The military services in the United States have been organized on a volunteer basis since 1973, when President Richard Nixon abolished the draft. The end of conscription came as a relief to most people—to young men, their parents, and eventually the leaders of the military services, which had been plagued by internal dissent and a lack of professionalism, partly as a result of having so many unwilling members.

Though isolated voices have always challenged the shift to a volunteer military, their criticisms have recently become more widespread and more vocal. The main objections come from two quite different directions.

Some critics argue that the reliance on an all-volunteer, professional army has led to diminished public concern and vigilance with respect to the wars the government decides to fight. Limiting the burdens of military service to volunteers has, according to these critics, weakened inhibitions against the use of military force. When the Iraq War was debated in 2002–3, most citizens were not concerned that they or their children would be required to fight. This eliminated a powerful constraint against the resort to war. According to these critics, the reintroduction of some form of conscription is necessary to reestablish greater democratic control over the practice of war.

Other critics come from the ranks of just war theorists. Their concern is not with diminished public vigilance but with individual moral responsibility. They argue that volunteering for military service in current conditions is morally problematic. Those who join the military may be motivated by a desire to serve their country; to prove, improve, or reform themselves; to have a steady income with benefits; to get an education; to carry on a family tradition; or some combination of these. Whatever their motivation, they are committing themselves to become weapons controlled by others whose purposes cannot be reliably predicted. Some just war theorists question whether it can be permissible for people thus to convert themselves into instruments for killing without knowing whom they may be required to kill, or why. In contrast to the problem cited by the first group of critics, this one would be exacerbated rather than resolved by the reintroduction of conscription.

I will focus here on the second, less familiar critique. I believe that the all-volunteer military can survive the challenge, but only if it undergoes significant reform and acknowledges a right of selective conscientious objection.

Traditional Just War Theory
The idea that voluntary enlistment in the military can be morally problematic derives from a neglected tradition of just war thinking. This approach to the ethics of war informed the work of some of the classical just war theorists, such as the 16th century Spanish philosophers Francisco de Vitoria and Francisco Suárez. It was, however, gradually abandoned by thinkers whose views together constitute what I call “traditional just war theory.” The traditional theory has been ascendant since at least the 18th century, but the older approach has recently been resurrected by a group of “revisionists.” The best way to understand revisionist just war theory is to contrast it with the traditional theory, which has had a profound influence in shaping common sense thinking about the ethics of war, in part because it was developed in tandem with the international law of armed conflict.

According to traditional just war theory, a soldier does no wrong by fighting in an unjust war, provided that he or she obeys the rules regulating the conduct of war. This theoretical idea finds powerful expression in public sentiments. For centuries it has been regarded as not merely permissible but conspicuously noble and admirable for a soldier to go to war without any concern for whether the war’s cause was just. Although the famous lines in Tennyson’s “The Charge of the Light Brigade” refer to obedience to tactical commands within a war, they articulate a general Victorian ideal of soldiering: “Theirs not to reason why / Theirs but to do and die / . . . When can their glory fade?” Supreme Court Justice Oliver Wendell Holmes was even more explicit:

In the midst of doubt, in the collapse of creeds, there is one thing I do not doubt, that no man who lives in the same world with most of us can doubt, and that is that the faith is true and adorable which leads a soldier to throw away his life in obedience to a blindly accepted duty, in a cause which he little understands.

Traditional just war theory offers a theoretical basis for this familiar sentiment. A foundational tenet of the theory is that the principles governing the conduct of war (jus in bello) are entirely independent of those governing the resort to war (jus ad bellum). One implication of this fundamental dualism in traditional just war theory is that what it is permissible for soldiers to do in war does not depend on whether the war is just. Whether the cause is just or unjust, soldiers are prohibited, for example, from intentionally attacking noncombatants. And whether the cause is just or unjust, soldiers are permitted to attack and kill enemy combatants. Typically in a war between states, one state’s war is just while the other’s is unjust. Yet according to the traditional theory, the principles of jus in bello are neutral between soldiers on opposing sides and are equally satisfiable by all.

Even though soldiers do not make decisions about whether their states will go to war, they can make decisions about whether they will go to war.

One rationale for this traditional dualism is provided by a view about responsibility. Soldiers cannot do wrong by violating the principles of jus ad bellum because, it is claimed, those principles do not apply to them. The principles of jus ad bellum govern decisions concerning the resort to war, but ordinary soldiers have no control over those decisions and thus are not responsible for them. Only those who are involved in making such decisions—political leaders—are capable of either violating or complying with ad bellumprinciples. But even though ordinary soldiers cannot be responsible for whether a war is fought, they are responsible for how it is fought. And according to the traditional theory, there are permissible and impermissible ways of fighting even in a war that is unjust.

