A well established principle in the morality of war is the principle of

advertisement
Guy Sela
Fighting without Proportionality
Dear all,
As will be obvious to anyone who takes even a brief glimpse at this draft, this is a very sketchy
one (even for a draft). One reason for the sketchiness (apart from my inability to provide
anything clearer at this stage) is that my thoughts on the matter are still changing rapidly as I
keep thinking about it. As you will also see the foreword promises a lot more than the paper
delivers. What I have here is basically the first part of the argument on how ‘proportionality’
considerations work (or rather what work is done by considerations that appear to be
proportionality) in individual self-defense cases. The second part which discusses the application
of this framework to war is missing altogether. In my talk, however, I hope to complement this by
focusing mainly on warfare and talking about self-defense only inasmuch as it is necessary for
understanding the framework. As I noted, my views here are still changing and I’m constantly
being influenced by discussing these issues with others, so I really look forward to discussing this
with all of you. It goes without saying that I’d appreciate it if you could refrain from citing this
draft without permission.
When we think about war and self-defense, we often assume – an assumption which is
supported by our intuitive judgments of hypothetical cases – that there must be some
proportionality between the harm one tries to avert and the harm one may inflict on his
adversary in order to avert the former. This assumption, while widely shared, is not at all
obvious or unchallenged. Some thinkers have cast doubt on the validity of this
requirement and suggested that causing harm in self-defense may be permissible even
when the harm caused is not ‘proportional’ to the harm averted. In this paper I wish to
explore this claim and its potential implications for the ethics of war. In the first section I
will outline the argument against proportionality offered by Larry Alexander. In the
second section I will corroborate Alexander’s argument by suggesting that our intuitions
regarding hypothetical cases can be explained away. My suggestion is that while
sometimes there is no requirement of proportionality, in some cases where the harm
averted is ignorable, the victim may have a duty of beneficence towards the aggressor
-1-
which makes it impermissible for him to harm the aggressor in self-defense. I will then
move on to examine what implications this argument might have regarding the
permissibility of fighting in war. I will argue that even if war is taken to be self-defense
en masse, it does not follow that the conditions in which Alexander’s argument applies
are the conditions normally prevailing in a situation of war.
1. Disproportional Defenses
Many people believe that if someone confronts me in a dark alley and tries to take my
bag in which I have something of small value (say, my favorite – perhaps even
irreplaceable – Barbara Streisand CD) and if my only means of keeping my bag is to kill
the robber, I must give up my bag (and the CD) rather than kill the robber.
Let us suppose, then, that I really like my Barbara Streisand CD1. In fact, I like it so much
that I’m buying a very sophisticated safe and put it inside the safe. Surely, nobody would
deny me the right to do that. But, still, I can’t sleep at night. Thoughts about my precious
Barbara Streisand CD being stolen keep disturbing my dreams, so I decide I need some
better means of protection. I take my CD (in the safe) and move to a small uninhabited
island. To be even safer, I hide the safe in one part of the island which I then encircle
with an electric fence and put some fierce dogs to watch it (I also put very clear signs
warning strangers of the existence of the dogs and the fence).
Suppose now that a burglar, who happens to be a huge fan of Barbara Streisand, decides
he will have my Barbara Streisand CD, no matter what 2. He learns of my hiding place
1
This argument is a slightly revised version of the argument made by Larry Alexander in [REFERENCE].
For those who think the example is a bit too outlandish, I highly recommend KieÅ›lowski’s final episode of
the Decalogue ‘Thou shalt not covet thy neighbor’s goods, nor anything that is your neighbor’s’. In that
2
-2-
and embarks on the island seeking the safe. He stands no chance against my sophisticated
defenses, but he optimistically (he is a Streisand fan after all) believes he does. He
approaches the fence and tries with his poor electrical skills to tamper with it. I watch
from afar, knowing he will soon be electrocuted. I could save him by switching off the
electricity, but am I really required to do that? Now suppose that I did (or that he
miraculously survived the fence without my aid). He now faces the dogs. Naively, he
tries to divert their attention with a slice of meat. My dogs are well fed and anyway they
are very well trained and won’t be distracted so easily. Soon they’ll be at his throat.
