity, and kept clearly within it. It must be shown that admonition or remonstrance to the persons on board the “Caroline” was impracticable, or would have been unavailing; it must be shown that daylight could not be waited for; that there could be no attempt at discrimination between the innocent and the guilty; that it would not have been enough to seize and detain the vessel; but that there was a necessity, present and inevitable, for attacking her in the darkness of night, while moored to the shore, and while unarmed men were asleep on board, killing some and wound[ing] others, and then drawing her into the current above the cataract, setting her on fire, and, careless to know whether there might not be in her the innocent with the guilty, or the living with the dead, committing her to a fate which fills the imagination with horror. A necessity for all this the government of the United States cannot believe to have existed.

Webster concluded that “if such things [as the attack on the Caroline] be allowed to occur, they must lead to bloody and exasperated war.”2

In September 2002 the Bush administration announced a fundamental shift in the official American “national security strategy.” The new strategy relies heavily on the preemptive use of force, and in defending it National Security Adviser Condoleeza Rice referred to Daniel Webster’s “famous defense of anticipatory self-defense.”3 But Rice missed Webster’s point. Webster sought precisely to limit the resort to preemption, even in the name of self-defense. Preemption, after all, initiates violent conflict, so it must meet demanding strictures. By drawing a sharp line between legitimate preemption and illegitimate aggression Webster sought to avoid “bloody and exasperated war.”

New World, New Doctrine?

The new Bush security strategy contrasts sharply with the official Cold War strategy of deterrence. The old idea was to protect the country by telling opponents—particularly the Soviet Union—that any attack would be met with devastating retaliation, and by building military forces sufficient to make the threat of retaliation credible. The new strategy is not so much to deter threats as to preempt them, to nip them in the bud, to “act against such emerging threats [from “our enemies”] before they are fully formed.”4 “Our best defense,” in short, “is a good offense.”5 But the new doctrine goes well beyond what might be considered justified preemption; rather, it is a preventive offensive war strategy.

This shift in strategy emerged soon after September 11. In October 2001, Secretary of Defense Donald Rumsfeld said, “There is no question but that the United States of America has every right, as every country does, of self-defense, and the problem with terrorism is that there is no way to defend against the terrorists at every place and every time against every conceivable technique. Therefore, the only way to deal with the terrorist network is to take the battle to them. That is in fact what we’re doing. That is in effect self-defense of a preemptive nature.”6

The U.S. National Security Strategy released in September 2002 draws out more fully the case for preemption. The argument begins from the permissibility of preemption for the sake of self-defense:

“For centuries, international law recognized that nations need not suffer an attack before they can lawfully take action to defend themselves against forces that present an imminent danger of attack. Legal scholars and international jurists often conditioned the legitimacy of preemption on the existence of an imminent threat—most often a visible mobilization of armies, navies, and air forces preparing to attack.”7

Greater reliance on preemption, in a wider range of circumstances, is now necessary, the administration argues, because the nature of war has changed: “It’s a different world,” according to Colin Powell. “It’s a new kind of threat.”8

The world is different in particular, the administration argues, because “terrorists” seek “martyrdom” and leaders of “rogue states” are often risk-prone and willing to sacrifice the lives of their people; because preparations to attack the U.S. will often not be visible (they may use “weapons of mass destruction” that “can be easily concealed, delivered covertly, and used without warning”); and because attacks may be devastating. For these reasons we need to revise our understanding of when a threat is “imminent”: “We must adapt the concept of imminent threat to the capabilities and objectives of today’s adversaries.”9 The U.S. cannot wait for a “smoking gun” if it comes in the form of a mushroom cloud. Therefore, “The greater the threat, the greater is the risk of inaction—and the more compelling the case for taking anticipatory action to defend ourselves, even if uncertainty remains as to the time and place of the enemy’s attack. To forestall or prevent such hostile acts by our adversaries, the United States will if necessary, act preemptively.”10

I do not dispute the administration’s moral premise: that the right to self-defense sometimes permits preemption. But even in the new environment, distinctions between short- and long-term threats and between different sorts of potential adversaries remain fundamental. Denying the importance of these distinctions, as the administration does, is morally unacceptable and will lead to greater instability.


The distinction between immediate threats and long-term potential threats underpins the classical distinction between preemption and preventive war. Although preemption is not mentioned in the United Nations Charter, where only self-defense in the case of attack is allowed under Article 51, preemption has historically been considered a particular kind of self-defense against immediate threats.