The traditional theory seeks to reinforce its claim that one can permissibly fight in an unjust war in part by observing that all soldiers fighting in a war pose a threat to others. Because they pose a threat, even if they are fighting for a just cause, they forfeit their right not to be attacked. But even though they forfeit this right, they retain their rights of self- and other-defense. Hence all soldiers, including those who fight for unjust aims (unjust combatants), are permitted to attack enemy combatants because those combatants pose a threat and are thus liable to attack. The traditional principle of noncombatant immunity is an extension of this line of argument. Noncombatants pose no threat and thus do not forfeit their right not to be attacked. Hence no soldiers, even those fighting in just wars (just combatants), are permitted to attack noncombatants.

We can refer to the traditional theory’s claim that it is always permissible to fight in a war, whether just or unjust, provided that one obeys the principles of jus in bello, as the permissibility of participation. It is a moral principle that is echoed by a parallel permission in the law of armed conflict, which allows that it is legal to fight in a war that is itself illegal.

The Revisionist Critique
Revisionist just war theorists reject the permissibility of participation. They think it is in general impermissible to fight in an unjust war even if one complies with the principles ofjus in bello, interpreted in the traditional way. Consider the argument that unjust combatants are permitted to attack just combatants because all soldiers make themselves liable to attack by posing a threat to others. That argument for the permissibility of participation has no force. For it is false that posing a threat to another is sufficient to make a person liable to defensive attack. When, for example, a police officer, having no other option, attempts to kill a rampaging murderer, she does not thereby become liable to attack, and the murderer has no right of self-defense. If the murderer kills the officer, he will be guilty of another murder.

In addition to having no theoretical grounding, the permissibility of participation is implausible on its face. Can one really believe that a Nazi soldier did nothing morally wrong by invading Polish territory and killing Polish soldiers who, perhaps just a few days earlier, were civilians who enlisted to help defend their society against foreign conquest? More generally, how can it possibly be morally permissible to kill people who have done no wrong as a means of achieving ends that are unjust?

Traditional just war theorists respond by claiming that the moral principles that apply in war are different from those that apply outside the context of war. Yet there is no reason to suppose that morality is schizophrenic in this way. War is on a continuum with other forms of violent conflict, and traditional just war theorists have never offered a criterion for distinguishing wars from conflicts that fall just short of war that would explain why entirely different principles come into effect when a conflict crosses the threshold to become a war. Suppose, for example, that we could identify a point at which the violence in Syria ceased to be a domestic conflict and became a civil war. Can anyone explain why what occurred at that transition was sufficiently significant to suspend the application of one set of moral principles and bring a different set into effect?

If the principles governing the morality of killing in war were different from those governing killing outside of war, one would often be unable to determine whether an act of killing was permissible without knowing whether it was part of a war, or of a conflict short of war. Yet if a person were involved in a conflict and were deliberating about whether it would be permissible to kill another person, it would be absurd for him to base his decision on whether the conflict satisfied the criteria, whatever they might be, for being a war.

The traditional theory’s claim that the priniciples of jus ad bellum—principles about when it is permissible to resort to war—apply only to those who participate in decision-making about the resort to war is also false. Even though soldiers do not make decisions about whether their states will go to war, they can make decisions about whether they will go to war. Hence it is entirely coherent to suppose that the principles of jus ad bellum apply to those decisions.

The revisionists therefore argue, in my view correctly, that the permissibility of action in war cannot be divorced from the ends that the action serves. To use the language of just war theory: they deny that jus in bello can be independent of jus ad bellum. They reject the dualism that lies at the heart of the traditional theory. If military action serves ends that are unjust, and bad from an impartial point of view, and in particular if the means employed include attacking people who have done nothing to forfeit their right against attack, that action cannot be morally permissible.

For centuries it has been regarded as not merely permissible but noble for a solider to go to war without concern for whther the war’s cause is just.

These theorists add that unjust combatants, fighting for an unjust cause, cannot in general satisfy even the principles of jus in bello, properly understood. Consider, for example, the principle of proportionality. Simplified somewhat for purposes of exposition, this principle requires that the good effects that an act of war can be expected to achieve not be outweighed by the expected bad effects, mainly collateral harms to innocent bystanders. (In recent debates about Syria, for example, those who have argued against limited strikes on the grounds that they would cause excessive civilian casualties and provoke the Assad regime to commit further atrocities have, in effect, been denying that such strikes would be proportionate.) But when the aims that military action is designed to achieve are unjust, and bad from an impartial point of view, the action cannot be expected to have good effects that outweigh the bad. Hence it cannot be proportionate.

A final point about the permissibility of participation is practical. Not only is this doctrine theoretically indefensible, but its incorporation into common sense thought about the morality of war also has had terrible consequences. For, by reassuring soldiers and conscripts that they can always fight with a clear conscience, even if they are certain their war is unjust, it has facilitated the initiation of unjust wars.