Again, I watch from afar, considering whether I should activate the whistle that will call
them to halt. Am I under an obligation to do so? Finally, the burglar got past the dogs
(with or without my help) and is now confronting the safe. I see him sweating and getting
extremely tensed over not being able to break the code and (having some medical
expertise) I can tell that he will soon get a heart attack. This can of course be avoided if I
provide him with the code so that he need not make further effort, but does anyone think
seriously that I ought to do that?
It seems, then, that I am allowed to protect my relatively non-valuable Barbara Streisand
CD by lethal means. But if that is the case, why am I not allowed to protect the same CD
by shooting a robber who tries to obtain it from me? One answer might be that whereas in
the island example I provided proper notice about the danger to which potential robbers
expose themselves, the same is not true about the dark alley case. This answer does not
seem convincing though. I may be able to warn the dark alley robber, show him the gun
episode, two brothers inherit their father’s rare stamp collection, which they then become obsessed with
and try (without success) to protect against theft. Feel free to change Barbara Streisand’s CD with any of
the rare stamps featuring in that film.
-3-
and even shoot a warning shot. Another, more promising, line of argument suggests that
what might make the difference is that in the dark alley case I will be killing the robber
whereas in the island case I will merely be allowing him to die. The problem with that
answer is that it seems that in the island case my role is far more active than merely
allowing the death of the burglar. It cannot be that I am similar to another guy who ends
up on the island and does not switch off the current, activate the whistle etc. After all, I’m
the one who put the fence/dogs there in the first place. Moreover, suppose that the current
all of a sudden went down and the dogs fell asleep. I see the burglar’s boat approaching
the island. May I not actively switch on the current and wake up the dogs knowing that
they will kill the burglar? And, finally, suppose that I forgot to switch the current on and I
now see the burglar tampering with the fence. May I not switch it on now, killing him at
once (after warning him, making sure he’s seen the notices etc.)? After all, that’s
precisely what the fence is there for. And if I may switch on the current after providing
appropriate notice, why may I not shoot the gun after doing the same?
This, in a nutshell, is Larry Alexander’s argument against proportionality. The argument
is meant to demonstrate that when we are protecting our rights (as opposed to when we
are inflicting harm as a matter of retributive desert) the amount of harm we may inflict
need not be proportional to the goods which we are entitled to protect. If Alexander is
correct, one may exercise a disproportional amount of force when acting in self-defense.
What should we make of this argument?
2. Some other Problems with Proportionality
-4-
A possible objection to Alexander’s argument – one which I already mentioned and
which I think poses a serious threat to the argument – is that we shouldn’t make much of
intuitive judgments about outlandish cases. While intuitions sometimes play an epistemic
role in telling us what our moral duties are3, this is rarely the case in situations which
extend far beyond our day to day experience. When it comes to situations which we find
it hard to imagine ourselves in, our intuitive judgments – if we have any – bear little
weight.
I think that this objection is powerful and plausible4 but it is not one which Alexander’s
adversaries can invoke. The objection assumes (quite plausibly) that it’s hard to imagine
oneself putting so much effort into protecting a Barbara Streisand CD (or something of
equal value) and that it’s even more difficult to imagine anyone risking their lives to
obtain it. Surely, there must be some other missing feature of the scenario, one which will
make it at least somewhat easier to conceive. Maybe the robber is not only after the CD
but also after the rest of the valuables I left on the island, or maybe he is simply insane,
and who says he’s not going to harm me just to make sure I don’t try to get my CD back?
This is all sound and true, but what about the case we began with? Can’t the same
objections be made there too? How do I know that once the robber took my bag and
found out that all I have is a Barbara Streisand CD he won’t try to harm me in his
frustration? And suppose that he tells me that he’s only after the CD, how can I believe
that5? And, finally, suppose that I pull out the gun, tell the robber that all I have is a lousy
3
[Sinnott-Armstrong, Liao etc.]
I have a long list of doubts about the use of intuitions in contemporary philosophical writing which I hope
to develop elsewhere.
