JSCOPE 2004

“Five Degrees of Separation: A Plan for Non-Combatant Protection”

JSCOPE 2004

Dr. Pauline Kaurin

Pacific Lutheran University

DRAFT

Introduction

One of the greatest contemporary challenges to the traditional combatant/non-combatant distinction is the question of practicality: in warfare that is not strictly conventional, where it is not clear whom the combatants are, where the non-combatants may appear hostile and the combatants may try to hide in non-combatant populations, how do soldiers determine who is a combatant (and a legitimate target) and who is a non-combatant (and not a legitimate target)? In a previous paper “Innocence Lost: The Future of the Combatant/Non-Combatant Distinction” I made both moral and practical arguments for a five level distinction to expand the current two level distinction. The basic line of argument (which I will recap in the first section of paper) is that since there are multiple ways in which one can be a combatant and multiple ways that one can be a non-combatant, a distinction that takes account of this fact is morally necessary and practically more likely to be implemented and observed on the ground.

In this paper I want to more fully develop what exactly this five level distinction would look like, in what each level would specifically consist and how each level might be implemented in various kinds of situations, including ‘traditional’ combat, guerilla combat, peacekeeping and humanitarian scenarios. The aim of this second paper is to develop the five level distinction in some detail and show how it could be applied in ways that would answer objections that such a model would be impractical, too hard to teach/learn, or would erode the moral intentions that now support the traditional distinction.

Before laying out the details of the five-part distinction, it will be important to discuss the basic line of argument in favor of a multi-part distinction, as opposed to the traditional two part combatant and non-combatant distinction.[1] What is it about combatants that justify them being the object of violence? What is it about noncombatants that does not justify the use of violence? The most accessible fact about warfare seems to be the power differential between combatants and non-combatants: some are armed and have a great deal of ability to be a threat to others; some, not armed, are vulnerable to those who are. Looking to Michael Walzer’s view, we see that combatants are dangerous individuals; they carry weapons, they can/will inflict this harm on others.[2] They are dangerous because they can inflict harm and put others in dangerous situations; they can cause others to have their rights violated, especially their right to life, but also the right to autonomy or non-interference with their livelihood. In short, they have a certain kind of power – an ability to be a threat, to be dangerous.

It is this core intuition that will provide the foundation for both the moral and practical arguments for an expanded combatant/non-combatant distinction. There must be something different about the actions of the persons in the category that justify different treatment; there are not just different administrative categories, but there are different categories of actions – multiple ways to be/act like a combatant and multiple ways to be/act like a noncombatant. First, there is a requirement of reason that one treat like cases alike for the sake of logical consistency. If there are persons who by virtue of their gender, age or social function would normally be considered “non-combatants,” who are acting in a hostile, threatening manner or causing harm, it would seem to be illogical to argue that they should be treated in the same fashion as others who are not acting in that fashion, who are not evincing threats or hostility.

However, it also seems illogical to lump those “hostile” or threatening “non-combatants” as being worthy of the same treatment as those who are clearly identified as “combatants,” who threaten openly by visible signs (carrying a weapon, wearing a uniform with fixed visible signs etc.) and have the advantages and protections of a military or para-military structure. It would be a category mistake not to make more fine-grained distinction as to how these individuals are to be treated, if there are fine grained-differences between how these individuals act and the effect that they can have on others. Additionally, it is a requirement of justice to treat like cases in a like manner and that either morally or legally speaking it is unjust to treat like cases in a different manner, unless there is a compelling reason to do so. Here it is possible to raise concerns about undue partiality: How are we applying the standards of justice in this case? What is the rationale for its application? Given this application, how will future cases be handled?[3]

In order to fairly and justly accommodate analogous nuances within warfare, we will need more than the two traditional categories. At a minimum the two traditional categories will be necessary, but also categories for unconventional belligerents and a category for “non-combatants” that might pose some threat. Such an expanded distinction would allow for a kind of flexibility that is required both by considerations of logical consistency and justice, but a flexibility which is not ad hoc or unduly partial. This distinction is further supported by the intuition that there is “something about that person” that requires different treatment – even if they are both non-combatants – and tries to articulate and classify what that “something” looks like.

