Final Draft
Adam C. Gastineau
07/2011
Targeting Targeted Killing
By: Adam C. Gastineau
In order to have a coherent discussion about the moral permissibility of targeted killing
we must first clearly define what we mean when we use the term. I will propose that
targeted killing is best defined as the authorized, intentional killing of a specific
individual or group of individuals who are targeted on the basis of their role in a given
armed conflict.1 In the first section of this paper I will unpack this definition a bit further.
I will argue that the definition I have offered allows us to incorporate a broader range of
acts that could be considered as acts of targeted killing. Other definitions imply that acts
of targeted killing only occur during military conflict and can only be carried out by state
actors. However as international terrorism, humanitarian intervention, and other
developing security concerns blur the lines between civilian and military paradigms of
conflict a definition that can cross these lines is needed. By widening the scope of the
terms ‘combatant’ to include those in task-defined ‘combatant’ roles and ‘armed conflict’
to include conflict in both military and at least some civil contexts I am attempting to
construct a definition that allows for this blurring effect and is at the same time is of some
use in determining when such acts might be morally permissible. I will then argue that
this definition is beneficial as it clearly distinguishes targeted killing from similar
instances of intentional killing such as assassination and extra-judicial execution. 2
Distinguishing targeted killing from these other acts of intentional killing serves several
purposes. First of all, it allows us to evaluate targeted killing’s moral value without any
of the normative baggage carried by the terms ‘assassination’ and ‘extra-judicial
execution’ much in the same way that distinguishing an instance of killing from an
instance of murder does. 3 Drawing this distinction also acknowledges that targeted
killing, assassination, and extra-judicial execution are morally justified, if in fact they are
morally justifiable, by different criteria. These differences are important when trying to
determine if a particular act of intentional killing can be morally justified. I will address
these criteria and problems that arise from them in the final section of the paper.
1
This definition is similar to that proposed by Steven David David, S. R. (2003). "Israel's Policy of
Targeted Killing." Ethics & International Affairs 17(1): 111-126 but differs in that it does not specify
explicit government authorization.
2
Defined by Amnesty International as ‘[a]n extrajudicial execution is an unlawful and deliberate killing
carried out by order of a government or with its acquiescence. Extrajudicial killings are killings which can
reasonably be assumed to be the result of a policy at any level of government to eliminate specific
individuals as an alternative to arresting them and bringing them to justice. These killings take place
outside any judicial framework.’ International, A. (2003). Israel and the Occupied Territories: Israel Must
End its Policy of Assassinations, , Amnesty International.Note 1.
3
Murder is generally considered to be ‘wrongful’ in some way. If all killings were ‘murder’ then all
killings would, necessarily, be wrongful. This would then pre-suppose the moral value of any killing
regardless of circumstance, and make the term ‘murder’ superfluous. Pacifists aside, most would find this
to be a gross overgeneralization, and certainly unhelpful in normative discussion about killing.
1
Final Draft
Adam C. Gastineau
07/2011
Defining Targeted Killing
I have chosen to define targeted killing as the authorized intentional killing of a specific
individual based on their role in an armed conflict. In this section I will attempt to clarify
some issues raised by this definition before moving on to illustrate how acts of targeted
killing, so defined, are clearly distinguishable from other acts of intentional killing such
as assassination and extra-judicial execution.
Authorization Part One
The first of these issues is the meaning of the term ‘authorization’. It is generally thought
that in cases of war, a proper authority must give the order to engage in warfare. If
authorization is required to commit a group to war it seems just as necessary that acts
intended to eliminate a specific targeted individual would require similar authorization,
even during an armed conflict, before the act can legitimately be carried out. However,
targeting specific individual(s) in that conflict seems to be normatively different from the
killing of combatants in the conflict. Combatants killed during a conflict are not singled
out to be attacked in the same way as in acts of targeted killing, assassination, or extrajudicial execution. Even in cases where one might try to argue that they are (e.g. sniper
activity, sentry elimination etc) those targeted are simply targeted as ‘a’ combatant.
Specifically targeting individuals for a particular reason and then ‘hunting them down’
seems to require a different kind of authorization.4 Prior to the order being given there
would need to be some deliberation over which individuals would be appropriate to target
in this way. The authorization for attack would stem from this deliberation, which would
need to provide good reasons for the attack to take place.
In cases of targeted killing it is arguable what type of authority is necessary to authorize
the act. Altman and Wellman suggest that a new institution be set up that would have the
power to authorize assassinations or targeted killings (Altman and Wellman 2009).5
However, I do not see that such an institution would be necessary (or at least as
necessary) to morally authorize instances of targeted killing, as Altman and Wellman
seem to think. By their argument it appears to me that currently existing institutions, such
as the UN, and perhaps international military institutions (NATO etc),6 that are able to
authorize humanitarian intervention could authorize acts of assassination using Altman &
4
As I will come to in a moment, the reasons one might be targeted differ in cases of assassination, extrajudicial execution, and targeted killing. Those targeted in assassinations are targeted on the basis of their
prominence or the political purpose of their death. Those targeted in extra-judicial executions are targeted
out of retribution for some past act. Those targeted in acts of targeted killing are targeted on the basis of
their role in the conflict. What I mean by ‘role’ I will come to later in the paper.
5
Altman, A. and C. H. Wellman (2009). A Liberal Theory of International Justice. Oxford, Oxford
University Press.Chapt. 5, pg 118. In my opinion the authors, like many others, conflate assassination and
targeted killing. I will attempt to draw a distinction between the two in a later section.
6
I am unsure of the legal standing of a country unilaterally authorizing a humanitarian intervention.
Judging from the brief discussion the issue has been given in the literature that I have read for this paper it
would seem to be a matter of some controversy: a conflict between national sovereignty and the possible
existence of a legal responsibility to protect the innocent. As Altman and Wellman are presenting a moral
argument and not necessarily a legal one, I will stick to speaking only about the moral legitimacy of acts of
intentional or deliberate killing.
2
Final Draft
Adam C. Gastineau
07/2011
Wellman’s criteria. Altman and Wellman argue that armed intervention may morally be
carried out by one or more sovereign states against another state if that state has; (a) lost
its legitimacy by grossly violating the rights of its citizens and (b) the risk to the safety
and security of those citizens is not disproportionate to the rights violations that would be
averted if the outside state(s) intervened. They then take this argument further suggesting
that this same formula could be used to justify some acts of assassination if such an
assassination would bring about the same result. They even argue that in some cases
assassination would be preferable to humanitarian intervention as killing just one (or a
few) member(s) of the leadership would better satisfy the second condition, assuming it
would be successful in bring about the same result.
It seems that if a state(s) were granted the authority to use military force to oust such a
regime on the basis of the above two conditions, they could use that same authority to
eliminate those responsible for the unfair conditions if doing so would produce the same
result. After all, the targeted killing of a few responsible individuals would be much less
risky to the citizens of such a regime than a full-fledged armed invasion in may cases. In
cases where the harm to citizens could be disproportionately high, such action should not
be authorized, as it would fail to meet the second standard for armed intervention. One
might also argue that deposing an illegitimate government is normatively different than
killing some, or all, of its members. There is a normative difference between the two acts,
but if we assume that both of Altman and Wellman’s conditions have been met, it would
seem to be better to kill members of the government responsible for the harms done to the
populous than to kill innocent members of the populous7 in the course of an armed
intervention on a larger scale. It should be noted that such an act would not be
permissible if done as an act of retribution by Altman and Wellman’s standards. The
authorization of the act would come from the fact that those targeted had forfeited their
legitimacy through their gross violation of others rights and because it would protect
those ‘others’ from further harm without at the same time putting said ‘others’ at greater
risk.
Regardless of whether a new institutional body is necessary to delegate authority, it still
seems reasonable that an appropriate ‘authorizer’ would be similar to that proposed by
Michael Walzer in Just an Unjust Wars. (Walzer 1977) The model put forward by Walzer
would place ‘proper authority’ with the governments of state actors in most cases but
could allow non-state actors to derive the necessary authority from the justice of their
cause. This differs from the definition of targeted killing proposed by David (David
2003), which explicitly requires government authorization. I will elaborate further on
governmental authorization vs. some other form of ‘legitimate authorization’ in my
discussion of extra-judicial execution below. What exactly a legitimate authority should
look like will be discussed further in drawing that distinction, but for now it is worth
noting that in order for targeted killing to be plausibly justifiable under some form of Just
War Theory an ‘appropriate’ authority would be required to give permission to carry out
the act.
7
Either directly (‘collateral damage’) or through lack of resources or infrastructure destroyed or blocked by
combat.
3
Final Draft
Adam C. Gastineau
07/2011
Putting the “Intent” in Intentional Killing
In order for a particular act to be considered targeted killing that act must be intentional.
By intentional killing I mean an act of killing that incorporates a prior intent to carry out
the act by both the individual who decides to carry it out or the authority that authorizes
the act. This reflects the idea of Intentionality laid out by John Searle’s work in
philosophy of mind. (Searle 1983)
I will not go deeply into Searle’s philosophy of action here, but I will highlight several
distinctions that he draws between intention-in-action and prior intention, and then
differentiate prior intention from reason(s) for action. I will also show how each of these
states causally relates to the others in bringing about the acts we are concerned with here.
For Searle all actions require two things, some sort of Intentional content and the physical
act. The Intentional content can take one of two forms in cases of intentional action and
sets the conditions of satisfaction that the physical act must meet to be considered an
intentional action. The Intentional content may be a prior intention or what Searle calls
“intention-in-action” which brings about a ‘spontaneous’ act. This intention-in-action is
inseparable from the physical act itself but sets the conditions of satisfaction for the
physical act. To use Searle’s example of raising one’s arm: one does not generally think
to themselves that they are going to raise their arm prior to doing so, one simply raises
one’s arm. The physical action of arm going up is the condition of satisfaction and if the
arm does not go up we say one has ‘tried and failed’ to raise their arm.
Such ‘spontaneous’ intentional actions are distinct from actions with prior intent, which
Searle refers to as “prior intentional actions”. These actions are those in which we have
formed some prior intent and the action is carried out by means of fulfilling the
conditions of satisfaction set out in the prior intent. To go back to our basic example of
raising one’s arm, one may form the intention to raise one’s arm “I will raise my arm”
prior to acting. This prior intention determines the conditions for satisfaction (i.e. that I
raise my arm). An act has a prior intention if and only if that prior intention causes the
physical act as a means of fulfilling that prior intention. So, briefly put, a prior intentional
act is one in which one has a prior intention, which causes an intention-in-action which
causes a particular physical act by means of bringing about the prior intention. So, if I
form a prior intention to raise my arm, and then forget about that prior intention, but later
in the day raise my arm (intention-in-action causes the physical act of my arm going up)
we do not say that the raising of my arm was a prior intentional act, as the prior intention
did not cause the intention-in-action, and so the intention-in-action did not bring about
the act as a means of fulfilling the prior intention.
The causal chain must also come about in the right way. To reiterate the example above
more generally, I may form a prior intention to do act ‘x’, then forget about my prior
intention but still intentionally perform act ‘x’ spontaneously as a result of intention-inaction. In such a case we can say the action was intentional, as intention-in-action
provided the necessary Intentional content to cause the act by means of carrying out that
intent, but the prior intention to do act ‘x’ did not cause the intention-in-action, so the
action was not a prior intentional action. Searle uses a case of intentional killing to
4
Final Draft
Adam C. Gastineau
07/2011
illustrate another way in which this causal chain might break down. Bill forms an
intention to kill his uncle. On his way to kill his uncle he is so excited by his prior
intention that he fails to see a man in a cross walk and hits the man with his car,
accidentally killing the man. The man turns out to be Bill’s uncle. We do not say that this
act is an act of intentional killing here because the prior intention did not properly cause
the intention-in-action (there is no intention-in-action in this case, as striking the
pedestrian was accidental) to cause the killing of Bill’s uncle. Bill did not intend to run
over anyone in the crosswalk prior to or during the act, therefore we couldn’t say that his
prior intention to kill his uncle caused the act in the proper way.8
In cases we are concerned with we can often point to multiple prior intentions that
operate as means and ends. I may form the prior intention to do A by means of doing B,
C, D, and E. B, C, D, and E are caused by the prior intention to do B, C, D, or E or
simply be caused by intention-in-action, but all are caused by, and incorporated into, the
primary prior intention A.9 Think of a soldier lying in ambush.10 He has established a
prior intention to ambush and kill enemy combatants. This prior intention includes other
prior intentions: such as to go to a particular place for the purpose of killing enemy
combatants and to lie in wait in that place. In cases such as this the prior intention to
ambush and kill enemy combatants forms the Intentional content of the act and defines
the conditions of satisfaction for the particular act. If the soldier does not kill the enemy
we say that he has tried and failed to do so. This example also illustrates a case in which
prior intentional actions may be multi-layered. The soldier’s prior intent to ambush and
kill enemy combatants incorporates at least two other prior intentions (going to the
particular place as a means to ambush and kill the enemy, waiting for a period of time in
that place as a means to ambush and kill the enemy), and also intentions-in-action (lying
down in high grass, the act of putting one foot in front of the other walking to the
location, pulling the trigger to fire the weapon etc). All of these are incorporated within
the prior intention to ambush and kill enemy combatants.
There is one further distinction that needs to be drawn here before moving on. Prior
intentions are distinct from reasons that apply to them. These reasons may inform,
motivate, or even justify the prior intention and in doing so help set the conditions of
8
There are other cases in which the prior intention caused an intention-in-action but the intended outcome
was caused by some instance of deviant causation rather than directly caused by the intention-in-action, (an
assassin’s bullet misses the target, but stampedes a herd of buffalo killing the target) but I will leave such
cases aside in the interests of space and be satisfied to simply say that the causal chain linking prior
intention, intention-in-action and action must come about in the ‘right way’.
9
For a less militant example, when I form the intention to get a cup of coffee various other intentional
actions are part of this intention. I stand up, I open the door to my office, I walk down the hallway etc. In
this case all of the actions named are caused by intention-in-action caused by my prior intention of getting a
cup of coffee, though in some cases other prior intentions may be included. There is nothing wrong about
forming a prior intention to stand up, open my door, walk down the hall etc, it just seems strange to do so in
this case. Regardless, all of the previous actions are caused in this case by the primary prior intention of
getting a cup of coffee.
10
By using this example, I am not claiming that such an ambush is an act of targeted killing. As I will
argue later, such an action has the required prior intention, but that prior intention lacks the ‘targeted’
specificity necessary to make the act one of targeted killing.