Conventional just war theory proposes standards for legitimate preemption close to those that Webster argued for in 1837. That theory has two elements.Jus ad bellum criteria describe conditions for legitimately undertaking a war: the cause must be self-defense, war must be a last resort and necessary in the sense that no other methods would work, the attack must be proportionate, and the war must have a chance of success. In such cases, Michael Walzer argues, “states may use military force in the face of threats of war, whenever the failure to do so would seriously risk their territorial integrity or political independence.”11 Jus in bello criteria concern the legitimate conduct of war, and include injunctions of proportionality and discrimination between combatants and noncombatants, where noncombatants are not legitimate targets.

Preemptive war is directed against an opponent who has not yet attacked or harmed anyone. So, it is a grave step and should only be undertaken if it is both prudent and morally justified. Building on Webster and just war theory, I argue that a legitimate preemptive use of force must meet four conditions:

1. Self to Be Defended Narrowly Defined

The party contemplating preemption should have a narrow conception of the “self” to be defended. On the face of it the self-defense criterion seems clear. When our lives are threatened we must be able to defend ourselves, using force if necessary. But self-defense could come to have a thicker sense, that our “self” is expressed not only by mere existence but also by our free and prosperous existence. For example, even if a tyrant would allow us to live, but not under institutions of our own choosing, we may justly fight to free ourselves from political oppression. But how far do the rights of the self extend? What values may actors legitimately defend with military force? If someone threatens our access to food, fuel, or shelter, can we use force? Or, if they allow us access to the material goods necessary for our existence but charge such a high price that we must make a terrible choice—between food and health care, or between mere existence and growth—are we justified in using force to secure access to a good that would enhance the self? With economic interests and vulnerabilities understood to be global, and the moral and political community of democracy and human rights defined more broadly than ever before, the self-conception of great powers expands. But a broad conception of self is not obviously legitimate, nor are the values to be defended apparent. When the self is defined too expansively, too many interests become vital. But war itself—and certainly preemption—is not justified to protect imperial interests or assets taken in a war of aggression.

The United States has increasingly defined its “self” in broad terms. According to the most recent Report of the Quadrennial Defense Review, the “enduring national interests” of the United States which are to be secured by force if necessary include “contributing to economic well-being” which itself includes “vitality and productivity of the global economy” and “access to key markets and strategic resources.”12 Further, the goal of U.S. strategy is to maintain “preeminence.” As the president said at West Point, “America has, and intends to keep, military strengths beyond challenge. . . .”13 TheNational Security Strategy also fuses ambitious political and economic goals with security: “The U.S. national security strategy will be based on a distinctly American internationalism that reflects the fusion of our values and our national interests. The aim of this strategy is to help make the world not just safer but better.”14 And, perhaps most strikingly, the administration claims that “[t]oday the distinction between domestic and foreign affairs is diminishing.”15 But if the self is defined very broadly and threats to this greater “self” are met with military force, self-defense will look, at least to outside observers, like aggression.16

2. Justified Fear of Imminent Attack

To justify preemption there must be strong evidence that war is inevitable and likely in the immediate future. Immediate threats are those which can be made manifest within hours or weeks unless action is taken to thwart an attack. This requires clear intelligence showing that a potential aggressor has both the capability and intention to do harm in the near future. Capability alone is not a justification.17

As Michael Walzer argued persuasively in Just and Unjust Wars, simple fear cannot be the only criterion for launching a preemptive attack. Fear, already omnipresent in world politics, increases in the context of a terrorist campaign. The nature of fear in the wake of a devastating assault may be that a government and people will, justifiably, be vigilant. Indeed, they may out of this heightened fear be hypervigilant about threats—seeing small threats as large and brutally squashing potential threats. The fearful may then overreact to threats that do not risk the territorial integrity or political independence of a state. Or, the threat of “uncertainty” may trigger preemptive attacks.

The threshold for credible fear is necessarily lower, then, in the context of contemporary counterterror war, in which terrorists have the advantage of surprise. But the consequences of lowering the threshold of fear may be increased instability and the premature use of force. If simple fear justifies preemption, then preemption will have no limits since, according to the Bush administration’s own arguments, we cannot always know with certainty what the other side has and where it might be located or when it might be used. And if fear of a surprise attack was once clearly justified, when and how will we know that a threat has been significantly reduced or eliminated?