Secondly, a multi-part distinction accords better with both classical and contemporary Just War theory, in particular the jus in bello considerations of discrimination and proportionality. The principle of discrimination (or non-combatant immunity) calls for soldiers to distinguish between those who are legitimate targets and those who are not, refraining from attacking those who are not legitimate targets. The rationale for the immunity of non-combatants is tied to the idea that these persons have rights that cannot be violated, even for a legitimate military purpose. Take the case of munitions workers (who are often in that ‘gray’ area between combatant and non-combatant): according to Walzer they can only be attacked “…when they are actually engaged in activities threatening and harmful to their enemies….We call them innocent people...which means they have done nothing, are doing nothing which entails a loss of their rights.”[4]

Even though Walzer makes the distinction a matter of rights, the rights are rooted in whether the actions of the people in question warrant having their rights violated. A multi-part distinction maintains that there are crucial differences between combatants and non-combatants, but also recognizes that there are important differences – largely in terms of power and vulnerability – within these categories as well. If there are these differences within the categories, then it also follows that there are different types of rights within these categories that ought to be taken seriously. Failure to recognize and accommodate this fact could mean that the traditional two-part distinction (in attempt to preserve the rights of the corresponding two groups) ends up violating the rights of the other groups, i.e. hostile non-combatants or unconventional belligerents.

An expanded distinction also accords better with the principle of proportionality. The idea of proportionality has played a central role in the Just War tradition both in its own right, and also in connection with the idea of self-defense.[5] If one uses more force than is necessary to achieve the end, one is inflicting unnecessary (and therefore unjustified) suffering which could in turn endanger the possibility of restoring a lasting peace after the conflict has ended. The traditional distinction generally recognizes two categories of force as legitimate: against combatants lethal force is justified, but against non-combatants lethal force is not justified, and further they are entitled to protection from force. Therefore, it is possible that having only two categories ends up inflicting more harm than necessary to achieve the end. An expanded distinction, on the other hand, is able to recognize gradations of power, danger and threat within the categories of combatant and non-combatant and can require that one use only enough force to stop or deal with that threat, thereby avoiding the infliction of unnecessary suffering. With an expanded distinction the question becomes “How much force is necessary to deal with this threat?” rather than the question “Is the use of force (usually lethal) justified or does this situation preclude any use of force?”

However, we must also look at the principle of proportionality in terms of self-defense. The question of killing in self-defense was an important one in early Just War thinking and was generally only justified in the case of the common good.[6] Why? Presumably because if it were for the common good, then whatever good that would come from the actions would outweigh whatever harm was inflicted, whereas if it were individual self-defense, that might not be the case. However, one could only use means, even for the common good, that would not tip this balance by causing harm and suffering beyond what is necessary to achieve the end in question. If unconventional belligerents are more vulnerable than traditional “combatants,” one is not only making a category mistake or violating an intuition about justice to treat them the same, but also violating the principle of proportionality. One could be using more force than is necessary, more force than can be morally justified, in responding to them. If neutral non-combatants are less vulnerable than traditional “non-combatants,” treating them in the same manner as the traditional “non-combatants” also could violate proportionality since more force could be required, and therefore justified, to defend oneself.

In addition to the moral considerations in favor of a multi-part distinction, there are also important practical arguments for expanding the traditional combatant/non-combatant distinction. First, actual field practice in recent conflicts would suggest that multiple distinctions within the traditional distinctions are already in effect. Looking at the examples of Somalia and Bosnia, these kinds of fine-grained distinctions already seem to be made on the ground. If in the face of these practices and complexities in war, we insist on keeping the part distinction, and only those two parts, soldiers may abandon the distinction altogether since it does not reflect the realities of war or at least their experiences of war in the field.

Second, despite the fact that combat and police work do have important differences, there is one similarity: in contemporary contexts both groups have to deal with using force in situations where non-combatants are present in areas where there are combatants (or dangerous individuals). In any given encounter, police officers are often working in situations where there are "non-combatants" whom they have a duty to protect, other "combatants" in the traditional sense (other uniformed officers) and suspects who might or might not be threatening. They have to be able to assess the level of threat and to be able to use the proper force required. Is deadly force warranted? Could non-lethal weaponry do the job? If they use more force than necessary they might kill or harm unjustifiably, while if they use less force than required they could endanger the "non-combatants" by not stopping and/or apprehending the suspect. Such a scale of threat assessment is not considered impractical, unteachable and irrelevant for police work. It stands to reason that it could be made workable for the military – although clearly such an approach could not be adopted wholesale, but would have to be adapted to the various needs of combat, humanitarian and/or peacekeeping missions.

II.