5
Final Draft
Adam C. Gastineau
07/2011
satisfaction that the prior intention lays out. In short, these reasons form the basis for the
prior intention. However, the reasons are nonetheless separate from the prior intention.11
These reasons cause, or at least shape, the prior intention of an act. If asked why one
raised their arm one might explain that the sun was in their eyes, that this made them
uncomfortable, and so believing their hand would block the sun, they raised their arm.12
Likewise using our example above, the soldier may have reasons for forming the prior
intention to ambush and kill enemy combatants. He may consider it his job as a soldier to
do so. He may wish to avenge the deaths of others in his unit. He may believe that it is
tactically correct given the situation in the area. He may have been ordered to do so. All
of these reasons contribute to the soldier forming the prior intention to carry out the act.
This process of formulating the prior intention to kill a particular target is necessarily
deliberative, so it may be better to refer to acts of assassination, targeted killing, and
extra-judicial execution as a subset of acts of intentional killing, perhaps by referring to
such acts as acts of deliberative killing, rather than acts of intentional killing. Both acts of
deliberative and intentional killing are intentional acts, but acts of intentional killing may
be undertaken, as in acts of self-defense, without reference to any prior intention or
deliberative process and so would not necessarily be acts of deliberative killing.
Regardless of how we choose to refer to these acts as a whole, an act of targeted killing,
like assassination and extra-judicial execution, is a reason based complex prior
intentional act. In cases of targeted killing, the term ‘targeted’ also bolsters the argument
that targeted killing is a prior intentional act, as one cannot target something without
having some prior intention to hit it. The requirement of proper authority further supports
defining targeted killing as a reason based complex prior intentional act. As previously
mentioned, the act must be caused by a prior intention to carry out the act. This prior
intention must be authorized in cases of targeted killing. This authorization is based on
reasons that form the basis of the prior intention that is to be authorized (e.g. killing the
individual targeted). Once given the authorization also provides a reason for the prior
intention. What we have then are two distinct reason-based fully intentional actions, in
which one, the act of authorization, becomes a reason informing the prior intention of the
second, the act that has been authorized. This can also be seen in our ambushing soldier
example. The soldier may have a variety of reasons for forming the prior intention to
ambush and kill enemy combatants, including his being ordered or otherwise authorized
to carry out the act. This prior intention (ambushing and killing the enemy) then acts as a
reason for the other two prior intentions incorporated in the larger prior intention to
ambush the enemy. The soldier has gone to a particular place and waited in that place for
the purpose of ambushing and killing the enemy. Such an act is then a complex prior
intentional act. If we stipulate that our soldier has been authorized to kill a specific
11
I may have reasons to get a cup of coffee, I am sleepy, I am sick of staring at the computer screen, I want
to get out of my office, etc, but these reasons, while they may motivate, inform, or justify my prior
intention are nonetheless distinct from it and form part of the deliberative process that brings the prior
intention about.
12
This explanation could be offered for either a ‘spontaneous’ intention-in-action or a prior intention to
raise one’s arm. The reasons would be given after the fact as no prior intention caused the physical action.
In cases of prior intentional action reasons that form the prior intention come before the act and shape the
prior intention to carry out the act.
6
Final Draft
Adam C. Gastineau
07/2011
individual (perhaps bin Laden) we have a model of targeted killing that shows the act to
be a complex prior intentional act. The fact that a combatant has been authorized to shoot
bin Laden or some other unique ‘target’ in particular provides specific Intentional
content, which differentiates targeted killing from other types of killing that may occur
during armed conflict. For example, if during fighting in Tora Bora Osama Bin Laden
had been killed in the bombing of a Taliban position this would not be considered to be
an act of targeted killing. The Intentional content of the act was the destruction of enemy
position (i.e. Al Qaeda or Taliban combatants in general), not Bin Laden specifically.
As mentioned above “targeted” implies a requirement of specificity. In order for the act
to have the required Intentional content, the act must have an Intentional object (i.e. the
set of conditions of satisfaction) at which that Intentional content is directed. Specificity
in acts of targeted killing may be determined in a variety of ways. The definition of
targeted killing I am proposing indicates that this specificity is determined by the
person’s role in a particular conflict (either their position itself, or the task or job they
perform).13 However, some have specified the target by name or other similar identifiers.
Michael Gross highlights this when he refers to targeted killings as ‘named
killings’(Gross 2006) and argues that this specificity removes the moral equality of
combatants.14 I do not find this argument particularly persuasive, but even we do agree
that Gross is correct, his view does nothing to undermine the idea that targeted killing
must be directed at a specific individual to be considered an act of targeted killing, how
ever that individual or group may be identified.15
Identifying Identity
Identifying the particular target does raise the question of what exactly constitutes the
identity we are targeting in cases of targeted killing. It also raises the question of how
wide a scope that identified target might cover. It does not seem that targeting a large
number of individuals seems to be specific enough to be called ‘targeted’ yet we could
still form a prior intention to target a large group of people. I will take the first question
first.
13
It has been noted that bin Laden may have only had a minimal role at the time of his death, and so
perhaps he was targeted for reasons other than the role he was fulfilling at the time. While this is a perfectly
legitimate issue to raise in this case, it is not to the point I wish to make here. I am attempting to show on
what basis one is targeted in acts of targeted killing according to my definition: i.e. one’s role in a particular
conflict. Whether bin Laden’s role at the time of his death was sufficient for him to be legitimately targeted
in an act of targeted killing is a different question than the one I wish to address here, which is “On what
basis is one targeted in an act of targeted killing?” Whether or not bin Laden was targeted on this basis is
an interesting, but separate, question I will not attempt to answer at this point.
14
In “Assassination and Targeted Killing: Law Enforcement, Execution or Self-Defense?” Gross puts
forward the argument that by naming a particular combatant as a target in armed conflict one robs them of
their moral equality as that by naming them the name-er, at least implicitly, declares that the individual
named is particularly deserving of death over other un-named combatants.
15
Gross himself rejects this argument after being taken to task on it by David Statman, who points out
(among other things) that the reason names are used in these cases is because there is no other way to
accurately identify those being targeted. Naming the individuals has little if anything to do with their guilt
or innocence. It is simply a means to identify targets that wear no uniform or insignia marking them as
combatants. Gross, M. L. (2010). Moral Dilemmas of Modern War : Torture, Assassination, and Blackmail
in an Age of Asymmetric Conflict. New York, Cambridge University Press
7
Final Draft
Adam C. Gastineau
07/2011
It seems that identity can be thought of in one of two ways. We can focus on an
individual entity (Barack Obama) and regard any definite descriptions that may point to
that entity as changeable in the case that such a description no longer applies to that
particular entity or otherwise ceases to be true (President of the United States). In such
cases our primary focus is on the entity itself and not any definite description we may use
to identify them. On the other hand, we may shift our focus to the definite description
(President of the United States) and then try to determine which entity actually fits that
description (For the moment Barack Obama but after the next election it could be
someone else). In this case we look from the definite description to the individual entity
rather than from the individual entity to the definite description. In the former case where
our focus is on the individual entity we generally ‘name’ the entity while understanding
that that particular entity may also be referred to by a definite description that applies to
it. In the latter case we refer to the definite description keeping in mind that this definite
description may indicate one entity at one point in time, but a different individual entity
at another. For example, “The Leader of Group Z” is the definite description of a
particular entity ‘x’. In this case the particular entity ‘x’ is named Bill. I can form a prior
intention to kill The Leader of Group Z, Bill. The Intentional object of this intention may
be to kill the entity ‘x’ called Bill, or it could be to kill The Leader of Group Z, a definite
description that, for the moment at least, happens to be an entity ‘x’ called Bill. If the
prior intention is to kill the entity ‘x’ called Bill then the focus is on the entity itself
regardless of other definite descriptions that may refer to it. If Bill changes his ‘name’ in
an attempt to avoid being targeted, either by no longer going by Bill or by stepping down
as leader of Group Z so that definite description no longer applies, the entity specified by
my Intentional object does not change. I am still targeting entity ‘x’ that was once called
Bill or The Leader of Group Z but is now called something else. If my Intentional object
in this case were to kill The Leader of Group Z and Bill stops being The Leader of Group
Z, then my Intentional object would shift to a different entity but would stay with the
same ‘name’. So, if entity ‘y’ Bob becomes the new Leader of Group Z, killing entity ‘y’
would become my new Intentional object.
As I will explain in a moment, cases of targeted killing are concerned with this later
formulation of identity, as targets of targeted killing are selected on the basis of their role
as defined by their ‘task’ in an ongoing armed conflict. This is in contrast to acts of
assassination, which are could be concerned with either of the two, and to acts of extra
judicial execution, which are only concerned with the former formulation of identity. As
we will see, acts of assassination differ from other types of intentional killing such as
murder because they target prominent individuals. (Knoepfler 2010) However,
assassinations are also said to be intentional killings carried out for political purposes.
(Kasher 2005) For example one might form a prior intention to kill Joseph Stalin. One
might argue that Stalin could not have committed the acts he did without having first
been in the prominent role he occupied, which focuses on the later formulation of
identity, or that killing him would serve a particular political purpose because his
successor would not do the same things he did because he as an entity was so ruthless,
bloodthirsty or what have you, which focuses on the entity of Stalin himself. It seems that
assassination could use either of the two formulations of identity to identify a target. If
8
Final Draft
Adam C. Gastineau
07/2011
the individual targeted is not prominent or their death will serve no political purpose, the
assassin would not identify that individual as a target, but in this case, it seems both
apply. Given this is the case, the individual targeted could be either a named individual
entity (Joseph Stalin) who also happens to meet the definite description of a prominent
role, (the leader of an authoritarian state, in this case the USSR) or the individual could
simply be whoever happens to meet that definite description included in the prior
intention of the assassin, such as the leader of the authoritarian state, whomever that
might be at the time.
In cases of extra-judicial execution, the targets are selected on the basis of desert and
killed as punishment for some past act. This seems to imply that the target is the entity
that carried out, or is believed to have carried out, a particular act or acts rather than the
task defined role that the individual may occupy. It might be argued here that one might
punish a group for a particular act by eliminating a member in a prominent role.
However, as we will see in the following sections distinguishing these three categories of
intentional killing (targeted killing, assassination, and extra-judicial killing) from one
another, if one hopes to offer a normatively neutral definition of extra-judicial execution
in which the act is not wrongful (or righteous) by definition, the focus must stay on the
retributive nature of punishment. Retribution seems to necessarily focus on an individual
entity that has done something that is deserving of punishment insofar as the act could
only be justifiable in cases where the individual killed was guilty of committing a wrong
that might justify his or her death. Killing an individual in retribution for some past act
does not seem justifiable at all unless one could plausibly show that the individual killed
had done something to deserve to be killed. If I wrong someone while acting in a
particular role, and they kill the next occupant of the role who has done nothing, they act
unjustly even if I suffer some harm, and thus some punishment for my act, as a result.
This seems to keep the focus on the individual entity that committed the act, rather than
the definite description of role that they occupy. This would hold true even in cases
where the two formulations of identity overlap to some degree.
However, there is still a question of the degree of specificity required. I have stated
previously that the act of attacking a position held by Al Qaeda/Taliban combatants and
‘accidentally’ kills Osama Bin Laden does not have the proper Intentional content to
qualify as an act of targeted killing. What about cases where the pilot has been advised
that the attack is targeting ‘high-level Al Qaeda operatives and leadership’? This would
allow the pilot to hold killing Al Qaeda leadership as the object of their intent. In this
case it would seem that the pilot would have the proper Intentional content for it to be a
case of targeted killing, but is that intentionality specific enough? Is it ‘targeted’ enough
to qualify as an act of targeted killing?
The short answer to this question is ‘yes’. The act would qualify as an act of targeted
killing of ‘high-level Al Qaeda operatives and leadership’, which may include Bin Laden,
but the act is not an act of targeted killing against Bin Laden. While Bin Laden can be
identified as part of the group that is ‘named’ as Al Qaeda leadership, he, as an individual
entity, cannot be. If the prior intention was to kill the entity identified as ‘the Leader of
Al Qaeda’ then we could say that the killing of Bin Laden was an act of targeted killing,
9
Final Draft
Adam C. Gastineau
07/2011
but because the name of the role targeted applied to several entities in this case we cannot
claim that the act identified Bin Laden as the target in the Intentional content of the act.
Again, we should refer back to Searle’s formulation of intentional actions. The pilot in
this case has a specific Intentional object: to kill Al Qaeda leadership in his or her attack
in a particular way. 16 Assuming that the object of intent was realized, i.e. Al Qaeda
leadership and/or high-level operatives were killed by the attack; one could call the
killing intentional by Searle’s formulation of an intentional act. However, one could not
say that killing Bin Laden himself was one of the conditions of satisfaction determined by
the Intentional content of the prior intention.
There appears to be a slippery slope problem here, but I think looking at the previous
discussion of intentional acts and identity can mitigate it enough to be eliminated in the
deliberation process required for authorization. If the intentional object grows too broad,
say to eliminate all Taliban and Al Qaeda combatants at a particular location, we begin to
lose the necessary specificity required to identify a particular intentional object in the
necessary sense. However, ‘High-level Al Qaeda operatives and leadership’ names a
specific group or perhaps a specific group of persons by virtue of their occupancy of taskdefined roles, that is, leadership and high-level operatives only, which is specific enough
to link specific individual entities with the definite description ‘high-level Al Qaeda
operatives and leadership’. Saying that one intends on ‘targeting’ all members of Al
Qaeda in a particular context or location does not seem to be specific enough to link
particular entities with the definite description. It might be said that the description is not
‘definite’ enough to form a prior intention with a specific intentional object based on
identity in either sense of the term in such cases.
This then raises one other problem that needs to be briefly addressed. How can we know
what a particular agent’s intentional object is in these cases? In short, it might be objected
that we don't know whom some attacker is aiming at. It has been stated that in order for
an act of killing to be an act of intentional killing, that act must include both an intention
to carry out an act and thereby realize a particular intentional object, in this case killing a
particular individual or group of individuals. I have also stated that in cases of targeted
killing the agent must have a prior intent to bring about the object of their intent, as
without this prior intent it would be difficult (if not impossible) for the act to be
deliberated upon and authorized. If the pilot includes killing Bin Laden as part of his or
her object of intent when setting out on the mission to kill Al Qaeda leadership, would
this mean that they had carried out an act of targeted killing against Bin Laden?
To answer this question, we need to refer to what object of intent was authorized prior to
the attack. In this example what has been authorized is a targeted attack to intentionally
kill a specific group of individuals based on their role within a given armed conflict. This
would seem to fit the definition I have proposed for targeted killing. However, I have also
stated that if Osama Bin Laden were killed in this attack, it would not qualify as an act of
targeted killing against Bin Laden, because killing Bin Laden specifically was not
identified as part of the object of intent that was authorized. The pilot’s (or soldier’s, or
16
Say by striking a particular compound with a missile fired from a drone.