If simple fear does not suffice, then how much of what kind of fear justifies preemption? We need to tread a fine line. The threshold of evidence and warning cannot be too low: simple apprehension that a potential adversary might be out there somewhere and may be acquiring the means to do harm cannot trigger the offensive use of force. This is not preemption but paranoid aggression, and it promises endless war. We must—stressful as it may be psychologically—accept some vulnerability and uncertainty. We must also avoid the tendency to exaggerate threats and inadvertently heighten our own fear. For example, though nuclear weapons are more widely available than in the past, as are delivery vehicles of long range, nuclear weapons and long-range delivery vehicles are not yet in the hands of hundreds or even dozens of terrorists. A policy that assumes such a dangerous world is, at this point, paranoid. Rather than assuming that we live in such a world, or soon will, we should work to make this outcome less likely.

On the other hand, the threshold of evidence and warning for justified fear cannot be so high that those who might be about to do harm get so far along in their preparations that they cannot be stopped or the damage limited. Assuming a substantial investment in intelligence gathering, assessment, and understanding of potential adversaries, what we need is a policy that both maximizes our understanding of the capabilities and intentions of possible adversaries and simultaneously minimizes our physical vulnerability. While uncertainty—about intentions, capabilities, and risk—can never be eliminated, it can be reduced.

Aggressive intent coupled with a capacity to do immediate harm is the right threshold for legitimate preemption. We should judge intent by considering two questions: first, have potential aggressors harmed us in the recent past or said they want to harm us in the near future? And second, are potential adversaries moving their forces into position to do significant harm? While it might be tempting to assume that secrecy on the part of a potential adversary is a sure sign of aggressive intentions, it may simply be a desire to prepare a deterrent force that might itself be the target of a preventive offensive strike. For example, consider the September 11th attacks. On these criteria it would have been entirely permissible to arrest the hijackers of the four aircraft used as weapons. But prior to September 11, taking the war to Afghanistan to attack al Qaeda camps or the Taliban would not have been justified preemption unless it was clear that such action could have thwarted imminent terrorist attacks.

3. Preemption Likely to Succeed

Justified preemption must be likely to succeed in reducing or eliminating the threat. Specifically, there should be a high likelihood that the source of the military threat can be found and that the damage it was about to do could be greatly reduced or eliminated by a preemptive attack. If preemption is likely to fail on either of those counts it should not be undertaken. The prosecution of a successful counterterror war is very difficult. Terrorist operatives are hard to find because they are generally few in number, mostly inactive and concealed, and tend to be co-located with civilians. Preemption may be easier against states simply because the preparations of governments, even for surprise attacks, tend to involve larger-scale mobilizations that often are more visible.

4. Military Force Required

For military preemption to be justified, force must be necessary. There must be no time for other measures to work, or those other measures must be unlikely to avert a devastating attack, the preparations for which are already underway. If an attack can be thwarted by arresting a potential terrorist, for example, then military strikes should not be used, even if they would also be effective. The requirement that military force be necessary thus puts the onus on defenders to work to resolve conflicts with potential adversaries.

Even if these four criteria for undertaking a justified preemptive strike are met, the use of preemptive force must also meet jus in bello criteria of proportionality and discrimination. Specifically, the damage caused by the preemptive strikes should not exceed what was put at risk by the attack that is preempted. Thus, when the administration suggests that nuclear weapons might be used to preempt the acquisition of chemical and biological weapons, it proposes a disproportionate response.18

Preemptive action should also avoid killing innocents and using measures that harm the prospects of future peace. As Immanuel Kant argued more than two hundred years ago, “Some level of trust in the enemy’s way of thinking must be preserved, even in the midst of war, for otherwise no peace can ever be concluded and the hostilities would become a war of extermination.”19Discrimination is extremely difficult in the case of counterterrorist preemption because terrorists do not live in separate garrisons as regular armies do, but among civilian populations who may be unware of their presence. Civilian deaths are thus both unavoidable and foreseeable even if the preemptive strike involves the use of precision guided weapons or commando raids.

Preemptive actions must also have limited military objectives, and the preemptive action should cease when the threat is eliminated or significantly reduced. A legitimate preemptive motive does not give license to actions that go beyond reducing or eliminating an immediate threat.