Before working out the details of a five part distinction for military operations – whether combat, peacekeeping or humanitarian intervention missions – it will be helpful to look at some other models and explore their strengths and weaknesses, and also to consider both analogies and disanalogies to military operations. The most obvious source of such a model would be in police work – the so called “Use of Force” models which are often based upon a continuum-type framework. The idea behind these models is that as the suspects behavior escalates, then the means that the officer can use to contain and/or arrest the suspect can also escalate – even to the point of using lethal force. According to John C. Desmedt, who formulated one of the first models used in the 1980’s, “The purpose of a use of force model is to accurately guide a reasonable response for virtually any situation. Overreaction may not only be unconstitutional, it may also be ineffective. Under reaction may allow harm to come to the officer or others.”[7] Such models are used in training, testing and operational analysis, as well as for litigation to ensure that the department policy of a given police unit is consistent with the law, as well as its own training objectives, administrative actions and policies.

In most of the models, there will be an axis on one end which will, in descending order, delineate various levels of subject behavior usually including some of the following categories: the Cooperative Subject, the Resister/Uncooperative Suspect, the Assailant. Within each of these categories there are often two sub-categories (sometimes called passive and active) which distinguish different features of that category which may require different force. First, the passive, but cooperative suspect is one who is compliant, but initiates her own compliance without verbal cues from the officer; the active cooperative subject is compliant with the officer when given verbal cues and instructions. Second, the passive Resister or Uncooperative Suspect is unresponsive to verbal compliance, is passive in response to cues from the officer; while the active Uncooperative Suspect is unresponsive and non-compliant and initiates active steps to avoid any attempts at control by the office. Third, the passive Assailant (or more accurately a low level assailant) is aggressively offensive but is not yet placing the officer in danger of serious bodily injury or death, while the active assailant does just that.[8]

On the other axis of the model there will be varying levels of force that the officer can use to control the situation (remembering that the purpose of the use of force model is control of the situation and the participants.)[9] In Desmedt’s model one has the following hierarchy of control mechanisms (in escalating order): social control, verbal control, weaponless control techniques, electrical shocking devices, chemical agents, control instruments, impact weapons, firearms (or other weapons that can result in death or serious injury.)[10] Another use of force model starts with command presence, moving to verbalization skills, control techniques and eventually to defensive techniques and/or tools.[11] The Montgomery County, Maryland police department also uses a four level continuum: “1) Communication/Voice Command, 2) Physical Force, 3) Protective Instruments, 4) Deadly Force.”[12]

How do these two parts of the models interact? From the standpoint of U.S. Constitutional law, it is the standard of a reasonable amount of force, not the minimum amount of force, that is what will determine whether the force used by a police officer is excessive or not.[13] So the idea in these models is that one would look at the behavior of the subject or assailant to determine what amount of force would be reasonable to control that level of behavior. If we imagine these two axes with a diagonal line running between them, on one side of the diagonal lies an areas of probable ineffective control and on the other side lies probable excessive force. In between these two areas is where reasonable force lies, the area where the officer can have effective control of the situation (to also avoid harm to the officer and other civilians in the vicinity) without using excessive (and unlawful) force.

For our purposes, what is valuable in these models is that fact that subjects are categorized into multiple categories and that the level of force to be used is indexed to their behaviors that the officer can observe in the situation. However, this does not mean that these models are uncontroversial or without concerns, especially with regard to their implementation. First, it has been objected that these models are too mechanical and automatic to meet the legal standard of reasonableness.[14] In legal cases one wants to know what the reasonable police officer would have thought in that situation, how the reasonable police officer would have perceived the threat posed by the behavior of the suspect and the force necessary to control that situation. This suggests that these kinds of assessments are the product of training, intelligence and professional experience which cannot be gained from a simple chart; they must be able to use these skills and experiences in the assessments, for which these models can be a guide, not the beginning and end of the model.

Second, some commentators have raised the serious concern of hesitation in field situations.[15] The argument is that since many of these continuum models are progressively sequentially in nature, the officer has to go through the process of the various steps of the model in deciding what force to use and this puts hesitation into the situation. This hesitation can be deadly, either because it can allow the situation to escalate more quickly as the subject sees there is hesitation or because the officer may hesitate too long in eliminating the less intrusive force option, when in fact stronger force is called for to quickly control and deescalate the situation. This concern centers around the time that is necessary for even the well-trained officer to go through the continuum and make her assessment about the behavior exhibited and the appropriate level of force; clearly in volatile, fluid and unpredictable law enforcement situations, time is one of the most precious commodities. Instead, the argument goes, the focus should be on training officer to learn the skills of subjective threat assessment and to accurately be able to know when deadly or serious force is necessary and apply it effectively.