10
Final Draft
Adam C. Gastineau
07/2011
any other agent’s) object of intent is not what determines the specific target of the action,
rather it is the object of intent authorized prior to the act that determines who is or are
targeted in cases of targeted killing. If the pilot sets out on his or her mission to kill Al
Qaeda leadership and bombs a wedding party instead, we say they have failed to bring
about the conditions of satisfaction incorporated in the Intentional content of the prior
intention and thereby fail to realize the object of intent and thus have failed to carry out
an act of targeted killing. Likewise, if that pilot is sent out on a mission to kill Bin Laden,
and they fail to do so, we say they have failed in a similar way. However, if the pilot sets
out on a mission to kill Al Qaeda leadership and succeeds in killing some leadership
figures, but does not kill Bin Laden, we cannot say that the pilot failed to realize the
object of intent that they were authorize to fulfill. The pilot has committed an act of
targeted killing against Al Qaeda leadership. The fact that Bin Laden was not killed in the
attack, even if the pilot included that object in his or her own object of intent, does not
mean the attack has failed because the object of intent that was authorized was achieved.
Who is Targeted in Targeted Killing?
If we allow that targeted killing must involve the intentional killing of a specific
individual or group of individuals we must answer the question of how that particular
individual(s) might be targeted. In cases of assassination targets are selected for either
their prominence or their political value depending on whose account you find most
credible. Many scholars such as Kasher & Yadlin (Kasher 2005) differentiate
assassination from other forms of intentional killing by saying that assassination is
carried out for some political purpose 17 , while others such as Knoepfler cite the
prominence18 of the target as the differential factor rather than the purpose of the attack
(Knoepfler 2010). However, both of these criteria seem to miss the point when it comes
to instances of targeted killing.
Rather than basing the ‘who’ of targeted killing on the target’s prominence or the political
purpose that the target’s death might serve it seems more accurate, both empirically and
normatively, look to the role that the targeted individual(s) play in a given conflict. This
‘role’ could have one of two meanings in this context: One’s position in a hierarchy or
within a particular societal or organizational context that points to the status or power one
holds in that particular organization or society, which would seem quite similar to the
prominence argument, or the task-defined role a person is in. A task-defined role is
defined by a particular pattern of actions (i.e. tasks) indicative of a role that one has or is
17
Though they define “political” in very broad terms to include the duty the state has to protect its citizens.
Thus, killing an individual because that individual constituted a threat to the citizenry of the state is killing
for a political purpose.
18
Knoepfler defines prominence by position that is “someone who is a leader of some sort or is
particularly famous and important” Knoepfler, S. (2010). "Dead or Alive: The Future of U.S. Assassination
Policy Under a Just War Tradition." New York University of Law & Liberty 5(457): 457-499. However, it
seems to me that this prominence could also depend on the job or task a particular individual might be
assigned. A skillful bomb maker may be prominent in this sense while not having a particularly prominent
position within the organization or being very famous outside of that organization. Similarly, a suicide
bomber on his or her way to carry out an attack becomes very prominent to those trying to stop them, even
though they are in a relatively low position in the hierarchy of their organization and are unlikely to be
famous at all, except possibly after the attack.
11
Final Draft
Adam C. Gastineau
07/2011
reasonably expected to have on the basis of that pattern of actions.19 I will use the term
‘positional role’ when referring to the first and ‘task-defined role’ when referring to the
second. The term positional role may refer to one’s place in a hierarchy but may also
simply refer to the identity of a target by means of the position an entity holds within the
organization or society they belong to. ‘Suicide bomber’ or ‘police officer’ are positional
roles that name the position held and may or may not focus on the position held within a
particular hierarchy. In cases of targeted killing however, the deliberation on who to
target should focus more on the task-defined role of the individual(s) to be targeted. In
cases of targeted killing one is targeted as a combatant or performing a particular
combatant-like task that constitutes an ex ante threat within a given conflict.20
These two types of roles are not necessarily independent of one another, as one’s
positional role is often influenced by the task-defined role associated with the position,21
but it is possible for one to occupy a certain positional role without having a particular
task or job that the role might imply. For example, a police officer that works in an
emergency call center has a positional role as police officer. They may be sworn, wear
the same uniform, and/or have undergone similar training to a patrol officer, but the taskdefined role that they have is quite different. To use another example that may have
greater bearing on the current discussion, the political leader of a given country may have
a positional role as the ‘commander in chief’ of that nation’s armed forces, but it is
unclear as to whether or not that individual has a task-defined role that would qualify that
person as a combatant in the same way that a military officer, even a very high ranking
one, would be. This could also be said of the political leadership of non-state actors such
as Sinn Fein, Hamas, or Hezbollah and so will become quite relevant when trying to
determine the justifiability of acts of targeted killing.
Finally, according to my definition acts of targeted killing only occur during armed
conflict. By armed conflict I mean something more general than ‘war’, which could be
plausibly defined as a state of armed conflict between nation-states or other political
factions within those states. An armed conflict occurs when one party, which may be a
group or an individual (in this case acting qua member of some group or organization),
imposes or attempts to impose, their will upon another by force of arms, and the other
party has the ability to resist in some way. While this is not the standard formulation of
‘conflict’, it seems very similar to cases in which one nation-state, group, or individual
aggresses against another. The aggressor seeks to impose its will on those it aggresses
against and those attacked have the choice to capitulate, flee, or fight the aggressor. The
conflict exists for as long as one side attempts to resist the will of the other either
19
For example, if one’s pattern of action in a particular conflict consists of arming oneself, helping other
combatants by storing supplies and equipment of their own freewill, helping to plan and carryout attacks
etc one is considered to be acting in the task-defined role of a combatant in that conflict.
20
It should be noted here that a task-defined ‘role’ is not the same as a ‘job’ or ‘occupation’. It may be a
manner of acting in a particular context that ends as soon as the context changes (e.g. the conflict ends) or
the actions cease (e.g. the agent surrenders or is captured). It is also not a ‘one-off’ act but a pattern of
actions indicative of a particular role in a given context.
21
Firefighters do fight fires after all.
12
Final Draft
Adam C. Gastineau
07/2011
passively, by fleeing, or actively, by fighting back or otherwise attempting to stop the
attack.
This formulation would include everything from conventional and asymmetric warfare,
to terrorism and some armed criminal enterprise, while excluding cases in which one
party is entirely at the mercy of the other, such as torture. A hostage taker attempting to
escape from police by threatening hostages or police with violence is engaged in armed
conflict with both the hostages (as long as they have some means to resist) and the police.
The hostage taker is attempting to use violence, or the threat of violence, to force both the
police and his or her hostages to comply with their demands. By doing so this individual
is taking on a role as a combatant within the conflict insofar as they are actively exerting
force and constituting a threat to both the hostages and the police attempting to protect
them. In this case the role ends as soon as the conflict ends, but it is not a ‘one off’ action.
The hostage taker engages in a large number of actions within the conflict, exerting
physical force, and threatening, making demands ‘or else’ etc. All of these actions are
incorporated into their task-defined role as a combatant, which lasts until they cease the
pattern of action that indicates that one is occupying that role. Should they stop acting in
this role, by surrendering or otherwise ceasing to use physical force, threats etc, they
would no longer be morally liable to attack. However, for as long as the individual
continues to act in this task-defined role, they are liable to be attacked, just as a
combatant acting in the task-defined role of a soldier or sailor or Marine in conventional
conflict is liable to attack until they surrender or are incapacitated and so cease to be a
threat to the opposing side. By the definition I am proposing a police sniper killing the
hostage taker (at least in this case) would be an instance of targeted killing in that the act
would be an authorized (prior) intentional act, identifying and specifying the target by
means of the task-defined role the individual held as a combatant in an armed conflict
and the justifiability of the act would depend on principles of discrimination22 and
proportionality.23
Similarly, a sniper randomly killing innocent people in a public park is engaging in
armed conflict with those he or she is firing upon and is taking on the task-defined role of
22
It might not be justifiable for the sniper to ‘shoot the hostage’ in order to get a clear shot at the hostage
taker for example.
23
Similarly, it would not be a justifiable instance of targeted killing if the hostage taker were threatening
the police and hostages with a rubber gun, though it might be excusable if the sniper was unaware that the
gun in question was a fake. A more difficult case might be one in which the hostage taker is torturing a
helpless victim and is shot by a police sniper in order to stop the torture. In this case, while we can say that
the torturer is liable to attack, we cannot say that the torturer is in armed conflict with their victim, (as the
victim has no means of resisting) nor can we say that the torturer is in armed conflict with the police, at
least in any conventional sense. I am tempted to argue here that because of the role police officers play in
society attacks or threats against innocents, even helpless innocents, bring the attacker into armed conflict
with the police: perhaps because the police offer a legitimate form of armed resistance on behalf of the
members of that society. However, proving this intuition would require a considerable amount of argument.
I do not have space to address the issue properly here and so for now entertain the idea that in some
instances police snipers and others agents in similar positions may not be considered as agents who can
engage in acts of targeted killing as the aggressing party is not in ‘armed conflict’ with the police in the
proper sense.
13
Final Draft
Adam C. Gastineau
07/2011
a combatant.24 The sniper is using the weapon to impose his or her will; in this case to
cause fear or death, on those people he or she is firing at. Again, those being aggressed
against are faced with a choice of capitulating and hoping they won’t be killed, fleeing or
otherwise attempting to avoid the attack, or fighting back against the attacker. While in
this situation one of these options is tactically preferable to the others, the conflict
between the shooter and his or her targets occurs nonetheless. Whether or not an incident
such as this would qualify as an act of targeted killing would depend on the sniper
having, or not having, a prior intent to kill a specific individual in the park that motivated
the action he or she then undertook to bring about that prior intention.
Advantages
Defining targeted killing as an authorized intentional killing of a specific individual based
on their role in a given armed conflict offers at least three advantages. First of all, we are
able to differentiate it from other acts of intentional killing that may occur during armed
conflict such as assassination and extra-judicial execution (and, arguably various forms of
sniper killing). As will be shown in the next section, acts of assassination do not require
authorization, are carried out only against prominent individuals, and, arguably, are done
only for political purposes. In cases of extra-judicial execution, the definition states that
such acts require government authorization, making it impossible for a non-state actor to
carry out such an act. The definition given also implies that the act is carried out in
retribution for a crime or wrong committed by the intended target. Targeted killing, as I
have defined it, differs from this, as it is simply a method of attacking one whose taskdefined role makes them liable to attack during an armed conflict. The act is not
necessarily motivated by retribution, but rather by the tactical or strategic importance of
eliminating that particular combatant at that particular time.
Secondly, because the act can be distinguished from other acts of intentional killing that
may be difficult to justify, we are able to avoid the negative moral ‘baggage’ associated
with such acts and perhaps evaluate the normative value of such acts more easily. If we
can distinguish between the three acts we can more easily establish which normative
standards need to be applied in which circumstance. For cases of assassination normative
problems stem both from the subjectivity of the act, insofar as assassination does not
have the requirement of authorization, and from the ‘nefarious’ or ‘treacherous’ nature of
the act. This problem of treachery may have some root in the issue of authorization but is
also related to the fact that acts of assassination are not contextually defined. Such acts
may occur during conflict or peacetime and be carried out by anyone against any
prominent figure. The retributive nature of acts of extra-judicial execution raises
normative problems around how one might prove that the entity identified as the target is
(a) guilty of the wrong they are accused of and (b) that the wrong necessarily justifies that
entities death without first trying the individual in a court of law and thus allowing the
individual to defend him or her self or offer mitigating evidence.
24
This would not necessarily be an act of targeted killing unless the sniper had a prior intent to kill a
specific individual or group of individuals in the park. It could be that the sniper in this case simply
intended to kill ‘some’ people. As there was no specific intended target or targets, it does not seem that one
could call such killing “targeted”. Cases such as these would also most likely fail to meet the authorization
requirement.
14
Final Draft
Adam C. Gastineau
07/2011
Finally, the definition I have offered also allows us to incorporate the full range of acts
that could be considered as acts of targeted killing. Other definitions imply that acts of
targeted killing only occur during military conflict. However as international terrorism,
humanitarian intervention, and other developing security concerns blur the lines between
civilian and military paradigms of conflict a definition that can cross these lines is
needed. By widening the scope of the terms ‘combatant’ and ‘armed conflict’ I am
attempting to construct a definition allows for this blurring effect while at the same time
offering some guidance on when such acts might be morally permissible. While it may be
the case that targeted killings can not occur in some policing scenarios it does appear that
some conflicts in policing qualify as armed conflict by the above definition, broadening
the scope of the targeted killing beyond strictly military conflict.
Distinctions
In this section I will compare the act of targeted killing as I have defined it with other acts
of intentional killing with which it is often conflated, assassination and extra-judicial
execution. The following table illustrates the similarities and differences between these
three acts.
Assassination
Targeted Killing
Extra-Judicial
Execution
Intentional Killing
4
4
4
High Degree of
Specificity
4
4
4
Authorization
6
4
4
Prominence (Position)
4
6
6
Prominence (other)
4
6
6
Retributive
6
6
4
Role (Task)
6
4
6
Political Purpose
4
6
6
Armed Conflict
6
4
6
Nefarious/Treacherous
4
6
?
Why bother distinguishing between assassination, extra-judicial execution and targeted
killing? In many ways they are quite similar, and it could be argued that each may be
justifiable under certain circumstances. The answer to this question is, in a word, clarity.
15
Final Draft
Adam C. Gastineau
07/2011
Like the word ‘murder’ the terms assassination and extra-judicial execution carry heavy
negative moral connotations, and with good reason. Assassination is often distinguished
from other forms of intentional killing by referring to its ‘nefarious’ means. Assassination
is perceived as an intentional killing that is carried out by means of treachery, deception,
and deceit and because of this is automatically assumed to be a wrongful act or at least
produce wrongful outcomes. (Gross 2006) Similarly, extra-judicial execution is
considered, by definition, to be a form of intentional killing for retributive or punitive
purposes. (International 2003) This smacks of injustice in that it punishes the target
without proving that the target is guilty of wrongdoing, which implies that it must be
morally wrong in some way.
Just as incorporating the term ‘murder’ into any definition of assassination or extrajudicial execution taints that definition with the idea that it must be morally wrong,
(Knoepfler 2010) referring to acts of targeted killing as assassinations or extra-judicial
executions muddies the water from the start. If we are to produce a definition of targeted
killing that will be action guiding that definition must allow us to start from neutral
ground. Conflating assassination and extra-judicial execution with acts of targeted killing
implies a particular moral conclusion from the start, and so affects where our conclusions
will end up. To refer back to our bank scenario, if we say that the gunman was
“assassinated” or “executed” by the police sniper, this would imply to most people that
the sniper shooting the robber has done something morally wrong. However, because the
term ‘targeted killing’ carries less, if any, negative implications we are able to start to
evaluate the moral value of the sniper’s action from more neutral ground.