Preventive Wars?

If preemption may sometimes be legitimate, is the Bush administration right to extend the case of justified preemption to preventive offensive wars? If all threats are considered imminent and unavoidable without the use of force, then yes. But although war has been transformed along many of the lines the administration suggests, not all threats are immediate and unavoidable.

The threat posed by terrorism is significant. Unconventional adversaries, prepared to wage unconventional “asymmetric” war, can conceal their movements, weapons, and immediate intentions and may conduct devastating surprise attacks.20 Nuclear, chemical, and biological weapons, though not widely held, are more readily available than they were in the recent past. And of course the “everyday” infrastructure of the United States can be turned against it, as were the hijacked planes on September 11. Terrorists in particular are extremely flexible. Unlike conventional militaries they can project power with great efficiency: they do not have to develop weapons and delivery vehicles; they may live among their target populations; and they require comparatively little in the way of logistical support. It is also true that although physical risk to terrorism could certainly be reduced in many ways, as Rumsfeld acknowledges, it is impossible to achieve complete invulnerability.21 And though the United States was open to serious threats in the past, Americans are perhaps more emotionally aware of that exposure today since, as Condoleeza Rice says, “9/11 crystallized our vulnerability.”22In sum, when combined with the advantage of surprise terrorism is a formidable military strategy which costs many times more to defend against than costs terrorists to conduct.

On the other hand, terrorists do not hold all the cards. For example, their sources of funding, often tied to illicit transactions and black market economies, are vulnerable to disruption through determined law enforcement. And while terrorists can piggyback on the infrastructure of their targets, they are also vulnerable to detection via that same infrastructure as they use phones, faxes, the Internet, and other electronic media. Finally, although there are far too many leaks in the containment of technologies and weapons of mass destruction, many of those weapons are still relatively expensive to acquire and difficult to produce in any quantity. Still, if we imagine all possible scenarios, the potential for devastation seems limitless and it ostensibly makes great sense to get them before they get us.

Perhaps for this reason Bush administration defense planners made a shift, even prior to September 11, from basing military planning on intentions and likely threats to what they call a “capabilities-based approach,” which attempts to “anticipate the capabilities that an adversary might employ” and “focuses more on how an adversary might fight than who the adversary might be and where war might occur.”23 Indeed, the scenarios proliferate according to General Ralph Eberhart, in charge of the military’s role in homeland security: “The list goes on and on. We can all envision the terrible things that might happen.”24

But in estimating potential threats, the intentions of a likely adversary are much more important than capabilities that “might” be employed by someone. So the assertion that the United States faces rogue enemies who “hate everything” about it must be carefully evaluated. While there is certainly compelling evidence that al Qaeda members desire to harm the United States and American citizens, the National Security Strategy makes a questionable leap when it assumes that “rogue states” also desire to harm the United States and pose an imminent military threat. Moreover, the administration blurs the distinction between “rogue states” and terrorists and essentially erases the difference between terrorists and those states where they reside: “We make no distinction between terrorists and those who knowingly harbor or provide aid to them.”25 But these distinctions make a difference when a country is deciding whether to initiate the use of force.

Current U.S. policy does not, however, respect these distinctions. As President Bush said at West Point: “We must take the battle to the enemy, disrupt his plans and confront the worst threats before they emerge. . . . Our security will require . . . a military that must be ready to strike at a moment’s notice in any dark corner of the world. And our security will require all Americans to be forward-looking and resolute, to be ready for preemptive action when necessary to defend our liberty and to defend our lives.” Indeed, since September 11 and the administration’s gradual articulation of its new security doctrine, the United States has sent troops to fight terrorists not only in Afghanistan but in the Philippines, Yemen, Indonesia, and former Soviet Georgia and has confronted both Iraq and North Korea over the issue of weapons of mass destruction. A limitless preventive offensive doctrine thus entails an expanding list of force commitments that might spread U.S. military forces thin at the same time that it risks escalating military conflicts. Such uses of force, while seeming sensible in an atmosphere of perceived heightened vulnerability, may at best be unnecessary. At worst there is risk of backlash fueled by fear and resentment, not least because discrimination between combatants is difficult in preventive war. Indeed, because they have not yet made an aggressive act nearly all those killed or injured by a preventive war will be noncombatants.