What would a threat assessment model look like? First, it is important to define precisely what one means by threat: “A threat is a capability to do harm joined by hostile intent.”[16] Therefore, threat assessment training should emphasize indicators of capability to do harm, as well as indicators of hostile intent, since the goal such training is to allow officers to recognize an immanent threat and to respond in a timely matter. Indicators of intent can be difficult to isolate, but might include verbal and non-verbal communication, along with noncompliance with clear verbal commands of an officer. Indicators of capability are more tangible and can include: possession or access to a weapon (even the officer’s), demonstrated combat skill or ability, size or fitness of the suspect or the presence of multiple subjects. Looking at the FBI’s training methods, one can see the following four examples of a deadly threat: 1) possession of or an attempting to gain possession of a weapon; 2) armed suspect running to gain cover; 3) subject who demonstrates intention or capability to inflict serious injury or death without a weapon; 4) subject is trying to escape vicinity of violent confrontation in which he inflicted or attempted to inflict injury or death.[17]

The idea with a threat assessment model, as opposed to the use of force model, is that if a clear and serious threat exists then the officer can immediately respond with adequate force, without the delay that it is argued the use of force model brings. If threat assessment suggests that it is not a deadly or violent threat, then the officer can take the necessary time to consider a range of non-deadly options. However, it should be noted that other options are only to be considered if the officer assesses the threat to be non-deadly or non-violent; otherwise this model suggests that the officer should respond immediately with strong force. The difference between the two models is that the use of force model, it is argued, tends to wait and see how the threat manifests itself and then to respond more stronger as the threat escalates; the threat assessment model wants to look to see if strong, especially deadly force, is appropriate and if it is, then implement it quickly and if it is not, then take time to consider the appropriate level of force. In short, the idea is for the officer to associate the use of force with the presence of a threat; when the threat ceases or decreases, then the type and amount of force can be modified accordingly.[18]

With these two models in mind, how applicable might either or both of these law enforcement models be to how soldiers use legitimate force (against combatants, not against non-combatants) in warfare? What are the analogies to warfare? To answer this question, it might be helpful to have a working definition of police work, as a point of comparison. One source defines policing as follows: “They are representatives of the state whose responsibility is to intervene proactively and reactively to a wide range of disturbances…police are professional interveners.”[19] Further, police must do the following four things in any encounter: 1) they must seek information about the situation; 2) they must attend to the overt behavior of the civilians in their presence; 3) they must be treated by civilians in a respectful manner, otherwise their status is threatened; 4) they must achieve an appropriate resolution or outcome to the situation.[20]

Given this understanding of police work, one could argue that both police and soldiers use lethal force in their encounters and must further, need to know when to use this force, on whom it can be legitimately used and when it cannot be legitimately used (particularly against uninvolved civilians.) Further, in both police work and combat one is responsible and accountable for both the force that one uses and for the protection of uninvolved civilians/non-combatants. In both areas the use of excessive force is viewed as morally and legally illegitimate and can be a crime for which the individual can be held responsible. Such excessive force is viewed as endangering the community that is represented, but also as a betrayal of the code to which police and soldiers claim obedience. On the other hand, insufficient use of force can endanger other colleagues (other officers or soldiers) as well as the civilians that one is sworn to protect. Both groups are involved (especially in contemporary warfare which can involve standard combat, peacekeeping, humanitarian missions – or more often a combination of all of these) in fluid and potentially volatile situations where the combatants and non-combatants (officers, suspects and civilians) are in the same physical proximity. Consequently, individuals need to make quick, but accurate assessments based in information/intelligence in the field, usually without much time for deliberation or reflection. Finally, it could be argued that in both situations training and teamwork with ones colleagues are essential for a good, timely and ethical response. Without these elements, the situation could get further out of hand and result in unnecessary injury or death.

Despite the considerable analogies between military operations and police work, there is a considerable and important objection that there are also a number of disanalogies between the two, that they are very different kinds of things. While police officers can and do interact using lethal force, very few (proportionally speaking) of their encounters involve the use of lethal or potentially lethal force; most of their interactions involve verbal and non-verbal communication as a major tool of control and achieving their desired outcome. However, with military operations (notably combat) this is not possible and in fact, many more encounters will involve the use of lethal or potentially lethal force. (One can argue that this analogy holds better with peacekeeping and humanitarian interventions which seem to involve more police-type encounters, but there is the still the consideration that lethal and potentially force is employed to a greater degree than in domestic police work.) In the military, lethal and potentially lethal force is the form; in police work it is the largely the exception.