In the following sections I will distinguish the act of targeted killing as defined above
from the acts of assassination and extra-judicial execution. I will also attempt to show
that in cases where assassination or extra-judicial execution may be justifiable the act is
closer to the definition I have proposed for targeted killing than the concepts of
assassination or extra-judicial execution, as they are commonly understood. In such cases
referring to such acts incorrectly may confuse our judgment of such acts.
Assassination vs. Targeted Killing
The term assassination has had many definitions. However, there are some characteristics
that seem consistent across definitions. Acts of assassination are intentional killings. That
is, like acts of targeted killing, people who have a prior intent to kill a particular
individual carry them out. They are generally directed against prominent individuals and
carried out for political purposes. Finally, many who consider assassination to be a
wrongful act link this wrong, at least in part, to the nefarious or treacherous means by
which the act is carried out. Using this template, acts of assassination differ from acts of
targeted killing as I have defined it in three important ways: targets are not necessarily
prominent individuals nor are they necessarily chosen for political purposes, and the act
itself is not nefarious or treacherous. Acts of targeted killing also differ from acts of
assassination in that they only occur during armed conflict between the parties involved
in the conflict, and acts of targeted killing require legitimate authorization.
16
Final Draft
Adam C. Gastineau
07/2011
Prominence
Prominence can be determined in different ways. One can be prominent by way of the
task that they perform, (Lee Harvey Oswald) or one can hold a prominent position (i.e.
high level) within a given organizational context such as the President of the US, or one
can be otherwise prominent, e.g. a celebrity. As discussed previously during my account
of roles, these types of prominence can exist in tandem. Often this ‘positional
prominence’ exists because of the job or task one in such a position is expected to
undertake. The president of a particular country is prominent both because of their
position within the government of that country and because of the tasks that the president
is expected to carry out. However, these types of prominence are not necessarily
connected. It is possible for one to have positional prominence without having task
prominence. Most monarchs today have positional prominence. They are held in high
regard and are famous figures within their own countries if not internationally. However,
few if any have much task prominence when compared to their prime ministers or other
government officials. They do not generally dictate policy, and the tasks they undertake
are only prominent because of the position of the monarch doing the task, rather than the
other way around.
Prominence of this type is based on one’s position within a particular organization or
society but not necessarily a hierarchical position. That is, some celebrities are simply
famous for being famous. Celebrities of this type are not necessarily part of any
institutional hierarchy. One might argue that the fact that they are idolized places them
above others in a society hierarchically, but one cannot be ‘promoted’ to the rank of
‘celebrity’ in the same way one is promoted within a monarchy, business, or other
hierarchical institutions. Celebrities may be richer, more visible, or more popular than
others in a particular society, but it does not follow from this that anyone with the
position of ‘celebrity’ is necessarily placed higher up than others within that society in
any real sense. Despite the non-hierarchical nature of this form of prominence, to be a
celebrity is to occupy a particular ‘position’ within society. Paris Hilton might be an
example of a celebrity who has this type of non-hierarchical positional prominence.
There is one other form of prominence that would fall under this non-hierarchical or
‘otherwise prominent’ model of prominence. It seems quite plausible to say that one
becomes prominent as the result of some task or act that they carry out. For example, an
otherwise unimportant individual may become extremely prominent because of a task or
act they have performed or are expected to perform. Individuals such as Lee Harvey
Oswald have no prominence independent of the act of assassination he carried out. His
prominence has nothing to do with his previous position in society, but rather comes from
the fact that he carried out the ‘task’ of assassinating John F. Kennedy. The fact that he
carried out this particular task defines his prominent position (good or bad, high or low)
in society. Che Guevara would be another example of an individual with task
prominence. Similarly, groups may have task prominence that stems from the importance
or impact of the task that they do or have done. Firefighters, the 9/11 Hijackers, the Bali
9, etc. all have prominence not because of their position within society previous to
undertaking the act or task associated with them, but because of the task or acts they
carry out or have carried out in the past. This task prominence leads to position in society
17
Final Draft
Adam C. Gastineau
07/2011
and often their celebrity, but it is still the case that it is the task, rather than the position
prior to the task that confers the prominence on the individual or group in question. This
form of prominence differs from a task-defined role in that a task-defined role cannot
result from a ‘one-off’ act. A task-defined role is a pattern of actions indicative of a
particular role. Task prominence on the other hand may be achieved by one single act, as
in the case of Lee Harvey Oswald or the 9/11 Hijackers, or multiple acts, as in the case of
Che Guevara or firefighters. In cases where multiple acts are the source of one’s
prominence one may have task prominence, which can be said to derive from one’s taskdefined role. For example, Che Guevara had a task-defined role as a combatant and it was
the actions he undertook in this role that gained him the prominence he had at the time of
his death. In cases of assassination it would be the prominence (and possibly the political
purpose, but I will come to that in a minute) that would mark him as a target. In cases of
targeted killing he would be targeted on the basis of his task-defined role. If he ceased to
act in that role (e.g. he ceased his pattern of actions indicating that he was a combatant
and instead became an official in the Cuban government) then he would no longer be a
target for an act of targeted killing.
Knoepfler does not distinguish between these types of prominence when defining
assassination.(Knoepfler 2010) For him any type of prominence is enough to classify the
intentional killing of a prominent person to be an act of assassination. By his definition
the killing of John Lennon was as much an act of assassination as was the killing of John
Kennedy. However, when looking at acts of targeted killing, prominence is by and large
irrelevant. Any value ‘prominence’ seems to hold in cases of targeted killing comes more
from the task-defined role that the individual plays in a given situation rather than the
prominence of their position. This also holds if we accept the narrower definition of
assassination that requires that the killing be carried out for political purposes. It is the
task-defined role the individual plays in the political situation, rather than the effect their
death might have, that defines them as a target. Now it is certainly the case that many
times the role one plays is directly tied to the prominence of one’s position and/or the
political effect one’s death might have, but it does not follow that this is true in every
case. The prominence of the individual targeted has less to do with why they are targeted
than the role they play in the situation. The political leadership of Hamas in power in the
Gaza Strip is quite prominent in both the positional and task-based sense, and killing
them might well serve some political purpose,25 but unless they can be shown to have a
particular task-defined role in the ongoing armed conflict between Israel and Palestine it
is not clear that they would be legitimate targets for an act of targeted killing as I have
defined it.
Acts of targeted killing may be carried out against any specific individual who occupies a
task-defined role in a given armed conflict. The prominence of the individual targeted
does not define him or her as a target; rather they are targeted because of their actions
during the conflict. In cases of assassination however this is not the case. Both the narrow
definition presented by Kasher & Yadlin and the broad definition given by Knoepfler
agree that assassinations can only be carried out against prominent individuals in either
25
Forcing a change of leadership for example.
18
Final Draft
Adam C. Gastineau
07/2011
the positional or ‘otherwise’ sense of the term ‘prominent’. That is, only those in
prominent (generally hierarchical, but not always, as in the case of John Lennon)
positions within a particular organization or society are targeted in assassinations. This
seems to indicate that targeted attacks carried out against less-than-prominent individuals
do not qualify as assassination, but must be some other form of intentional killing, such
as targeted killing, that does not target individuals on the basis of the prominence of their
position alone. This is particularly true when dealing with non-state actors, such as
terrorist groups, insurgents or guerrilla forces where individuals may have very important
task-defined roles but may not necessarily be highly placed within the organization, or
where individuals who are highly placed in the political structure of the organization have
task-defined roles that are of little importance when it comes to an ongoing conflict. For
example, political figures in various non-state organizations currently or historically
involved in armed conflict. Such figures would have little if any task prominence with
regard to the conflict itself, presuming they were not actively involved in high-command
positions, but would nonetheless have a degree of positional prominence in terms of the
hierarchy of the organization, or as figureheads of ‘the movement’, that might make them
targets for assassination, but not necessarily targets for an act of targeted killing.
As I stated above, it seems that terrorists who are targeted for ‘assassination’ in the cases
presented by both Kasher & Yadlin and Knoepfler are not targeted because they are
prominent, but because of the role they play within a certain conflict. In fact, the
terrorist’s prominence may come from the fact that they are targeted, not the other way
around. While it is certainly the case that prominence and role are often closely related, it
does not necessarily follow that the more prominent the person, the greater the role they
have in a particular conflict, or that that role is the result of one’s prominence. If we base
the definition of targeted killing on the prominence of the individual targeted, we have
needlessly confused the issue. The prominence of the individual seems to matter only
insofar as their prominence allows them to cause a threat to the population being
defended. The task-defined role the individual has in constituting that threat is what
designates the individual as a target, not their prominence.
Knoepfler’s definition of assassination as “the targeted killing of a prominent person”
places assassination as a sub-set of intentional killing based on the prominence of the
target and nothing else. When it comes to defining what Knoepfler believes morally
justifiable assassination might look like the justifications for the act hinge on the target’s
role as a threat rather than their prominence and the existence of a state of armed conflict.
To use an example stated above, the political leadership of Hamas in power in the Gaza
Strip, Hezbollah in Lebanon, or Sinn Fein in Northern Ireland are quite prominent in
terms of their positions in the organization, as well as within the political context of the
conflict, but unless they can be shown to have a particular combat role in some form of
armed conflict attacking such individuals would not be an act of targeted killing. Any
attempt to kill them might be an instance of assassination under the definition offered by
Knoepfler based on their prominence, but the act would not be an instance of targeted
killing, as their task-defined role would not necessarily designate them as a combatant in
an armed conflict.
19
Final Draft
Adam C. Gastineau
07/2011
Political Purpose
Kasher & Yadlin posit that assassination must be carried out for a political purpose to
qualify as an act of assassination. By their argument, if that political purpose was
justifiable, then the act of assassination itself could be morally justified, provided it was
the only possible way to achieve the desired result.26 This position reflects a common
view that one of the distinguishing factors of assassination is that the person killed is
killed for political purposes. However, if we are seeking a normatively neutral definition
for the term then we must be very careful about what we mean by a ‘political purpose’.
Killing another individual because of the pro tanto political end it may bring about does
not seem to be justifiable in any circumstance. It seems strange to say that we have the
necessary epistemological certainty of a particular political outcome necessary to justify
taking the life of a human being. Kasher & Yadlin understood this and, in order to
formulate a definition of assassination that was not morally wrong by definition, put
forward the idea that the term ‘political purpose’ must be re-defined as well, or at least
take on a different meaning in this context than the one ascribed to it when it is generally
used.
Kasher & Yadlin take a much broader view of what that ‘political purpose’ might be.
They argue that as the state has a duty to protect its citizens from harm, killing an
individual, or group of individuals, who pose a threat to that state’s citizens is killing for
a ‘political’ purpose. This is doubtless a strange concept of ‘political’ but it does offer an
advantage when trying to present a definition of assassination that is not wrongful by
definition. By defining ‘political purposes’ in this way, (i.e. by including the state’s duty
to protect its citizens as a political purpose) Kasher & Yadlin are able to morally justify
some acts of assassination on the basis of national self-defense and avoid the problem of
assassination being morally wrong by definition. As examples of how such a definition of
assassination might work, they hold up the hypothetical case of Adolph Hitler and the
killing of certain Hamas operatives by Israel.
In the hypothetical case, the act of killing Adolph Hitler falls quite neatly into the
definition I have proposed for targeted killing; Hitler was an active high-level combatant
engaged in an armed conflict, in so far that he was issuing orders and actively planning
both military actions against the Allies and harmful acts against his own people in the
form of using executions, the Gestapo, and the camps to rid himself of both political
opponents and certain groups he deemed to be threats in Germany, such as the Jewish
26
There is something of an epistemological problem here. Even in cases where there is a legitimate
political end, we cannot know that assassination will necessarily achieve this end and could wind up
bringing about more harm than good. However, as this problem exists for most intentional goal-based
actions it does not necessarily debunk the argument. One in favor of using political purposes, in Kasher &
Yadlin’s sense, to form a definition of assassination that would be morally justifiable might simply respond
that the act would only be justifiable in so far as it achieved a justifiable political purpose. If it was found
after the fact to have caused more harm than good, one might say that it was excusable to the degree that it
was reasonable for the authorizing agent to believe that the act would bring about the justifiable political
purpose
20
Final Draft
Adam C. Gastineau
07/2011
population. He would be targeted on the basis of his role in that conflict.27 However, the
act only fits Kasher & Yadlin’s definition of assassination if we accept the very broad
definition they propose of ‘political purposes’. Likewise, the killing of the members of
Hamas qualifies as assassination only insofar as we accept that the state fulfilling its duty
to defend its citizenry is a political purpose. This broad definition seems to place more
importance on the role the target has as a threat than the ‘political’ outcome in any
meaningful sense of the word. If the state must defend its citizens from such an
individual, it would be due to the threat posed by that individual fulfilling their role (in
terms of their task-defined role within the armed conflict).
Nefarious or Treacherous Means
Another feature that some thinkers use to distinguish assassination from other types of
intentional killing is the means by which the intention is carried out. Augustine, Grotius,
David (David 2003), and others make reference to the ‘nefarious or treacherous’ ways
that are used when some person or group seeks to assassinate, regardless of whom they
are seeking to kill or why. In fact, this feature of assassination has been one of the key
reasons why assassination might be considered morally wrong. The idea of using
premeditated trickery or deceit to kill another seems morally objectionable even in cases
where the individual may have acted in a way that would otherwise justify their death.28
For example, it may have been morally acceptable to kill Adolph Hitler in an attempt to
overthrow the Nazi government and end the war. However, von Stauffenburg may have
been committing a wrongful act in so far as a German officer who had sworn loyalty to
the German Reich should not attempt to kill the leader of that Reich. It could also be
argued that the regular use of nefarious and treacherous means could bring about harmful
results in the long run, even if the political purpose of the act was justified. Deception and
treachery can be seen as corruptive insofar as their use becomes tacitly or explicitly
permissible.
Regardless of if we agree with the moral argument that assassination is, by definition,
morally wrong on the basis of the means used, it does seem that nefarious or treacherous
means is a distinctive if not defining characteristic of assassination when comparing it to
other types of intentional killing such as targeted killing and extra-judicial execution.