To see how the administration has blurred the distinctions between preemption and preventive offensive war, consider their arguments about the threat posed by Iraq’s potential to acquire weapons of mass destruction. Vice President Cheney argues that Iraq poses a threat to the United States: “Many of us are convinced that Saddam Hussein will acquire nuclear weapons fairly soon. . . . Deliverable weapons of mass destruction in the hands of a terror network or murderous dictator or the two working together constitutes as grave a threat as can be imagined. The risks of inaction are far greater than the risks of action.”26 But here we must recall the distinction between preemptive war and preventive war. There is no imminent threat supposed or cited in this case since Iraq has no nuclear weapons at this point.

Moreover, while other so-called “rogue states” present more imminent nuclear threats, the administration still insists that Saddam’s potential future capability justifies immediate action. As one Bush administration official told the New York Times after North Korea announced that it would remove monitoring equipment on both its nuclear reactor and plutonium stockpiles—enabling it to produce several more nuclear weapons within a few months—”We still think Saddam is the bigger threat, but there is no question that the North Koreans, who already have superior firepower, may soon be in a position to threaten to deploy or sell its nuclear capability. Iraq is a long way from that.”27

Thus, the administration’s arguments about Iraq reveal both its tendency to conflate preemption and prevention and the weakness of its factual, legal, and moral claims. The new National Security Strategy is not actually “preemption” in the manner described by the just war tradition or any of the other “legal scholars and international jurists” that the administration might be referring to. Rather, the policy is more accurately described as one of waging preventive war. But preventive war is not justified under just war theory, nor is it likely to be judged legal under international law. The trend has been quite the opposite: to equate preventive war, such as was launched by Japan against the United States in 1941 and by Germany against Russia and the Soviet Union in World Wars I and II, with aggression. In any case, it threatens a world of “bloody and exasperated war.”


Foreign policies must be judged not only on grounds of legality and morality but also on grounds of prudence. Preemption is only prudent if it is limited to clear and immediate dangers and if there are constraints on its conduct—proportionality, discrimination, and limited aims. If preemption becomes a strategy—or if it becomes the cover for a preventive offensive-war doctrine—it may become self-defeating as it increases instability and insecurity.

Specifically, a legitimate preemptive war requires that states show that the potential aggressors have the capability and the intention of doing great harm. But while capability may not be in dispute the motives and intentions of a potential adversary may be misinterpreted. Specifically, states may mobilize in what appear to be aggressive ways because they are fearful or because they are aggressive. Some states may defensively arm because they are afraid of the “preemptive/preventive” state; others may arm offensively because they resent the preventive-war aggressor who may have killed many innocents in a quest for total security. A preemptive doctrine which has—because of great fear and a desire to control the international environment—become a preventive-war doctrine is likely to create both more fearful states and more aggressor states.

In either case, whether states and groups arm because they are afraid or because they have aggressive intentions, instability is likely to grow as a preventive war creates the mutual fear of surprise attack. In the case of the United States’ preemptive/preventive-war doctrine, instability is more likely to increase because that doctrine is coupled with the goal of maintaining global “preeminence,” and the United States has said that it will discourage any military rivals from challenging it.28

Further, a preventive-war doctrine undermines international law and diplomacy, both of which can be useful—even to hegemonic powers. Preventive war short-circuits nonmilitary means of solving problems. If all states reacted to potential adversaries as if they faced a clear and present danger of imminent attack, tensions would escalate along already tense borders and regions. Article 51 of the U.N. Charter would lose much of its force. In sum, a preemptive/preventive-war doctrine moves us closer to a state of nature than a state of international law.

Moreover, while preventive-war doctrines assume that today’s potential rival will become tomorrow’s adversary, diplomacy or some other factor could work to change the relationship from antagonism to accommodation. As Bismarck said in 1875, “I would . . . never advise Your Majesty to declare war forthwith, simply because it appeared that our opponent would begin hostilities in the near future. One can never anticipate the ways of divine providence securely enough for that.”29

In sum, one can understand why any administration would favor preemption and why some would be attracted to preventive war if they think it could guarantee invulnerability. But this psychological reassurance is at best illusory and the effort to attain it may be counterproductive. Preventive wars are imprudent, because they bring wars that might not happen and increase resentment. They are also unjust, because they assume, as Bismarck said, perfect knowledge of an adversary’s ill intentions when such presumptions may be premature or false. The temptation to slide over the line from preemption to preventive war is great because that line is vague and because of the extraordinary stress of living under the threat of terrorist attack or war. But that temptation should be resisted. Vulnerability is a fact of life. And the stress of living in fear should be reduced by true prevention—arms control, disarmament, negotiations, confidence-building measures, and the development of international law.