5
When I presented this as an objection to the weight of our intuitive judgments about proportionality,
someone responded that we could imagine that we know this guy is only after the CD because he
committed numerous robberies in the past and in all of them he only took CDs and left the owners intact,
4
-5-
CD and threaten to kill him. I even fire a warning shot and still the guy keeps coming
towards me. What would be more natural than to think that the guy is insane?
It seems, then, that to the extent that we shouldn’t rely on our intuitions about outlandish
cases, not only Alexander’s cases must be disqualified as too far from our daily
experience but also the cases which suggest that self-defense is subject to proportionality
requirements. We need a new argument for proportionality, one which does not rely on
intuitive judgments, but it is not easy to find one. More importantly, it seems that the
rationale that underlies self-defense is not in line with such arguments. Nonconsequentialist accounts of self-defense6 suggest that there’s something about the
aggressor’s doing which makes it permissible to harm him. It may be the fact that he
forfeited his rights, or the fact that he is culpable, or the fact that his actions transgress on
the victim’s protected sphere or something like that. Whichever it is, it can be invoked as
easily when someone tries to dispossess me of my Barbara Streisand CD as when he tries
to dispossess me of my life or limb.
Even more importantly, the notion of proportionality suggests some arithmetical relation
between both sides of the equation. The relation needn’t necessarily be a straight line
function (of the form Harm Inflicted <= N * Harm Avoided) but it does need to tell us
something about the relation between the harm inflicted and the harm avoided.
Proponents of the proportionality requirement, though, do not provide any explanation of
this relation; all they tell us is that someone may not violate a very important interest of
the attacker (his life) in order to preserve a negligible interest of the defender (like $20 or
this is well known because it was published (alongside the robber’s picture) in every newspaper etc. I think
the reply demonstrates my claim about outlandishness nicely.
6
I shall focus in this paper exclusively on non-consequentialist accounts of war and self-defense.
-6-
a Barbara Streisand CD). Whenever we start asking questions about other kinds of
interests – may one kill to avoid being raped? may one break the robber’s jaw to preserve
his CD? – proportionality advocates tell us that these things are too complicated to tell. It
should be stressed that this objection is not about the vagueness of the proportionality
requirement or the lack of a clear formula. Many moral questions have only vague
answers and it is no (serious) objection to a theory that it cannot provide clear boundaries.
The claim here is rather that the proportionality requirement is so under-developed that it
raises the suspicion that it has nothing to do with ‘proportions’ at all.
Finally, suppose that I am attacked not by a single assailant, but rather by a horde of them
and I need to kill each and everyone to defend myself. If the proportionality requirement
is valid then, presumably, there will be appoint in which the number of assailants (20?
100? 1,000?) will be so vast that it will be out of proportion to kill all of them in order to
save my life. But most people intuitively think that the number of assailants is irrelevant
and that I may kill as many of them as need be to save myself. How can that sit
comfortably with the claim that the harm caused needs to be proportional to the harm
averted?
3. An Alternative Approach to Proportionality
It seems that it might be worthwhile to explore different ways of understanding what’s
going on in ‘proportionality’ cases7. Let’s leave self-defense for a moment and look at a
different kind of examples where what one gains and what one loses are not the same.
Consider the following. You pass by the pond in the park, as you do every day on your
7
Since I argue that there are no proportionality restrictions on the right of self-defense, when talking about
the proportionality conditions in self-defense the word proportionality will sometimes be in inverted
commas.
-7-
way to work, when you notice a toddler drowning in the pond. His parents are nowhere to
be seen, there’s nobody around and he seems to be in real life danger. On the other hand,
the pond is not deep and you swim quite well. You could easily jump in and save the
toddler. The problem is that you’re carrying your favorite Barbara Streisand CD with you
and if you jump to the pond it will be ruined (you don’t have time to lay it down and then
jump, etc.). I believe that many people would agree that in that situation they ought to
save the child at the expense of the CD. The moral idea underlying this intuition is that
sometimes we have duties to benefit others. While many of our duties are duties not to
harm others, sometimes morality requires us to actively benefit others; and sometimes we
are required to do that even though it involves a setback to our own interests.
Duties of beneficence, unlike duties not to harm, are mostly imperfect duties. Imperfect
duties are duties we have to fulfill, but we don’t necessarily have to do so here and now.