Assassins are generally portrayed as working in secret, concealing themselves up to the
point they carry out the act. Often assassins are individuals who have some duty of
loyalty to the individual or group they are assassinating, and it is precisely this duty of
loyalty that allows them to get close enough to strike at their target. This might take a
strong form where loyalty is actually sworn, as in the case of von Stauffenburg and the
other conspirators who attempted to kill Hitler in 1944, or a weaker form where a citizen
27
In this case the term role could be defined as either the task-defined role or the positional role. Which of
these roles is most important in judging the moral value of an act of targeted killing will be discussed later
in the paper
28
S.R. David stresses the ‘deceptive’ aspect of assassination as one of the key criteria in distinguishing
acts of assassination from acts of targeted killing. He points out “The United States had little trouble
defending its efforts to kill Muammar al-Qaddafi in 1986 or Osama bin Laden in 1998 using bombs and
cruise missiles. Precisely because they were military operations and not carried out under false pretenses,
the ban against assassination did not apply.” David, S. R. (2003). "Israel's Policy of Targeted Killing."
Ethics & International Affairs 17(1): 111-126.
21
Final Draft
Adam C. Gastineau
07/2011
of a particular country attempts to kill a representative of that country. Citizens are
generally considered to have some level of duty to abide by the rules of the state and not
attempt to get rid of the leadership of that state, for whatever reason, by killing them.29
Even in cases where the assassin is actually an individual with no duty of loyalty to the
target of their attempt, they often pose as one at least until they are in a position to kill
their target(s).
Targeted killing is not, by definition, nefarious or treacherous. Individuals who are
targeted for attack are targeted on the basis of their role in a given armed conflict. This
role constitutes a threat to the opposing side of the conflict, and they are targeted on the
basis of that threat. This does not seem to be particularly nefarious or treacherous, as an
individual involved in an armed conflict is generally a) aware that they are involved in an
armed conflict and so b) actively threatens harm to the side they are opposing.30 It does
not seem that they have any legitimate expectation not to be targeted for some kind of
attack if they are behaving as a combatant within the conflict.
It might be objected here that plenty of acts of assassination take place during armed
conflict, and that in these cases the task-defined role, if not the positional role, of the
individual to be assassinated eliminates any expectation of immunity they might have,
just as in cases of targeted killing. To use the example given above, Hitler constituted a
threat to many people in his role, both directly to the Allies, and indirectly to Germans by
his continuation of an un-winnable war and his methods of keeping order in Germany. If
this is true than it could be argued that von Stauffenburg and his co-collaborators were
being no more ‘treacherous’ than a drone operator striking a target from several miles
away when they tried to kill him by planting a bomb during a strategy meeting. After all,
most people find little objectionable about commando raids, sniper attacks, surgical
strikes, and other highly targeted ‘hidden’ tactics when they are used against combatants
or others in combatant-like task-defined roles during armed conflict. If we allow that
Hitler was a legitimate target for attack on the basis of his task-defined role in
perpetuating an un-winnable war that was harming both those he was opposing and his
own citizens, why would von Stauffenburg and company be using treacherous and
nefarious means? Would the act, and other such cases, be better classified as an act of
targeted killing?
We are prevented from calling the assassination attempt against Hitler in 1944 an act of
targeted killing by the fact that von Stauffenburg et al., while engaged in armed conflict,
were not engaged in armed conflict against the Third Reich at the time they carried out
their attack. As members of the German army, and as German citizens, they were not
actually engaged in armed conflict against the Third Reich but were still operating as a
part of it. Even though they objected to the government and its policies, they still
operated as though they were a part of that government until they struck. If they had
29
There may be something wrongful about a citizen who has at least tacitly given their loyalty to their state
and its leadership betraying that loyalty by plotting to kill that leadership in the guise of a loyal citizen.
Again, the moral value of the act is not my main concern at this point, I merely mention it to link the idea
of treacherous or nefarious means with the definition of assassination.
30
Assuming they don’t give up, which would end their role in the conflict.
22
Final Draft
Adam C. Gastineau
07/2011
broken off from the German government in some way, by declaring themselves to be
partisans for example, or by staging an open protest, or otherwise breaking ties with the
German government, then an attack against Hitler by one of their members might be
considered an act of targeted killing provided it was authorized by the group and targeted
Hitler or others in the Reich on the basis of their role within the conflict between their
group and the Reich. Such an attack would not be nefarious or treacherous, as the splinter
group would no longer be attacking members of their own group. However, because they
did not break off prior to the attack, by either arguing as a group against the government,
or by defecting, the attack that they carried out involved an element of treachery and so
should be classified as an assassination attempt rather than a failed act of targeted
killing.31
This also raises an interesting question when looking at the enlistment or hiring of
others to carry out acts of targeted killing. If, for example, the Allies had enlisted or hired
von Stauffenburg to carry out the attack on their behalf, would this be an act of
assassination or an act of targeted killing? It could be argued that the Allies would be
authorizing, or at least facilitating, an act of assassination rather than an act of targeted
killing, as the means of using an active German officer to strike at German command
might be considered to be treacherous or nefarious. In such cases it would be helpful to
determine if in accepting an offer from the British von Stauffenburg had sufficiently
declared himself to be in conflict with the German Reich. If such was the case than an
attack carried out by him under Allied command might meet the definition of an act of
targeted killing as, unbeknownst to those in German High Command, he would be
engaged in armed conflict with them.32
Authorization
I have specified above that acts of targeted killing require some form of authorization
prior to the act. Assassinations, on the other hand, do not. Lee Harvey Oswald was not
authorized by any institution or group to carry out the killing of the President of the
United States.33 The proverbial ‘lone gunman’ has carried out other acts of assassination,
some with clear political purpose such as the shooting of Martin Luther King Jr., and
some that seemed focused only on the prominence of the target, such as the killing of
John Lennon. It does not seem necessary for an assassin to be authorized to carry out the
act by anyone but him or her self. Targeted killing on the other hand is not an act of
targeted killing unless it is authorized in some way. This is fairly straightforward when
31
In cases such as this, the high cost of defection to those defecting should be considered when weighing
the moral value of such an act. Such an act of assassination could be excusable, or even justifiable, if the
cost of openly opposing the government or defecting is high enough. Again, I am not interested in
addressing such issues in this section and will address the moral value of such acts later in the paper. I bring
up this example here only to illustrate how assassination may be differentiated from targeted killing on the
basis of treacherous or nefarious means.
32
Again, the knowledge of the target that they are in armed conflict with a particular agent does not seem
to be relevant here, as various ‘hidden’ tactics which depend on the target not knowing they are being
targeted (such as snipers and ambushes) are not seen as nefarious or treacherous in armed conflict when
directed against the opposing side in the conflict.
33
Unless, of course, you choose to believe the conspiracy theories to that effect. However, in light of
available evidence, such theories seem to be implausible to me and so I use this example.
23
Final Draft
Adam C. Gastineau
07/2011
we are talking about states engaging in acts of targeted killing but is less clear when
looking at the conditions under which a non-state actor undertakes an act of intentional
killing. How do we determine what is assassination and what is targeted killing in these
cases?
I have said that both assassination and targeted killing are acts of intentional killing. That
is, the action itself is the result of an Intentional component (I will kill x) and object of
intent (x’s death by means of fulfilling the Intentional component) wherein the
Intentional component has been determined prior to the act itself. This prior intent
differentiates acts of intentional killing from instances that may occur in conflict such as
firefights where the Intentional component is almost simultaneous with the act to realize
the object of the intent (intent-in-action), and instances where either the intentional
component or the object of intent is non-specific as in the case where a soldier goes out
on patrol and ‘ends up’ killing Bin Laden or Searle’s murderer who plans to kill their
uncle, but instead ‘accidentally’ runs them down with the car on the way to carry out the
act. For an act of assassination this specific prior intent is sufficient. However, for an act
of targeted killing there must be some level of legitimate authorization of the specific
object of intent prior to the act being carried out if it is to be considered to be an act of
targeted killing.
This doesn’t mean that assassinations are never ‘authorized’ in the sense that the assassin
has received orders from some sort of institution to carry out the act. An assassin may
have been told or commissioned to carry out the act, but this would not necessarily
authorize them to do so. A mob ‘button man’ may be told to make Paulie disappear by
the boss of his organization, but it doesn’t seem that this type of authorization is
sufficient to be considered an act of targeted killing. It is certainly an act of intentional
killing and may even be considered an act of assassination if Paulie is prominent enough
(for say arranging the hit on Don Corleone) or if it serves some ‘political’ purpose as
defined by Kasher & Yadlin (scares the other families into backing off and thus making
the ‘citizens’ of the Corleone family a bit safer), but it lacks the legitimate authority
necessary for it to be considered an act of targeted killing.
In the example above the organization that issued the authorization of the object of intent
is not a legitimate authority, legally or morally, as it is an ongoing criminal enterprise.
The legitimacy of the authority authorizing the act matters a great deal in cases of
targeted killing, as otherwise the act itself is not ‘authorized’, and so cannot be an act of
targeted killing. How this legitimacy might be determined will be discussed in the next
section, but for now I simply wish to make plain that while some acts of assassination
may be authorized (legitimately or otherwise) by institutions or groups, it is not necessary
that they be so to be considered assassinations. This authorization requirement, or rather
the lack thereof, also differentiates assassination from another type of intentional killing
that is often confused or conflated with targeted killing; extra-judicial execution.
24
Final Draft
Adam C. Gastineau
07/2011
Extra-Judicial Execution vs. Targeted Killing
Amnesty International has defined extra-judicial execution as:
“…an unlawful and deliberate killing carried out by order of a government
or with its acquiescence. Extrajudicial killings are killings which can
reasonably be assumed to be the result of a policy at any level of
government to eliminate specific individuals as an alternative to arresting
them and bringing them to justice. These killings take place outside any
judicial framework.”(International 2003)
By this definition acts of extra-judicial execution are acts of intentional killing, which are
authorized by a government and are carried out in retribution or as punishment for some
wrong or crime committed by the target rather than “arresting them and bringing them to
justice”. These two characteristics differentiate such acts from both acts of assassination,
which require no government authorization, and acts of targeted killing, which require
authorization, though not necessarily government authorization, but are not carried out as
a form of punishment or retribution against an individual or group for any particular act
or crime. Rather they are simply tactical or strategic strikes against combatants as a result
of the threat constituted by their task-defined roles qua combatants.
Government Authorization vs. Authorization
I have already indicated that authorization, of any type, is not a necessary characteristic
of assassination. On the other hand, both extra-judicial execution and targeted killing
require some form of authorization. Extra-judicial execution, as defined above, specifies
government authorization. Either explicit approval through “policy at any level” or tacit
approval by means of that government’s “acquiescence” is necessary for the act to
qualify as an act of extra-judicial execution. By this definition only state actors are able to
carry out acts of extra-judicial execution. This seems quite reasonable as the act is to be
carried out in lieu of judicial proceedings to punish or retaliate for a criminal offense that
would under other circumstances require some form of judicial proceeding. However, it
is unclear if this same requirement of government authorization is necessary in acts of
targeted killing. If so, then it would seem that non-state actors would be unable to
carryout acts of targeted killing as they would lack the necessary institutional structure
needed to authorize the act.
S.R. David seems to be of opinion that it is, as his definition of targeted killing specifies
that acts of targeted killing must be “undertaken with explicit government approval”.
(David 2003) However, it seems to me that when we are speaking about authority to
undertake an act of targeted killing the legitimacy of the authorization has less to do with
the institutional procedure or structures that are used to authorize the act then the
definition of extra-judicial execution might imply. It does not seem that we are talking
about requiring an institution to be ‘recognized’ by any particular body such as the UN or
the international community at large in order to be considered ‘legitimate’. We also are
not talking about a particular institution or group’s authority to impose punishment or
otherwise judge the targeted individual(s) as we are in cases of extra-judicial execution.
Rather it seems that in these cases what we are really concerned about is that the
25
Final Draft
Adam C. Gastineau
07/2011
authorizing body has some sort of moral legitimacy, similar to that held by state actors.
This moral legitimacy might be formulated along a contractarian model where the
institution represents the political will of some non-state group on whose behalf it acts, or
it could be formulated along more teleological lines wherein the legitimacy of the
institution is dependant on the normative legitimacy of the group’s ends. In such cases
institutions such as the mafia or other organized crime groups that might order acts of
intentional ‘targeted’ killing, even during ongoing armed conflict, would not be
considered to have ‘authorized’ such acts, as the groups lack the necessary institutional
legitimacy to do so.34
Given that acts of targeted killing occur only during periods of armed conflict, and that
the ‘targets’ of these acts are targeted on the basis of their roles within this conflict, it
would seem that justice of such a group’s cause would also bear some weight in the
legitimate authorization of the action. This is not to say that the justice of the cause is
sufficient to legitimately authorize an act of targeted killing by a non-state actor, but the
justice of the cause may affect the group’s ends and thus aid in granting the authorizing
institution legitimacy, especially under the teleological formulation. Authorization does
imply some form of collective action where a group decides that a particular course of
action is allowable or required according to certain rules or institutional procedures
accepted as binding by the members of that group.35 However, these rules and procedures
not meeting certain legal standards set by the international community, or failing to
comply with Robert’s Rules of Order, would not be sufficient to rule a particular nonstate actor to be ‘illegitimate’ with regard to authorizing particular acts, including acts of
targeted killing, from a normative standpoint.36
In cases where a non-state actor can be said to have just cause to use violence against
another state or non-state actor and the institution authorizing the act could be termed
morally legitimate, by either the contractarian and/or teleological accounts, it would seem
that the authority granted by that institution to its agents could be termed legitimate
without that authorization coming from a government institution. This being the case it
does seem that acts of intentional killing by non-state actors could be acts of targeted
killing when properly authorized by a morally legitimate institution. Such an institution
would derive at least a portion of this legitimacy from its moral legitimacy under either a
contractarian or teleological model (or perhaps both), which could include, in cases of
armed conflict, the justice of its cause.
34
It is not my intention to argue that one formulation is better than the other, simply to illustrate that either
could be used to determine the legitimacy of a particular group or institution.
35
The ‘lone-gunman’ is still an assassin, as an individual acting alone cannot meaningfully authorize his or
her own behavior to the world at large. Even in a hypothetical case where the agent is authorized by a
necessarily morally correct ‘God’ to carry out an act there are two parties involved: God (the authorizer)
and the agent (the authorizee), that work collectively according to mutually accepted norms
36
It could certainly be the case that this objection might be stronger when discussing legitimate legal
authority, but legal authority is not the topic of discussion here.