1. British Ambassador to the United States Henry S. Fox in a letter to U.S. Secretary of State Daniel Webster, 12 March 1841, in Kenneth E. Shewmaker, ed., The Papers of Daniel Webster: Diplomatic Papers Volume 1, 1841–1843 (University Press of New England, 1983), 42.

2. Webster in a letter to Fox, 24 April 1841, in Shewmaker, 62, 67–68.

3. Quoted in David E. Sanger, “Beating Them to the Prewar,” New York Times, 28 September 2002, B7.

4. George W. Bush, preface to The National Security Strategy of the United States of America. (Washington, D.C.: Office of the President, September 2002), ii. (Emphasis added.)

5. The National Security Strategy, 6.

6. Donald H. Rumsfeld, “Remarks at Stakeout Outside ABC TV Studio,” 28 October 2001,

7. The National Security Strategy, 15.

8. Colin Powell, “Perspectives” interview, New York Times, 8 September 2002, 18.

9. The National Security Strategy, 15.

10. Ibid.

11. Michael Walzer, Just and Unjust Wars: A Moral Argument with Historical Illustrations (Basic Books, 2000), 85.

12. U.S. Department of Defense, Report of the Quadrennial Defense Review(Washington, D.C.: U.S. GPO, 30 September 2001), 2.

13. Report of the Quadrennial Defense Review, 30 and 62; and George W. Bush, remarks at Graduation Exercise of the United States Military, Academy West Point, New York, 1 June 2002 (White House transcript).

14. The National Security Strategy, 1.

15. Ibid., 31.

16. As Richard Betts argues, “When security is defined in terms broader than protecting the near-term integrity of national sovereignty and borders, the distinction between offense and defense blurs hopelessly. . . . security can be as insatiable an appetite as acquisitiveness—there may never be enough buffers.” (Richard K. Betts, Surprise Attack: Lessons for Defense Planning[Brookings Institution, 1982], 142, 143.)

17. Yet current U.S. military doctrine is to defend against potential and actual military capabilities, not against likely threats.

18. See “National Strategy to Combat Weapons of Mass Destruction,” December 2002,

19. Immanuel Kant, “To Perpetual Peace: A Philosophical Sketch,” in Immanuel Kant, Perpetual Peace and Other Essays on Politics, History and Morals (Hackett Publishing Co., 1983), 107–143:110.

20. I spell out the nature of this transformation in Neta C. Crawford, “Just War Theory and the U.S. Counterterror War,” Perspectives on Politics, vol. 1, no. 1 (March 2003). Also see Richard K. Betts, “The Soft Underbelly of American Primacy: Tactical Advantages of Terror,” Political Science Quarterly, vol. 17, no. 1 (Spring 2002), 19–36.

21. Critics of U.S. foreign policy will say that much of this vulnerability is of America’s own making because it financed, armed, and trained some of the individuals, groups, and governments that have turned on the United States. But these issues of responsibility are perhaps beside the point now—assuming, that is, that the United States is wise enough to halt current and future support for such unsavory characters. This of course remains to be seen.

22. Condoleeza Rice, “Wriston Lecture, President’s National Security Strategy,” 1 October 2002, New York, NY.

23. Report of the Quadrennial Defense Review, 14.

24. Quoted in Philip Shenon and Eric Schmitt, “At U.S. Nerve Center, Daily Talks on the Worst Fears,” New York Times, 27 December 2002, A12.

25. The National Security Strategy, 5.

26. Dick Cheney, “In Cheney’s Words: The Administration Case for Removing Saddam Hussein,” New York Times, 27 August 2002, A8.

27. Quoted in David E. Sanger and James Dao, “North Korea Says it Regains Access to its Plutonium,” New York Times, 23 December 2002, A1, A10.

28. Report of the Quadrennial Defense Review, 30, 62.

29. Quoted in Gordon A. Craig, The Politics of the Prussian Army, 1640–1945 (Oxford University Press, 1955), 255.