Another disanalogy comes from the emphasis one could argue that the military places on teamwork, espirit de corps, in training, in codes of honor and professionalism and in actual operations. In police work, especially domestic police work, it is the individual officer (although backed up by others as necessary) that is the basic organizational unit; in the military, it is rarely the individual soldier acting alone that is the basic organizational unit, but a small group of individuals working together. Therefore in military operations, teamwork and the chain of command is much more important than the individual decision-making ability of one person. (Once again, one could modify this argument a bit to suggest that the group responsible in humanitarian and peacekeeping operations is much smaller and may, in fact, be an individual soldier – or a couple – and is more analogous to police work.) A reason for this difference is that the essential functions of these two groups are quite different: the police are charged with maintaining civil order and protecting the public and property; soldiers (at least in combat situations) at charged with defeating the enemy, getting the other combatants to surrender or cease fighting, and protecting or at least abstaining from inflicting harm on non-combatants.

So if there seem to be some analogies between military operations and police work and yet there are some important differences, where does leave us? In particular, how helpful will ‘use of force’ or ‘threat assessment’ models taken from police work be in working out a multi-part combatant/non-combatant distinction? First, I think that we can conclude that in peacekeeping, humanitarian interventions and other non-conventional combat situations there are more analogies to police work and therefore, the ‘use of force’ models could be helpful. However, because of the disanalogies (especially in conventional combat situations) any model should have a heavy threat assessment component. Any ‘use of force’ model we employ for military operations (regardless of the kind) should be approached in descending order – starting with the highest level of force applicable and reducing the force as the threat reduces or is controlled. Most ‘use of force’ models in police work assume an ascending model of force where one starts with the least amount of force and call for increasing it as the situation or suspect behaviors escalate; this approach will clearly be problematic in military operations where the use of lethal force is much more likely and it is less likely that one can use verbal or other non-lethal options to deal with the situation.

What I will argue in the next section is that a soldier use a threat assessment kind of approach to determine whether the individual belongs in the combatant on non-combatant category, and then within that category, she can employ something more like a use of force model (in descending order of threat) to ascertain what level of force is justified depending on where they fall within the combatant category (is the subject an unconventional belligerent or merely hostile?) or the non-combatant category (is the person neutral or vulnerable?) This allows us to make use of the insights that might be gained from domestic police models, while at the same time recognizing the important differences between police work and military operations.

III.

What exactly would this multi-part distinction look like? How might it work in the field? To begin with, this scale is designed to have the ability to be somewhat flexible in its application. Clearly there are areas of ambiguity that the traditional combatant/non-combatant distinction cannot accommodate, but it is important not to give up assignations of responsibility, nor to deny the moral difference between those who threaten and cause harm and those who do not. Rather, I want to suggest a pragmatic way in which this distinction can be understood and applied in the complex and fluid situations that are characteristic of contemporary war. This scale could especially help those in the field to make a determination of the potential threat that is involved in a given situation, and therefore give guidance on how best to act consistent with International Law and the ‘rules’ of war.

The idea behind this scale is to incorporate the necessity of threat assessment with the flexibility of the use of force models that we discussed above. The first determination that would need to be made is whether the person in question is a combatant or non-combatant (or is likely to be one or the other since it certainly can be unclear). Once this determination is made then the question of how to treat this person, precisely how much force is to be used and how can be made by looking at what kind of combatant they might be or what kind of non-combatant they might be. This is where, even in dealing with a combatant, one might modify the kind and extent of force that one is using depending on the circumstance – whether the situation is a tradition combat situation, a peacekeeping or humanitarian mission. To begin, let us look at each step in the model from the standpoint of threat assessment: What are the intentions of the subject? What are capabilities of that person to carry out any intentions?

The first and highest level of threat on this scale would be that of the uniformed combat personnel. This would be an individual with a standardized (at least to his own military context) uniform with discernable symbols and identification which announce him as such to those who might see him. This person carries a weapon openly, often travels in structured groups of his colleagues; it can be ascertained from a distance that his intentions are likely to be hostile, that he has the ability to carry out these intentions and therefore, poses the highest level of threat to a possible enemy. This category is how the combatant in conventional warfare has usually been defined and in a conventional combat situation, it is expected that this is the kind of combatant to be encountered. Since this type of individual clearly poses the largest threat on the scale, the level of responsibility he must meet should also be high.