Why Enlistment is Morally Contentious
If the revisionists are right that it is impermissible to fight in a war that is unjust because the aims it serves are unjust, then voluntary enlistment in the military is morally problematic. Some people draw that conclusion because they think that submission to authority is never permissible. As moral agents, we must always—they say—preserve our moral autonomy, which means that we should not act against our own judgment by submitting to authority. But that seems too extreme. Suppose you could be certain that your action would, over time, conform better to the requirements of morality if you were to obey some authority rather than be guided by your private judgment. Then it seems right to obey the authority. But even if we accept this more plausible view, we should still have doubts about the permissibility of voluntary enlistment in the military. For no government’s moral judgments about its own resort to war can be expected to be more reliable than those of a well-informed, impartial, and morally scrupulous individual.

The problem with voluntary enlistment is not so much that it involves pledging to obey another or committing one’s will to obedience. It is, rather, that by enlisting, a person assumes the risk of becoming an instrument of injustice, either because he will fail to perceive that a war in which he has been commanded to fight is unjust, or because he will recognize it as unjust but fight in spite of that knowledge. In short, if the permissibility of participation is false—if it can, contrary to the traditional just war theory be wrong to fight in an unjust war—then enlistment involves a risk of engaging in serious wrongdoing.

This risk is substantial. When soldiers are commanded to fight in a war that is objectively unjust, they almost invariably obey. There have been countless unjust wars but comparatively few instances of conscientious refusal to fight in them by active-duty soldiers. There are many reasons for this. One is that soldiers often lack access to relevant facts about the war. Another is that they are often lied to by their government—even in the most open, democratic societies. Yet they generally trust their government and their superior officers. And even if they knew all the relevant facts, they might still be unable to recognize that their war is unjust, for they are usually actively discouraged from engaging in independent, critical thinking about the orders they receive. Military life is filled with activities, such as drills, that are designed to ingrain habits of obedience. Furthermore, even soldiers who are rightly convinced that their war is unjust are likely to fight. Not only have they been trained to obey orders, but they also tend to believe that obedience is a professional duty that overrides other considerations. They feel loyalty to their country, to the military institution of which they are members, and to their comrades in arms. And they can usually expect to be punished, perhaps severely, if they refuse to fight.

Much of this was recognized by another Victorian, John Ruskin, who, in a lecture to a group of soldiers, observed:

you have put yourselves into the hand of your country as a weapon. You have vowed to strike, when she bids you . . . . all that you need answer for is, that you fail not in her grasp. And there is goodness in this, and greatness, if you can trust the hand and heart of the Britomart who has braced you to her side. . . . But remember, good and noble as this state is, it is a state of slavery. . . . In thus vowing ourselves to be the slaves of any master, it ought to be some subject of forethought with us, what work he is likely to put us upon.

Perhaps many of those who join the military trust themselves to be able to distinguish between just and unjust wars and to have the courage to refuse if they are ordered to fight in an unjust war. If so, all but a few of them are deluded. There is some reason, though far from conclusive, to believe that the majority of soldiers throughout history have fought unjustly rather than justly. This is suggested by the plausible claim that there are only three types of war: wars in which one side is just and the other unjust, those in which both sides have just and unjust aims, and those in which both sides’ aims are entirely unjust. (If the opposition to the Assad regime in Syria consisted exclusively of jihadists determined to impose an Islamic theocracy—which fortunately it does not—the current civil war there would be an example of the third type.) If, as I think is true, there cannot be wars in which both sides’ aims are wholly just, but there can be wars in which both sides’ aims are wholly unjust, that makes it probable that there have been more unjust wars than just wars, and thus that more soldiers have fought in unjust wars than in just wars. Yet the record of testimony by soldiers indicates that most soldiers have believed their war was just.

In particular cases, of course, the general statistics may be misleading. If one is considering whether to enlist, the probability that one will be led to fight in an unjust war depends substantially on the recent history of one’s state and the nature of its government. The moral risks are higher, for example, for those who enlist in the Sudanese army than for those who enlist in the Swedish army.

A Proposal
We face a dilemma. There seems to be a moral presumption against enlistment, based on the serious risk that enlisting will lead one to attack and kill people in the service of unjust aims. Such killings are unjust. Yet most political communities require a military to protect their security and ensure their survival.

It would be embarassing for a government if its war effort were met with substantial, morally motivated resistance from its own soldiers.

The best response to this dilemma is to adopt provisions for selective conscientious objection by active-duty soldiers. Volunteers could then enlist in the knowledge that if they are commanded to fight in an unjust war, they will not have eliminated all options other than wrongful obedience and submission to punishment for disobedience.