26
Final Draft
Adam C. Gastineau
07/2011
A Further Distinction: Retribution vs. Punishment
Before moving on to discuss the issue of retribution and punishment and how they relate
to instances of targeted killing and extra-judicial execution, I wish to highlight how one
can distinguish between the two.37 McMahan defines retribution as “the infliction on
wrongdoers of whatever penalties they deserve” (McMahan 2008) and points out that this
is only one of the justifications for punishment.38 Punishment may also be meted out for
other reasons including to protect innocent individuals from further harms committed by
the criminal, to dissuade or deter others from committing similar acts, and to express
society’s disapproval of such acts.39
Punishment then could be carried out for a variety of reasons, where as acts of retribution
are carried out only on the basis of desert. An elderly professor who throws water
balloons into his lecture hall full of students every time one of them falls asleep as
punishment for their inattention cannot be said to be acting in retribution unless he only
hits, or attempts to only hit, the sleeping students. Students actually hit by the balloons
may or may not deserve the punishment of being soaked to the skin. One could argue that
such acts are punishment by saying that by flinging the balloons haphazardly into the
lecture hall the professor both discourages the behavior of sleeping in class, and
expresses his, and perhaps the university’s, disapproval the behavior, but one could not
argue that the act of throwing the balloons and soaking the students was retributive.
Retribution or Punishment
The definition above specifies that an act of extra-judicial execution be carried out “as an
alternative to arresting them and bringing them to justice”. This seems to indicate that
acts of extra-judicial killing necessarily have an aspect of punishment involved in the
action. However, it is less clear about whether the definition is calling for punishment as
retribution, or punishment based on some other justification. When attempting to
determine the justifiability of such acts this piece of information would be quite
important. If the definition were referring to retribution, then one would have to show
that the individual deserved to be punished for the wrong that they had committed and
that the wrong was severe enough that they could be killed for it. If, on the other hand,
the definition is simply referring to punishment one would need to ask a much wider
range of questions pertaining to the probable effects of the action on the individual(s) and
37
The distinction between the two is drawn from Jeff McMahan’s arguments in the following: McMahan,
J. (2008). Aggression and Punishment. War Essays in Political Philosophy. L. May. New York, Cambridge
University Press: 67-84.
38
McMahan is unclear about whether or not those who are targets of retribution in fact deserve retribution
or are simply believed to deserve retribution and it does not seem to matter insofar as he is only seeking to
draw a distinction. This issue does become quite important when it comes to evaluating the moral value of
the act. It would seem, given the severity of the punishment in cases of extra-judicial execution, that one
would want to say the target actually deserved the punishment they received or that punishment would be
unjust.
39
McMahan also includes “reforming the criminal’s morals” in this list (pg 80). While I agree that
punishment may highlight areas in need of moral improvement, I do not see that anything except a
conscious effort by the criminal can reform their moral faults. I am probably reading McMahan too literally
here, but regardless of whom is in error I do not include this statement as I view it as a common but still
erroneous justification for punishment in my view
27
Final Draft
Adam C. Gastineau
07/2011
society being punished in addition to whether the punishment fit the crime, so to speak.
However, even in this case it might be claimed that retribution had some role to play as it
could be plausibly said that retribution is a necessary condition for punishment even if
other reasons, such as deterrence, exist. In both cases there would be questions about the
justifiability of punishing or retaliating against the individual without, in the case of
retribution, proving the individual(s) targeted actually deserved to be killed, or in the case
punishment, proving that the individual(s) targeted deserved to be punished and that the
punishment would have the desired effect.
One objection that is commonly raised by many groups, including Amnesty International,
is that such actions are wrongful by definition because by killing the target prior to any
judicial proceeding, these conditions for justification cannot be met.40 Instead they are
killed for a wrong that may or may not have been committed and/or may or may not have
been severe enough to deserve death. If one attempted to circumvent this objection by
pointing out that the act was punitive not retributive, and so the guilt of the individual
punished was not necessarily relevant, one could respond that given the severity of the
punishment in acts of extra-judicial killing, moral questions about the justifiability of
killing someone in order to express society’s displeasure or to dissuade others from
taking similar action would quickly be raised. Both of these objections seem to hinge on
some idea of desert: either the target being guilty and so deserving to be killed for a
particular act, or the target deserving such a severe penalty as punishment when it is
almost always unclear that such a punishment would produce the desired effect. If it
could be shown that acts of targeted killing are carried out in retribution or as punishment
for past wrongs or crimes it would seem that targeted killing would be susceptible to
similar criticisms, which would be similarly problematic when it came to justifying such
acts.
It seems that targeted killing can avoid these issues however. Cases of targeted killing are
carried out during an ongoing armed conflict, and those targeted are targeted on the basis
of their (task-defined) role in that conflict, it does not seem that punishment or retribution
enter into the equation as they necessarily do in cases of extra-judicial execution. Even
when they do, such motivations do not decide the target of the action as they do in cases
of extra-judicial execution. This is not to say that the intentional object of an act of
targeted killing must be entirely devoid of such ideas. In the emotionally heightened
conditions that exist during armed conflict it is hardly unlikely that the individual tasked
to carry out the attack, or even those authorizing the attack, may feel that one of the
reasons the attack is allowable is to ‘get that SOB’ in retribution for some act the target is
blamed for carrying out. While ‘crime’ is a slippery subject during such conflict, strikes
in retaliation for ‘atrocities’ carried out by the opposing side are not uncommon. Even in
40
Implicit in this objection seems to be the idea that without proper procedural justice it is difficult, if not
impossible, to determine what is substantively just as only one side has the opportunity to present their
‘case’. Someone sympathetic to this view may argue that even in cases where the death of the target could
be reasonably said to be substantively just (e.g. bin Laden’s death) we cannot know if it was or not until
both sides of the story are heard and judged by an objective (or reasonably objective) third party as in a
judicial proceeding. Because we cannot know, and because the harm is so severe, they might claim that we
are morally obligated to ensure that an individual accused of such wrongs receive procedural justice, as this
is the most reliable method to bring about substantive justice.
28
Final Draft
Adam C. Gastineau
07/2011
less emotionally charged cases killing those thought to be responsible for a state of
affairs, either on the basis of desert, or on the basis of punishing the individual or group
to achieve some other end, may seem to be a better option than capturing them for a
variety of reasons.41 However such reasons are not the basis for authorizing acts of
targeted killing by the definition I propose. Rather individuals targeted in acts of targeted
killing are targeted on the basis of their (task-defined) role in the conflict, and the
permissibility of attacking them depends on that (task-defined) role making them liable to
be killed (ex ante) rather than their deserving to be killed in retribution for past actions.
Retribution may enter into the calculus for the person carrying out the act, or even for the
institution authorizing the attack, but in the absence of an ongoing armed conflict and the
(task-defined) role that the target(s) have in the context of that conflict, such attacks
would not qualify as acts of targeted killing by definition. They would simply be acts of
extra-judicial execution and have to be examined by the criteria applying to extra-judicial
execution, namely desert, to determine their moral value.
Standards of Justifiability
So far, I have defined targeted killing as the authorized, intentional killing of a specific
individual or group of individuals who are targeted on the basis of their role in a given
armed conflict and distinguished this definition from assassination and extra-judicial
execution. All three acts are acts of intentional killing in that they are carried out with an
intention held prior to the act wherein the act is performed by means of bringing about
the prior intention. Furthermore this prior intention is directed at a specific intentional
object, such that if the object is realized in a way other than that included in the intent, the
act either fails (I have the experience of raising my arm, but my arm does not go up
because it is held) or the act is not caused by the prior intention in the right way, and so is
unintentional (as when I run over my uncle on the way to kill him without first realizing
that the man in the crosswalk is my uncle).
However, these three acts are distinct from one another in several ways. Assassination is
distinguished from extra-judicial execution and targeted killing in that it necessarily
targets those in prominent positions, regardless of task-defined role, and may (or must if
one agrees with Kasher & Yadlin’s definition) be carried out for political purposes,
widely or narrowly defined. Furthermore, the means used to affect the killing are
generally regarded to be nefarious and/or treacherous in cases of assassination. Extrajudicial execution is a government authorized act carried out in order to punish the
individual or group targeted on the basis of having committed some past wrong rather
than arresting and trying the target. This punishment is based on desert but is carried out
without giving the target a chance to defend him or herself as they would if arrested and
tried. Finally acts of targeted killing are authorized acts carried out within the context of
armed conflict against those whose task-defined role makes them a combatant in that
conflict, and thus liable to attack. The concepts of both armed conflict and combatant are
41
Including discouraging the behavior, or to ‘send a message’ that such behavior will not be tolerated by a
particular state or group. Pragmatic arguments about the cost (both in economic terms and in the number of
people who might be harmed) and likelihood of capturing and trying the target in a court of law might also
be raised.
29
Final Draft
Adam C. Gastineau
07/2011
broader than traditionally defined but the necessity of these two criteria serve to
distinguish such acts from the other two acts of intentional killing discussed. Combatants
are not defined by being part of a hierarchical group, wearing a uniform, or bearing arms
openly; rather one’s status is determined by looking at the overall task-defined role that
one has in the given conflict. If this task-defined role has to do with the active
continuance of the armed conflict by means of engaging in or facilitating combat, then
one is effectively a combatant and thus liable to attack. This allows for the inclusion of
the wide range of what have been termed as “civilian combatants” or “quasi-combatants”
(Gross 2010) that have become increasingly common in asymmetric conflict. The term
‘armed conflict’ has also been broadened in scope to include types of conflict that may
occur outside of the context of conventional warfare. Armed conflict occurs when one
party attempts to enforce their will on another party by force of arms and the other party
has some, but not necessarily equal, means to resist such force. Such conflict continues
until one side capitulates or otherwise withdraws from the conflict or no longer has the
means to offer any resistance. This model of conflict similarly allows for the inclusion of
various asymmetric conflicts both military and non-military such as conflicts that may
arise in policing, terror/counter-terror conflict, and other armed conflict between state and
non-state or non-state and non-state actors.
There are two benefits to drawing such distinctions. First of all, by illustrating the
differences between these three acts of intentional killing we can get a clearer view of
what criteria one should use to determine the moral value of a particular act. The
justification for an act of assassination depends on what we determine to be the political
purpose of the act, and, to a lesser extent, the nefarious or treacherous means used to
carry it out. The justification of extra-judicial execution hinges on both the legitimacy of
the authority authorizing the act and, most importantly, whether or not the punishment
inflicted can be justified on the basis of desert. Acts of targeted killing similarly depend
upon the legitimacy of the authority authorizing the act, but also depend on whether or
not the task-defined role of the target(s) make the person targeted liable to attack as part
of an ongoing armed conflict. Secondly, we can also demonstrate how one or more of
these acts can avoid problematic features of the others and so be able to set these features
aside as irrelevant in our consideration. This allows us to see the matter more clearly and
without the waters being muddied by any negative baggage the general concept may bear.
To clarify this further, let us consider the following case.
Imagine a government authorizes a military team to strike an individual
with whom they have been involved in an armed conflict.42 The individual
in question (let’s call him Old Boy Leroy or OBL for short) is the leader
of the non-state actor with whom the state actor is in armed conflict. OBL
has declared ‘war’ on the state multiple times in various public
announcements and has taken credit for authoring several attacks on the
state actor and others while also endorsing attacks carried out by other
individuals associated with the same group. Aside from authorization,
there is also evidence that he participates in the planning of some of these
42
Given the recent events that took place around May 2nd 2011, this shouldn’t be too difficult.
30
Final Draft
Adam C. Gastineau
07/2011
attacks, as well as the overall strategy used by the non-state actor. The
conflict in which the actors are involved is an asymmetric conflict but
attacks by both sides have been continuing for some time. Some of the
attacks by the non-state actor have deliberately targeted civilians and
civilian targets of no real military value. When the state actor strikes,
using military personnel and equipment, the leader of the non-state actor is
taken by surprise and is killed. In announcing the death of the leader, the
president of the state actor declares that he gave the order to ‘get’ the
leader and “bring him to justice.”(Obama 2011)43 Given this situation, is
this a case of assassination, extra-judicial execution, or targeted killing?
Assassination or Targeted Killing?
One could argue that such a case would be a case of assassination. Knoepfler would point
out that OBL could be said to have been targeted on the basis of his prominence in terms
of both the task and position that he had within the organization. One could also argue
that killing the individual served a political purpose in the narrow sense of decapitating
the organization, or in the broader sense of defending the population of the state actor
from a known threat by eliminating that threat. Are we then best served by defining the
intentional killing of Old Boy Leroy as an act of assassination to determine the
justifiability of the act?
I would argue no for the following reasons. First the act itself was not particularly
nefarious or treacherous, which while not a necessary condition of assassination, is part
of the general concept of assassination and so, if present, should push us in the direction
of calling such an act assassination rather than something else, all other things being
equal. Secondly, because we must judge the justifiability of assassination by the act’s
political purpose the only cases of assassination that might allow us to start from neutral
ground look remarkably similar to cases of targeted killing, which brings me to one last
reason. Because of this close similarity between a morally justifiable (but not necessarily
morally justified) case of assassination and a case of targeted killing, we may want to
avoid defining such an act as an act of assassination for clarity’s sake; that is defining the
act as one of assassination can negatively skew our thinking on the matter from the
beginning just as referring to a killing as ‘murder’ would. The term assassination does
carry negative moral connotations, and these connotations can affect our intuitions about
the case at hand. This would muddy the waters from the start.44 Given the fact that the act
occurred during armed conflict, and that the ‘political purpose’ of the act is only
justifiable if it occurs within such a context (for reasons I illustrate further below,
following Kasher & Yadlin’s arguments) it would seem preferable to define the act as an
43
I am assuming in this case that the team was in fact sent to kill the target, not merely capture him. It is
possible that the team be sent to capture the target but end up killing him in self-defense. Such a case would
lack the necessary intentionality needed for targeted killing or extra-judicial execution, as the intentional
object does not include killing the suspect. Such a case would ‘fail’ to capture the leader rather than be a
successful case of targeted killing or extra-judicial execution.
44
Both Knoepfler and Kasher & Yadlin have pointed out this problem and stated that it is why they
attempted to offer more neutral definitions. However, re-defining the term does not necessarily fix this
problem.
31
Final Draft
Adam C. Gastineau
07/2011
act of targeted killing in this case, and arguably in similar cases that occur during armed
conflict, in order to avoid the negative normative baggage the general concept of
assassination carries with it.45
In this case the act itself was not particularly nefarious or treacherous. The individual
targeted knew full well that he was engaged in an ongoing armed conflict with the state
actor, and so should have realized that he could be attacked as part of that conflict.
Granted this attack did take Old Boy Leroy by surprise, but surprise is an acceptable
tactic in modern conflict. Ambushes, sniper fire, and other similar types of surprise
attacks have become part of the standard set of tactics used to fight in armed conflicts.
This is especially true in asymmetric conflicts that take place between state and non-state
actors. While the targeted individual may not have expected an attack at the moment it
came it would be unreasonable for a combatant, even a high-level combatant, to believe
themselves to be immune from attack while the conflict in which they were participating
was still going on. It would seem then that one wouldn’t be able to use this characteristic
to determine the moral value of the act in this case.