The second level of threat would be an armed individual (or unconventional belligerent), but not necessarily or obviously in a standard uniform that clearly announces her intentions. Since this person is armed, it is likely (though not certain) that her intentions are hostile and since her intentions are not evident, but a weapon is, this comprises a fairly serious threat. It may turn out that she has no hostile intention at all, but the presence of a weapon at least raises concern and potential threat, thus placing her potentially in the combatant category. The fact that the intentions are not announced also points to the potential threat and therefore, she should be viewed as a threat and having a great deal of power – at least until such time as she announces her intentions (which would either move her up or down the scale definitively.) One might encounter this kind of combatant in a conventional combat situation, raising questions about whether she is also to be treated with the same force that one would use against a conventional combatant, but more likely this is the highest level of combatant that one would see in peacekeeping and humanitarian type missions.

The third level of threat would be classified as hostile. In this level the intentions of the individuals are not made clear by a uniform or other symbol which announces their intentions and it is also not clear whether or not they have weapons (or have easy access to them). However, these individuals may meet a certain description that fits potentially hostile forces, be present in an area where such individuals are suspected of operating or in some other way persuade the on-looker that they are neither neutral nor vulnerable (levels four and five of the non-combatant end of the scale.) While this is admittedly somewhat ambiguous, it is necessary to leave this category somewhat open in order to accommodate differences between combat missions and those that are humanitarian or peacekeeping in nature. (In the later two situations one may able to gather intelligence or use verbal skills to question the person or their compatriots, whereas in a combat situation one may have to assume they are hostile until non-verbal cues or behavior indicates otherwise.) Clearly, what looks ‘hostile’ in a combat situation may differ from other situations, or vice versa. Another important component in this level is the fluidity of the situation, such that events and situations are highly volatile and subject to change, which makes the ambiguity of intentions and weapon possession dangerous.

At this point, it is essential to stress that that these first three categories are all different variations on the traditional combatant category; these individuals are all to be treated (in relative degree) more or less as one would treat a combatant. This also means that there is a clear conceptual distinction between levels three and four/five (even though individuals may might move from the combatant categories to the non-combatant categories.) Levels four and five are non-combatants and are entitled to the protections associated with those groups (neutral powers and “innocents”) under International Law and the rules/customs of war.

The fourth level of threat would be classified as one who is neutral or non-hostile. In this case the individual appears to announce no hostile intentions (symbolically by dress or insignia, by the presence of a weapon, or by verbal or non-verbal communication) but neither has she announced friendly intentions nor is evincing open cooperation with military forces. In terms of capability, she may have indicated (or opposing soldiers may have discerned) that she has no weapon, no easy access nor made an attempt to gain one. As long as she maintains her neutrality/non-hostility, she could be considered to be a fairly minimal threat, but clearly if the neutrality appears to change then her status would move to either the third or final category. If the neutral stance is maintained then this person is not a threat, and responsibility should be rendered accordingly.

The final level of threat would be one who is classified as vulnerable. Any hostile intentions that may be exhibited (i.e. a prisoner in uniform, angry speech) are counter-acted by a clear inability or lack of desire to act on them. This individual poses a very minimal (minimal – not zero – because they might change their status and this has to be accounted for) threat and, in fact, may require positive protection from other hostile forces. Clearly many of the people who are traditionally classified as non-combatants would fall into this category (most children, women, the old, prisoners and the infirm or injured) and should be accorded the usual protections that we think of because they clearly pose little threat and may be threatened themselves – with no recourse to defend themselves.

So how might one implement this scale in the field? First, let us look at a conventional combat situation (i.e. ground war in Iraq.) In such a situation is it likely that one will encounter persons in the first and possibly the second levels as well as possibly the fourth and fifth categories. Clearly one must immediately ascertain whether one has a combatant or non-combatant. If the person turns out the be a non-combatant, then one can immediately lower the force used to be adequate to deal with level four (controlling the situation) while one ascertains what kind of non-combatant is involved. If one decides that this person is in fact level five (the traditional non-combatant in need of protection) then one can further modify ones tactics.