This suggestion may seem hopelessly utopian and naïve. But there are precedents. A right of conscientious objection—indeed a duty of conscientious objection—is now recognized by the United States in matters of jus in bello. Soldiers are required to disobey a “manifestly unlawful order,” even during emergency conditions of combat, which seems in various ways riskier than permitting a soldier to refuse on moral grounds to go into combat at all. Moreover, conscientious refusal to fight occurs more often than most civilians are aware. The public seldom learns of these cases because the military often reassigns the objector to some other duty, thereby avoiding unwelcome publicity.

There are legitimate and quite serious concerns about having a legally protected option for conscientious refusal to fight. Perhaps the most obvious is that the option could be exploited by malingerers whose aversion to fighting is merely personal rather than moral. A person could join the military, perhaps “see the world” (as the recruiting ads used to say), but then refuse to do what he has been paid and trained to do.

I see no good solution to this problem. There would have to be tests for assessing a soldier’s sincerity, but such tests are highly fallible. There would also have to be tests for the plausibility of a soldier’s reasoning. It could not, for example, be an acceptable ground for conscientious refusal to fight that a soldier claimed to have heard the voice of an alien, or even the voice of God, telling him not to fight. It would, moreover, be necessary to impose penalties even on successful conscientious objectors as a means of deterring malingerers. Conscientious objectors might be required to do onerous tasks unrelated to the war and perhaps even refund some of the wages they had been paid while being trained for the work they now refuse to do. This would be unjust to those who rightly believe that their participation in a war would be morally wrong.

Still, the difficulties of solving the problem of malingering need to be weighed against the case in favor of permitting conscientious objection. There are two principal elements to this case. One is that it would diminish the moral risks a person takes in enlisting. When those risks are reduced, they are more likely to be outweighed by the moral reasons people have to serve in the military, which are primarily concerned with a society’s ability to fight just wars. The other is that if a country’s soldiers were to have a conditional option of conscientious refusal to fight, this could help to deter the government from attempting to initiate an unjust war. For it would be a profound embarrassment for a government, and would threaten its domestic legitimacy, if its effort to fight a war were met with substantial, morally motivated resistance from its own soldiers.

One might respond that just as the threat of conscientious objection by soldiers could prevent the initiation of unjust wars, so too could it prevent the fighting of just wars. This seems highly unlikely, though, in the case of a war of national self-defense. Yet in the United States, self-defense has seldom been a serious issue on those occasions since World War II when the country has used military force. Apart from the initial strikes against al Qaeda bases in Afghanistan in 2001, the major uses of military force by the United States that might be considered just have been instances of collective defense (for example, the Korean War) and humanitarian intervention (for example, Kosovo). And some soldiers might well have sincere and understandable moral objections to military intervention in another country even when there is an objectively compelling case for humanitarian intervention.

I accept, then, that allowing selective conscientious objection by active-duty soldiers could inhibit the fighting of just wars of humanitarian intervention. (The same is true of the reintroduction of conscription.) But given that it could also inhibit the fighting of unjust wars, the main question seems to be which of these effects would be greater. As I observed earlier, soldiers are strongly disposed to believe that any war in which they are commanded to fight is just. This partially explains why governments have always found it relatively easy to initiate unjust wars. And the evidence available to soldiers and others that a war is unjust is likely to be much stronger when the war really is unjust than when it is just. If these two claims are correct, then soldiers are considerably more likely to believe that their war is unjust when it is in fact unjust than when it is just. This suggests that permitting conscientious objection is more likely to inhibit the fighting of unjust wars than to inhibit the fighting of just wars. If that is so, the trade off between these two effects seems desirable, and an improvement over the status quo.

Allowing selective conscientious objection by soldiers would not eliminate all the moral risks involved in enlistment. Volunteers may still lack relevant information, be deceived by their government, or simply fail to care about whether a war in which they have been commanded to fight is just. And even if they believe that the war is unjust, they will still face pressures to fight rather than avail themselves of the option of conscientious refusal. But having the option of conscientious refusal would appreciably decrease the moral risks that volunteers would assume by enlisting.

Serious physical risks are ineliminable from the job of soldiering. For obvious reasons, those who take those risks must in general be relatively young, with more to lose than older people. It is therefore unconscionable that civilians should require soldiers also to take graver moral risks than are necessary. There are at least two additional ways in which these moral risks could be reduced, both of which would aid soldiers in thinking responsibly about the option of conscientious objection. One is to include in soldiers’ training some education in the morality of war. Soldiers should have some familiarity with the major issues in just war theory and the positions that prominent writers in the tradition have defended. Our various service academies already offer this to many of our future officers, but moral education needs to reach enlisted personnel as well.

The second way that civilian society can reduce the moral risks that soldiers face is to provide sources of guidance on the morality of particular wars that are independent of the government. How this might be accomplished is a subject for future debate. I hope at least to have shown why such a debate is necessary.

Photograph: Senior Airman Desiree N. Palacios / U.S. Air Force