Kasher & Yadlin point out that the prominence factor of the target is indeterminate when
considering the normative value of the act. So, if we are to make a statement about the
moral value of the act we are forced to address the ‘political purpose’ of the act. What
about the ‘political purpose’ of the act? Regarding this ‘political purpose’ Kasher &
Yadlin say:
“The difference between a political purpose in the broad sense of the term and a
political purpose in the narrow sense is crucial. Assassination for a purpose of the
former kind is sometimes justified, morally and ethically, whereas assassination
for a purpose of the latter type is never justified, from the same points of view.
Broadly, the purpose of eliminating a danger to the life of citizens, when
jeopardized by terrorists, is political and justified when governed by the
principles of the military ethics of fighting terror, as presented in the second
section of the paper. Under certain conditions, specified by those principles, an
assassination is morally and ethically justified when it is an act of killing a
prominent person, selectively and intentionally, for the purpose of self-defense.
However, within the framework of intra-national relationships or international
relationships during peace, there is no moral or ethical justification for changing
the system of powers by killing a prominent person. Within the intra-national
arena, the moral principles of democracy and their legal implementation
delineate pursuits of political change in a way that excludes assassination. Within
the international arena during peace, the moral principles that pertain to
45
In this case it seems that we have enough evidence to at least make the claim that the act could be one of
targeted killing as it was authorized and it was an intentional killing of a specific individual that could be
said to be acting in a task-defined role in an ongoing armed conflict that made them liable to attack. If the
act was unauthorized, or illegitimately authorized, we could not even consider the act to be one of targeted
killing, as it would not meet the authorization requirement of the definition. Likewise, if the armed conflict
had ended before the time of the attack we would be left to decide between assassination and extra-judicial
execution. If the act were non-specific or unintentional then we could not claim it to be an act of
assassination, targeted-killing, or extra-judicial execution.
32
Final Draft
Adam C. Gastineau
07/2011
international relations and international law preclude assassination.”(Kasher
2005)
So, by the standards set out here, the ‘political purpose’ that allows for the possibility of
an act of assassination being morally justifiable is that of self-defense.46 This seems
plausible enough. However, this standard only seems to apply when the parties involved
are in a state of armed conflict: either asymmetric conflict with non-state actors, or during
wartime when in conflict with other state actors. This formulation then does not differ
substantially from acts that might be better defined as acts of targeted killing. This does
not mean that all such cases would be necessarily justified, as the ‘political purpose’ of
self-defense may or may not hold depending on the situation. However, if the criteria we
are focusing on to determine the justifiability of the act is the extent to which killing a
particular individual in a time of conflict constitutes self-defense, it would seem that we
are looking more at the (task-defined) role that the targeted individual is playing in the
conflict, and any harmful consequences that might ensue from that role, than the
‘political purpose’ of the act when we are trying to determine the act’s moral value. This
then would be a point in favor of calling the act an act of targeted killing in this case.
Given the fact that the general concept of assassination carries with it an idea of the act
being wrongful in some way, and that in cases such as the death of Old Boy Leroy the
‘political purpose’ of self defense necessarily refers to the (task-defined) role of the
individual targeted to determine the moral value of the act, it would seem to be preferable
to refer to such acts as acts of targeted killing rather than acts of assassination in order to
avoid any confusion the negative normative baggage in the general concept of
assassination may cause. On first brush this may seem to be a pragmatic rather than a
normative claim. It would certainly seem to be more practical to use neutral terms to
avoid confusion when trying to determine public policy. However, it seems to me that
there is a normative aspect to this as well.
First of all, it would seem that one has a duty to be as clear as possible when arguing for
or against important policy matters to ensure that the best decision possible is made.
Secondly, if one of the moral duties of the government is to formulate policy in the best
interests of its citizens, which includes allowing citizen evaluation in the case of liberal
democracies, then the government would seem to have a certain duty to talk about such
policy using general concepts that most people would understand, both to allow any valid
objections to be made and to accurately inform the public of what policy is actually being
followed. If the general concept of ‘assassination’ has a negative connotation in the same
way that ‘murder’ does, then using a term like “justified assassination” or “defensive
assassination” could be as potentially misleading as “torture-lite” or “illegal combatant”.
One might offer a normative argument along consequentialist lines that insofar as such
ambiguous terms lead the populace or institutions of that populous to conclusions with
overall harmful results, using such terms is wrongful. Contractarian thinkers could also
argue such a point by pointing to the institution’s duty to inform its citizens or a citizens’
right to be aware and understand policies enacted on their behalf. Ambiguity about what
46
By self-defense I take Kasher & Yadlin to mean some form of ‘national’ or ‘state’ self-defense wherein
the state authorizes a killing in order to prevent some harm from coming to its citizens, and so to its ‘self’.
33
Final Draft
Adam C. Gastineau
07/2011
terms reflected what acts could be said to be wrongful because such ambiguity, if left
unaddressed, could constitute a failing in the former duty or a violation of the later right.
It would seem preferable that when evaluating the moral permissibility of acts of
intentional killing we abandon the term assassination in cases such as these where the act
occurs as part of an ongoing armed conflict. This is especially true here, as it seems that
the (task-defined) role OBL played in this particular conflict is key to determining if the
act is morally justifiable or not. In this case it would seem preferable to use the term
‘targeted killing’ as I have defined it. The act of killing OBL was authorized, intentional
and specific, occurred during an ongoing armed conflict, and could be said to have been
carried out against one whose (task-defined) role in that conflict could have justified the
attack.47 Furthermore, the general concept of targeted killing does not have the same
negative conceptual baggage of being nefarious or treacherous that burdens our intuitions
about assassination, and so would allow us to begin our normative evaluation from more
neutral ground.
Extra-Judicial Execution or Targeted Killing?
So, based on the distinctions drawn in the first part of the paper and the argument given
above we have narrowed the field by one. This case should not be evaluated as a case of
assassination. However, we are still left to choose between evaluating the case as if it
were a case of extra-judicial execution or a case of targeted killing. Given the statements
made by the president of the state actor about the ‘justice’ of the act this choice is
difficult. The act was legitimately authorized,48 took place during an ongoing armed
conflict, and the (task-defined) role OBL performed could justify treating him as a highlevel combatant in that conflict. This could lead us to call the act an act of targeted
killing. However, assuming that the president’s statements were not merely rhetorical, the
act seemed to include the intent to punish, if not necessarily to get retribution for, the acts
authorized and endorsed by Old Boy Leroy. Furthermore, the authorizing agent in this
case was a government, which ticks another box in favor of evaluating the event as an act
of extra-judicial execution based on the definition offered above. Drawing the distinction
between extra-judicial execution and targeted killing will determine whether the moral
justification of the act should be judged by some normative evaluation of desert or
punishment, or if it should be judged by the extent to which Old Boy Leroy was or was
not acting in the task-defined role of a combatant in this particular conflict.
The government of the state actor authorized the killing of OBL, explicitly in this case.
However, both governments and non-state institutions or groups may authorize acts of
targeted killing provided that those institutions or groups are ‘legitimate’ in a normative
sense.49 In some cases, such as cases in which the government or institution lacks moral
legitimacy, (e.g. dictatorships or other forms of totalitarian rule that fail to meet
47
This is not to say that it did, merely that it could. It does not seem implausible to me that Old Boy Leroy
would be qualified as a combatant in this situation given his actions described in the scenario.
48
In the sense that a morally legitimate institution (in the case of targeted killing) or a government
institution (in the case of extra-judicial execution) issued the authorization, not necessarily in the sense that
the authorization was morally justified
49
See the section on Government Authorization vs. Authorization above.
34
Final Draft
Adam C. Gastineau
07/2011
requirements of moral legitimacy) this aspect of both extra-judicial execution and
targeted killing may become an issue. However, this particular feature does not help
much to determine the moral value of the act in this case. So, as with the characteristic of
prominence in the discussion of assassination, I set it aside for now.
This leaves us with the retributive value of the act. Retribution offers the strongest
possible justification for punishment in this case. If we try to justify the act by referring
to other reasons offered for punishment such as dissuasion of a particular behavior, or
punishment as an expression of society’s disapproval of that type of behavior, the harm in
this case seems to eclipse the benefit.50 As mentioned previously it seems that there are a
couple of prominent objections related to desert that must be overcome here if the act is
to be justified along these lines. First, we must deal with the objection that the
individual’s guilt or culpability has not been proven, so it is unclear that they deserve any
retributive punishment at all. Secondly, even if we can construct an argument that shows
the individual to be in some way ‘guilty’ of committing a particular wrong, we must
show that the severity of the punishment for that wrong is justified. In short, we must
show that the individual, being guilty of a wrong, deserved to die for committing that
wrong.
So then, how does retribution factor into the case we are looking at? Assuming that the
president’s statement about justice isn’t just rhetoric, this statement seems to indicate that
at least part of the intentional object was to bring about justice by killing OBL: that he
deserved to die and so killing him was “bringing him to justice”. Is this idea of killing
Old Boy Leroy in retribution for past wrongs sufficient to determine this particular act of
intentional killing is an act of extra-judicial execution?
My answer in cases like the one above is no. The inclusion of the idea that the individual
targeted deserves to be killed and thus brings about some form of ‘justice’ does not
automatically determine that the act is one of extra-judicial execution. I stated above in
the section on retribution that retaliating for an event may be part of the intentional object
of targeted killing, but it is not what one should focus on when trying to determine the
moral worth of the act. Why shouldn’t we focus on it by declaring the act an act of extrajudicial execution in this particular case? There are, I think, and least three reasons. First,
50
One could fashion a utilitarian or consequentialist argument that might justify punishing an individual so
severely in the name of maximizing utility or some other conception of the common good. For example, if
the death of OBL would be sufficient to completely destroy his organization, and thus prevent the harms
that the group might do in the future we might think that punishment by death would be allowable. There
are some problems with this line of thought however. First of all, there is the common epistemological
problem of knowing that such an act would be sufficient to achieve the desired end rather than some other
end (such as an increase in violence to retaliate). Secondly, we are still left with the feeling that even if a
consequentialist justification based on achieving some good end (say the common good in this case, or
peace, or some other good) came to pass, the ‘goodness’ of the act would be weakened substantially if it
turned out that Old Boy Leroy did not in fact have anything to do with the wrongs attributed to him, or that
the wrongs actually committed were insufficient to justify killing him. In cases where we are talking about
the intentional killing of another human being demanding the most stringent justification possible is
desirable to prevent reckless misuse of the ‘justified’ act. People, after all, are not potatoes. The deliberate
killing of a human being should only be allowable in very limited circumstances, insofar as it is allowable
at all.
35
Final Draft
Adam C. Gastineau
07/2011
the fact that the act occurred during an ongoing armed conflict. Secondly, given the taskdefined role Old Boy Leroy was fulfilling, it could be argued that he constituted a threat
ex ante and not just ex post. This would make him liable to be killed to prevent him from
further harming others or other such ends, rather than on the basis of desert, which would
be an end in itself. Finally, because, if the first two statements are true, calling the act one
of extra-judicial execution brings in negative normative baggage that confuses the issue
unnecessarily, which could lead to harmful results. Even if it is the case here that one
could define the act as either extra-judicial execution or targeted killing it seems
preferable to avoid the problems caused by the negative normative baggage inherent in
extra-judicial execution by labeling the act an act of targeted killing. Of course, this
option is only possible if the act has met the requirement of occurring in an ongoing
armed conflict and the requirement that Old Boy Leroy was acting in a task-defined role
that made him liable to attack, which it seems to have done in this case.
In this case, one could say that this was an act of targeted killing. The non-state actor and
the state actor had been involved in armed conflict for some time. OBL has said on
multiple occasions that he is at war with the state actor that killed him. He and his
organization act as though they are at war. His organization has a hierarchical structure
and plan and carry out attacks in much the same way that a state might. Though they may
not be ‘at war’ in any conventional sense, it certainly seems that they are involved in an
ongoing armed conflict. Additionally, the fact that Old Boy Leroy was killed by a
military team in a military operation, rather than by a law enforcement organization
attempting an arrest, or a by traitor or secret agent disguised as one of his colleagues
(which, being deceitful, might push us toward calling the act an act of assassination)
supports the contention that this particular act was an act of targeted killing rather than an
act of extra-judicial execution. This is not to say that it is impossible for an act of extrajudicial execution to occur during an ongoing armed conflict. It is quite plausible that an
individual whose task-defined role changes such that they no longer meet the standard of
a combatant (a former bomb-maker or operational planner for Hamas being brought into
a political position in Gaza for example) to be targeted for extra-judicial execution as
retaliation or punishment for previous ‘crimes’ (moral or legal) they were involved with
while in their previous task-defined role.51 However, it does not appear that such is the
case here.
The example of our ‘retired’ bomb-maker points towards the second reason to call this
particular case an act of targeted killing rather than an act of extra-judicial execution. Old
Boy Leroy was not retired. He was still acting in a task-defined role that made him liable
to attack. He was engaged in authorizing current and future actions against the state actor,
and so his task-defined role gave the state actor a plausible claim that he constituted an ex
ante threat. Because he was still a threat, the state actor could attack him and justify the
act on the grounds of some formulation of defensive action. This argument could look
much like the duty-based self-defense argument presented by Kasher & Yadlin, or it
could be formulated along more consequentialist lines by arguing that not killing OBL
would produce a harmful result. If Old Boy Leroy were a high-level combatant in a state
51
I do not claim that such an act would be morally justifiable, but it would qualify as an act of extrajudicial execution that occurs during an ongoing armed conflict.
36
Final Draft
Adam C. Gastineau
07/2011
vs. state conflict it would seem that he could be justifiably attacked as a combatant in that
conflict, regardless of any wrongs he may or may not have committed. Combatants are
liable to attack, even surprise attack, because attacking them serves some further purpose
(defending citizens from harm or avoiding an otherwise harmful result) not because they
somehow ‘deserve’ to be.52 Even in conventional conflict a ‘war criminal’ is not treated
as a criminal until after the individual has surrendered, is captured, or the conflict
otherwise ends. As long as the conflict continues the individual in question is still a
combatant, constitutes a threat on the basis of that status, and so be justifiably attacked.53
So if we look at OBL’s task-defined role relative to the context (that is of a combatant,
albeit a high-level one) then it would seem that this act could be an act of targeted killing
as it could be said that his task-defined role made him liable to attack in order to achieve
some further end rather than the attack being an end in itself.
In the case at hand these two reasons seem to give strong support to the contention that
we should view this particular case as an act of targeted killing rather than an act of extrajudicial execution. However, there is one other factor that we should consider: that of the
negative normative baggage that comes with the general conception of extra-judicial
execution. Extra-judicial execution seems to be unjust because the individual targeted is
targeted because they somehow deserve to be killed, but the person targeted is never
given an opportunity to show that they do not. However, it seems that in this case we are
talking about killing an individual who is liable to be killed because of actions ex ante
regardless of whether or not they deserve to be killed for actions ex post.