On the other hand, if one ascertains that the person is a combatant (either it being clear that they are or one has a suspicion of hostility) then the soldier should start with force in accordance with level one or two. If the person is a conventional belligerent and obviously so, one would start with that scale of force and if not, one starts with level two force and moves down the scale if it becomes clear that less force is necessary. It would be well to point out that in many combat situations (both conventional and asymmetrical) level three persons could actually be more dangerous simply in virtue of what is not known, so that the force differential between levels two and three may be minimal if there is any difference at all. In all three of these levels, the person is a combatant (or is considered as a combatant) and so lethal force could be legitimate.

Second, let us look at how this scale might work in the case of a peacekeeping operation (e.g. Bosnia) where U.S. forces are charged with keeping opposing forces in check and/or keeping civil order. In such a situation, one may encounter level one personnel as combatants, but it is much more likely that one will be dealing with levels two and three, as well as level four and five non-combatants. In any case, one will still need to be able to discern combatants from non-combatants as in the above combat scenario. If it is determined that one is dealing with combatants, then one would start with the highest level of force applicable (likely to be level two) and modify in descending order if that becomes appropriate. For example, it may turn out one has simply hostile forces who could turn out to be non-hostile non-combatants (in the event you determine they are not armed), but for now it is clear that they are hostile and not clear whether or not they are armed. In this situation, you might use less force than you would in a similar combat situation (because the individuals in question could be non-combatants and also it is likely that there are non-combatants in the area), but because of the potential danger (level three is still a combatant category) one may still need to keep ready lethal force.

Third, in a humanitarian intervention (e.g. Somalia where the US military is charged with delivering food) it is likely that you have even more non-combatants in the areas and probably are dealing with level three combatants (although level two combatants could also show up to disrupt the situation). In this case one probably has to assume level three combatants with a force appropriate to dealing with level of threat, but also not unduly endangering the surrounding non-combatants. To further complicate matters, it may turn out that the level three combatants are level two combatants in disguise and so the soldier has to be prepared for that possibility, but keep that level of force in check until the presence of weapons is ascertained. In peacekeeping and humanitarian missions, the force differential between levels two and three could be significant (lethal force vs. other methods of control more like conventional police work) and so the difference between these two levels becomes more important to establish. In these two kinds of operations, individuals could easy move from level three to level four (you ascertain they do not have a weapon) or vice versa (they begin exhibiting hostile behaviors or gain access to a weapon as a situation escalates) and so one has to modify the kind of force that is appropriate in level three accordingly.

To sum up, whatever situation the military operation finds itself in, the first matter of business is to ascertain whether the individual is a combatant or non-combatant since that will determine which end of the scale one begins with in determining the level of force. If the person is a non-combatant, then we start with a much lower level of force, if any at all; if the person is a combatant, then one starts with the highest level of force appropriate to the likely kind of combatants – modifying (usually reducing) the level force necessary as finer determinations are made about the kind of combatant involved and the threat posed. Clearly the difference in how one treats a hostile combatant versus a non-conventional belligerent in a combat situation may not be that drastic, but I believe those differences will be much more significant in peacekeeping and humanitarian missions. What I want to suggest is that we can still maintain the traditional distinction with its attendant moral considerations, but also recognize that there are various levels of gray (especially in how one might present as a combatant) that may call for different levels of force and may require some fluidity in how one applies that force. I think that a five part model such as I have outlined above can do just that.

IV.

Naturally there are numerous concerns and objections that would be raised to any such model, but I want to concentrate on three. First, one might object that this model relies too heavily on the law enforcement model and is, consequently at odds with how the military operates since the police and the military have too very different purposes. The majority of police work is concerned with maintaining civil order and apprehending and controlling those who resist or who are a danger to that civil order; the police do not need to defeat an enemy, but only need to be able to exert control in the situation long enough to resolve it. On the other hand, military operations are largely concerned with defeating the enemy and mastering control of a physical area beyond just restoring civil order, which involves sophisticated technologies, strategies and extensive teamwork to bring this about. It could also be argued that there are serious differences in organizational structure, training methods and institutional values which also require very different approaches. In short, we are comparing apples and oranges.

While it is true that this model does rely heavily on a multi-part distinction which has analogies to the use of force models found in domestic police work, I would argue that the five part model I have suggested is based upon military considerations – both moral and practical. In the last half of the 20th century militaries (especially in the developing world) were engaged in military operations where it seemed increasingly difficult to distinguish the combatants from the non-combatants; in engagements like Vietnam one could argue that the distinction had been abandoned de facto as the realities of the battlefield seemed to eclipse morality and military (not to mention legal) conventions. What I have tried to do is to use a five part distinction to argue that it is possible to maintain this important distinction (with its necessary moral and legal protections for non-combatants), but only if we expand the distinction to recognize that there are different kinds of non-combatants, and especially that there are different kinds of combatants. Using a “one kind of force fits all” model ultimately undermines the very distinction that it seeks to preserve, especially when battlefield requirements and realities seem to fly in the face of the distinction.