The distinction between liability and desert is often very hard to make in these cases,
especially when the idea of retaliation for past offenses becomes incorporated into an
ongoing armed conflict. Indeed S.R. David confuses the issue in his essay on targeted
killing by saying that “Israel’s policy of targeted killing, stripped of its utilitarian
justifications, is retribution, plain and simple.” (David 2003) A point which is wholly
rejected by Stein (Stein 2003) on moral grounds as unjust given that those targeted are
targeted without trial, or to use another turn of phrase, targeted extra-judicially, and so
have not been proven to deserve such retribution. If Israel’s policy of targeted killing is in
fact a policy of extra-judicial execution in that it is done to exact retribution or revenge
for Palestinian suicide attacks or missile strikes rather than to eliminate those whose taskdefined roles make them liable to attack, I am inclined to agree with Stein that such
actions taken for retribution alone are generally unjust. Given the harm caused by the
punishment, it would seem to me that taking such action without giving the individual an
opportunity to defend him or herself is difficult to morally justify, especially given the
difficulty of bringing about substantive justice without hearing both sides of the story. In
52
For further discussion on the distinction between liability and desert as they relate to defense and
punishment see pg 81 McMahan, J. (2008). Aggression and Punishment. War Essays in Political
Philosophy. L. May. New York, Cambridge University Press: 67-84.
53
McMahan and others have set aside the idea of ‘moral equality of combatants’ in various ways.
McMahan argues that only those fighting for an unjust cause can be justifiably attacked. I do not wish to
pursue the argument here and so will stay with the Walzerian position that combatants are more or less
equally allowed to kill one another during armed conflict for simplicity’s sake. For some of McMahan’s
arguments to the contrary please see: McMahan, J. (2006). "On the Moral Equality of Combatants*."
Journal of Political Philosophy 14(4): 377-393.
37
Final Draft
Adam C. Gastineau
07/2011
an armed conflict there are many instances in which the intentional killing of an
individual may be justifiable, but when it comes to doing so on the basis of desert, we
must have very (very) good reason to believe that the person we are killing does in fact
deserve to be killed. This can be a very difficult standard to meet given they are not
posing an imminent threat to us or someone else. There may be some cases in which we
can say an act of extra-judicial execution is morally justified or at least morally excusable
(that is a wrongful act all things considered, but excusable to some degree under the
circumstances),54 but I do not know that such a standard should be used when forming
national security policy. As Stein points out:
“Many people may sometimes feel a need for revenge and retribution, but these
feelings cannot be used to justify policies. States, as individuals, are bound by
laws. When Israel emphasizes that these killings are preventive rather than
punitive measures, it does so for a reason: punishment without trial is illegal and
immoral.” (Stein 2003)
Does Stein’s criticism holding punishment, of any type, without trial to be immoral hold
in this case? I think that it does, at least if we accept the premise that judicial proceedings
are more effective in bringing about substantive justice than other methods of meting out
punishment and/or the idea that the rule of law constitutes a moral ‘good’.55 In this case
we may have a large amount of evidence indicating that Old Boy Leroy committed
certain acts that may justify his death. However, because it has not been proven that (a)
OBL did in fact commit the acts in question or is otherwise culpable for them, and (b)
that culpability for these acts justifies the death of OBL (as opposed to some other form
of punishment) it does not seem just to kill him out of hand as punishment for those acts.
Doing so constitutes a wrong in that it both harms OBL without proving cause and in that
it disregards the idea of rule of law in criminal matters, which, insofar as rule of law is a
moral ‘good’, also constitutes an unjustifiable act of corruption.(Stein 2003) If OBL was
selected as a target on the basis of some past wrongful act, it seems that killing him rather
than detaining him would be difficult to justify in normative terms, as we would have to
leap these hurdles along the way. However, if we look at this particular case it would
seem that we do not need to worry about justifying the act on grounds of desert. Old Boy
Leroy was still acting in a task-defined role that would seem to make him liable to attack,
and so it would seem that one could justify an act of targeted killing based on that taskdefined role.
The ease of confusing the acts of targeted killing and extra-judicial execution is apparent,
especially in cases such as the one above where some idea of ‘justice’ is included in the
intentional object of the act. Ultimately the question we must ask ourselves in such cases
is this: “Is the target acting in a task-defined role that makes them liable to attack ex ante
or are they targeted ex post on the basis of desert?” Cases that meet the former standard
54
Such a case might be one like the killing of Pablo Escobar. For a case study and analysis see pp 148-149
Miller, S. (2009). Terrorism & Counter-Terrorism: Ethics and Liberal Democracy. Malden MA, Blackwell.
55
I am not claiming here that judicial proceedings are always necessary to bring about substantive justice,
though I think Stein might. Rather I am saying that insofar as such proceedings are more likely to bring
about substantive justice than other methods, an individual who is to be punished, especially punished
severely, for an ex post act seems to have some moral claim to such proceedings.
38
Final Draft
Adam C. Gastineau
07/2011
are cases of targeted killing; cases that meet the later are extra-judicial executions. The
standard that must be met to justify an act of extra-judicial execution is quite difficult to
meet,56 which raises another issue. This difficulty in distinguishing between the two acts
in some cases coupled with the relative ease of justifying an act of targeted killing could
lead an authorizing government to undertake acts of extra-judicial execution and then
portray the target as liable to attack after the fact. One might even argue that point in this
case, given the statement made by the president of the state actor.
I can see three reasons to set aside this objection. The first is rather simple. The difficulty
of the choice, or the temptation to choose one option over the other because it is easier to
justify, is irrelevant if we are actually interested in making the correct moral choice. Even
in cases where the choice is difficult we are still under an obligation to make the correct
choice, and the choice that is easiest to justify is not always the ‘right’ choice. Hard cases
are hard, but this does not mean that we are allowed to take whichever option is easily
justifiable. Instead we are obligated to look at the facts of the matter and make our choice
to the best of our abilities. Secondly, though governments may be tempted to fudge on
these matters, such ‘fudging’ can, and should, be controlled by institutional measures
preventing such actions.57 Institutional measures based on the distinction between
liability and desert that I have outlined can give clear guidance in the formulation of laws,
rules, or other types of policy, which set standards that can prevent abuse. In instances
where abuse occurs, such standards give grounds for the punishment of such abuse and
can demand accountability. Finally, if we are able to distinguish targeted killing from
extra-judicial execution in a general sense by distinguishing between cases of liability
and desert, the international community has grounds hold state actors, and to a lesser
extent, non-state actors, accountable for their actions in hard cases such as the one above.
Disagreements may still occur, but if we have an understanding of what criteria
distinguish one act of intentional killing from another, at least we know what we should
be disagreeing about. In cases such as the one above, if the state actor wants to morally
justify the act as an act of targeted killing rather than an act of extra-judicial execution,
they would have to show that Old Boy Leroy was acting in a task-defined role that made
him morally liable to attack; i.e. a task-defined role that made him an ex ante threat. If it
could be shown that this was not the case, the justifiability of the act could be challenged
on those grounds, or the claim could be made that the act was in fact an act of extrajudicial execution, and justification could be demanded on the basis of desert.
56
I would argue almost impossible without some form of judicial proceeding, but I will not do it here.
Such regulations might resemble the Israeli High Court decision laying out ground rules for the
conditions necessary for targeted killing. For an account of this decision see pp. 327-328 in Guiora, A.
(2004). "Targeted Killing as Active Self-Defense." Case Western Reserve Journal of International Law
36(2/3): 319-334.
57
39
Final Draft
Adam C. Gastineau
07/2011
Conclusion
I have proposed that the term targeted killing is best defined as the authorized, intentional
killing of a specific individual or group of individuals who are targeted on the basis of
their role in a given armed conflict. I have then argued that this definition offers several
advantages. First of all, it allows us to differentiate between targeted killing and other
types of intentional killing that may occur during armed conflict. Secondly, because it
allows us to clearly distinguish acts of targeted killing from other acts of intentional
killing that may be difficult to morally justify (such as assassination and extra-judicial
execution) it frees us from any negative normative baggage that may come along with
these concepts. This in turn allows us to begin our evaluation from a neutral position
when considering the normative value of a particular act and hopefully see the issue more
clearly. Finally, this definition allows us to identify acts outside of the military context as
acts of targeted killing. This is important as world conflict begins to shift away from
conventional ‘wars’ between states and more toward non-traditional asymmetric conflict
involving both state and non-state actors where the lines between civilian security roles
such as policing and traditional military security roles begin to blur and non-state actors
play a greater role. The definition I have proposed allows for this changing environment
by placing importance on the moral legitimacy of the authorization, rather than the
institutional status of the authorizer, the task-defined roles of the participants, rather than
the narrower traditional definition of ‘combatant’, and by broadening the meaning of
‘armed conflict’ to include more than just state vs. state warfare.
In formulating this definition, I have attempted to illustrate how targeted killing, so
defined, differs from other acts of intentional killing with which it is often conflated. Acts
of assassination require no authorization and target prominent individuals identified by
‘name’ (e.g. Joseph Stalin) or by ‘definite description’ (e.g. Premier of the Soviet Union)
for, arguably,58 political purposes (narrowly or broadly defined). Acts of extra-judicial
execution require government authorization and target individual entities identified by
‘name’ in retribution for some ex post act. Drawing these distinctions also acknowledges
that targeted killing, assassination, and extra-judicial execution are justified, if in fact
they are justifiable, by different criteria. Acts of assassination are morally justified on the
basis of their political purpose, widely or narrowly defined. Acts of extra-judicial
execution are justified on the basis of desert. Acts of targeted killing are justified on the
basis of the task-defined role of the target making them liable to attack (or not) within the
given armed conflict. Knowing which act has occurred therefore affects which criteria
we use to judge the moral value of the act. In the final section of this paper I have
attempted to illustrate how this might be done using the case of Old Boy Leroy. In this
case I have argued that the act of intentional killing of OBL is best judged to be an act of
targeted killing as the case meets the definition I have proposed, and because doing so
allows for clearer analysis by avoiding the negative moral baggage of both assassination
and extra-judicial execution, and also allows us to more easily judge the act by means of
the task-defined role of the target rather that a very wide definition of
‘political’ purpose or a difficult to justify concept of desert.
58
See Knoepfler, S. (2010). "Dead or Alive: The Future of U.S. Assassination Policy Under a Just War
Tradition." New York University of Law & Liberty 5(457): 457-499. for further discussion.
40
Final Draft
Adam C. Gastineau
07/2011
Altman, A. and C. H. Wellman (2008). "From Humanitarian Intervention to
Assassination: Human Rights and Political Violence." Ethics 118(2): 228-257.
Altman, A. and C. H. Wellman (2009). A Liberal Theory of International Justice. Oxford,
Oxford University Press.
Bellamy, A. J. (2005). "Responsibility to Protect or Trojan Horse? The Crisis in Darfur
and Humanitarian Intervention after Iraq." Ethics & International Affairs 19(2): 31-54.
David, S. R. (2003). "If Not Combatants, Certainly Not Civilians." Ethics & International
Affairs 17(1): 138-140.
David, S. R. (2003). "Israel's Policy of Targeted Killing." Ethics & International Affairs
17(1): 111-126.
Gross, M. L. (2006). "Assassination and Targeted Killing: Law Enforcement, Execution
or Self-Defence?" Journal of Applied Philosophy 23(3): 323-335.
Gross, M. L. (2010). Moral Dilemmas of Modern War : Torture, Assassination, and
Blackmail in an Age of Asymmetric Conflict. New York, Cambridge University Press.
Guiora, A. (2004). "Targeted Killing as Active Self-Defense." Case Western Reserve
Journal of International Law 36(2/3): 319-334.
Haydar, B. (2005). "The Ethics of Fighting Terror and the Priority of Citizens." Journal
of Military Ethics 4(1): 52 - 59.
International, A. (2003). Israel and the Occupied Territories: Israel Must End its Policy of
Assassinations, , Amnesty International.
Kasher, A. and A. Yadlin (2005). "Military Ethics of Fighting Terror: An Israeli
Perspective." Journal of Military Ethics 4(1): 3 - 32.
Kasher, A., Yaldin, Amos (2005). "Assassination and Preventive Killing." SAIS Review
25(Number 1 Winter-Spring): 41-57.
Knoepfler, S. (2010). "Dead or Alive: The Future of U.S. Assassination Policy Under a
Just War Tradition." New York University of Law & Liberty 5(457): 457-499.
Lang, A. F. (2009). Authority and the Problem of Non-State Actors. Ethics, Authority,
and War: Non-State Actors and the Just War Tradition. E. A. S. Heinze, Brent J. New
York, Palgrave Macmillan: 47-71.
McMahan, J. (2004). "The Ethics of Killing in War." Ethics 114(4): 693-733.
41
Final Draft
Adam C. Gastineau
07/2011
McMahan, J. (2006). "On the Moral Equality of Combatants*." Journal of Political
Philosophy 14(4): 377-393.
McMahan, J. (2008). Aggression and Punishment. War Essays in Political Philosophy. L.
May. New York, Cambridge University Press: 67-84.
Miller, S. (2009). Terrorism & Counter-Terrorism: Ethics and Liberal Democracy.
Malden MA, Blackwell.
Miller, S. (2011). Social Institutions. The Stanford Encyclopedia of Philosophy. E. N.
Zalta.
Obama, B. (2011). Remarks by the President on Osama Bin Laden. Washington D.C.,
Office of the Press Secretary.
Primoratz, I. (1990). "What Is Terrorism?" Journal of Applied Philosophy 7(2): 129-138.
Primoratz, I. (2010). Terrorism. The Stanford Encyclopedia of Philosophy. E. N. Zalta.
Richelson, J. T. (2002). "When Kindness Fails: Assassination as a National Security
Option." International Journal of Intelligence and CounterIntelligence 15(2): 243 - 274.
Scheffler, S. (2006). "Is Terrorism Morally Distinctive?" Journal of Political Philosophy
14(1): 1-17.
Searle, J. R. (1983). Intentionality An essay in the philosophy of mind. New York,
Cambridge University Press.
Statman, D. (2003). "The Morality of Assassination: A Response to Gross." Political
Studies 51(4): 775-779.
Statman, D. (2004). "Targeted killing." Theoretical Inquiries in Law 5(1): NA.
Stein, Y. (2003). "Any Name Illegal and Immoral." Ethics & International Affairs 17(1):
127-137.
Walzer, M. (1977). Just and Unjust Wars. New York, Basic Books.
42