Second, one could reasonably object (especially if one is charged with the teaching and training of young military personnel) that such a model is too complicated to teach and learn in the classroom, much less implement in the field. It is in answer to this objection that I think the analogies to law enforcement are useful, especially in the cases of peacekeeping and humanitarian missions. While the use of force models are not uncomplicated and not without their critics, the fact that they have been used in police departments and law enforcement agencies for nearly twenty years, have been taught in the course of standard training and have also legal, especially constitutional, scrutiny by lawyers and the courts suggests that it is neither impractical, nor impossible. That, however, is not to suggest that it is easy either. I recognize that my model does add an additional layer of discussion to the teaching of the combatant/non-combatant distinction that takes place in the academies, OTC, ROTC, basic training and in the field. Rather than just focusing on the differences between combatants and non-combatants, I am also asking students and teacher to reflect on what a combatant in its various incarnations, and a non-combatant in its various incarnations, might look like and what level of force is legitimate in dealing with them.

However, I would also argue that many of these conversations are already happening as students, teachers and military professionals alike grapple with the difficulty of applying and maintaining this distinction in complex, contemporary warfare. What this model does it to give a voice to those conversations, and a way to sort out that confusion both in the classroom and on the battlefield in a way that I believe is consistent with the moral considerations that drive the distinction (and the larger apparatus of Just War theory) and the military traditions and considerations that up it.

Third, the most serious objection is that such a model undermines the original intentions (especially the moral ones) of the traditional combatant/non-combatant distinction and would further erode the distinction, rather than shore it up – as I have argued. The main area of concern would likely be the third and forth levels where there is some necessary gray area. In level three, if it is not clear whether or not the individual has a weapon, why should they be considered a combatant? Would not be better to consider them a non-combatant until one can make that determination? On the other hand, persons who get assigned as level four non-combatants could very well combatants in disguise or be potentially dangerous, does the fact that someone does not have a weapon make them not a threat? Rather than clearing up the gray area that is so problematic in maintaining the distinction, this model actually seems to add to the problem by further institutionalizing this ambiguity.

The intention behind the distinction is to protect the non-combatants from being unjustifiably harmed and to ensure that force is used only against those whom it is justified, namely the combatants. Any model which asks soldiers to take seriously the question of how much and what kind of force they need to use against the enemy and helps them better identify those against who they should use this force, as opposed to those against whom it should not be used, will further this end. Rather than undermining the intentions of the distinction, I believe that such a model could potentially reduce the unfortunately incidents of “collateral damage” as soldiers have more options about how, when and to what extent they use force. The original distinction asks us to take serious that there are moral and practical differences between combatants and non-combatants which at the core of fighting a war justly; this model simply tries to more clearly articulate what each of these categories might look like – taking into account the complexities and ambiguities of contemporary combat, peacekeeping and humanitarian operations.

Conclusion

In this paper I have outlined and defended in more detail a five-part combatant and non-combatant distinction which is intended to do two things: first, this model tries to account for the considerable difficulty in sorting out the different varieties of combatant and the different varieties on non-combatants that one might encounter in military operations; second, rather than arguing for a sliding scale, this model simply expands the two traditional categories and thus maintains in tact the crucial moral insight that there is a difference between those who can be legitimately targeted and those against whom such targeting is a war crime. The ‘threat assessment’ and ‘use of force’ models from domestic police work do provide some useful practical insights in how this model might work, but it was also necessary to make some serious alterations to account for the differences between military operations (combat, peacekeeping or humanitarian) and domestic police work. Therefore, soldiers would first need to use a threat assessment model to ascertain whether the person in question is a combatant or a non-combatant (or is likely to be) and then determine within that category what kind of force or tactics are justified (i.e. neither endangering non-combatants or fellow soldiers, nor using more force than necessary) to achieve the end in question. The hope is that a distinction which maintains the moral difference, but also has some flexibility as to how force is to be applied – especially in different kinds of military situations – will encourage soldiers to continue to maintain and implement the distinction, despite the complexities and ambiguities of the contemporary battlefield.

NOTES