The classical example of such duty is the duty to give charity. Whereas many people
think that they are quite well off and have a duty to give charity, it is rarely the case that
they ought to discharge this duty at any given moment or with respect to any specific
pauper. Duties of beneficence can sometimes be perfect duties, i.e. duties we have to
perform here and now, like in the case of the child in the pond. In that case the magnitude
of the potential harm to the child, the fact that you are the only one able to help, the
relatively low cost of offering help, the immediacy of the required help and so on, all
contribute to making the duty of beneficence in this case a perfect duty. But in most cases
our duties to benefit others are imperfect. Think, for example, about less extreme cases of
saving people from harm. While it is surely the right thing to let the driver ahead of you
know that his rear lights are out of order, we don’t have to do it on each and every
-8-
occasion. We have a considerable amount of discretion as to when exactly we ought to
discharge our duty and this pertains not only to issues such as whether it is nighttime or
daytime but also to whether we can convey this message without much effort and how
late we are to our next meeting.
Consider another variant of the pond case. When I pass by the pond I see one of my
neighbors, a rather disagreeable character who always makes loud noises when I’m trying
to sleep, blocks my parking space and rudely ignores me when I greet him hello. He
stands dangerously close to the pond and when I greet him and gently point out to him
that he might want to move farther away from the pond he starts cursing me and shouts
‘Mind your own business!’. As he does that he clumsily falls to the water, his head hits a
rock and he is in great need of help. Am I under a duty to help him? Generally it seems
that when I see drowning people and I can offer them help without any serious risk to
myself, I ought to do so. But this does not entail that I must offer help to all needy people
that I see. Not even to all very needy people. Considerations like who else is there who
can offer help, how much effort will it require of me and so on are considerations I may
legitimately take into accounts. And so are considerations which involve the type of
relationship I have with my prospect beneficent.
If I only have to give charity on some occasions or to some of the people who ask for it, it
seems perfectly acceptable to decide to give to those whom I like and to disregard those
whom I don’t like. I am not suggesting here that the grounds for preferring some people
over others as recipients of benefits are completely arbitrary. Nor am I suggesting that my
choices in that respect cannot be subject to criticism. We would rightly criticize someone
who chose to give charity only to white paupers but not to black ones. What I’m rather
-9-
claiming is that inter-personal considerations may legitimately play a role in carrying out
duties of beneficence.
To appreciate this point, compare it to instances of the duty not to harm. Whether I like
my neighbor or not, whether he’s rude or nice, agreeable or disagreeable, are all
irrelevant to the question whether or not I may harm him. I am not allowed to hit my
neighbor (or anyone else) even if I really don’t like him (for good reasons). On the other
hand, that I dislike the pauper in the next street (for good reasons) is a perfectly good
ground for deciding not to give him charity. To summarize this point, duties of
beneficence are often imperfect duties, we have a lot of discretion as to where, when and
how we discharge them and inter-personal relations play a role in their fulfillment.
Now let’s go back to cases of self-defense. Suppose that there is no proportionality
requirement in self-defense and that I have the right to use as much force as needed to
repel an attack on one of my protected interests. To say that I have a right merely means
that it is within my discretion to use force or not to use it; it does not mean that using
force will always be the right thing to do (cf. I have the right to demonstrate for greater
inequality between men and women does not mean that it is the right thing to do). The
use of the word right in both sentences may be a bit confusing so perhaps it’s better to say
that it’s not always justified to use force even when it is at my discretion whether to use it
or not. What makes the use of force not only ‘within my right’ (at my discretion) but also
‘right’ (justified)? Many things can do the justificatory work. That I may suffer
substantial losses if I don’t use force (I’ll be killed or will lose a limb) sounds like one
excellent reason to exercise force to repel the attack. But it is not the only reason. The
fact that this person has attacked me a hundred times in the past is another good
- 10 -
justification; after all, I am not required to demonstrate an endless amount of restraint.
The same goes for a case when I have been attacked a hundred times by various different
people (rather than by the same assailant). And another justification might be that the
person who attacks me is someone who does not deserve leniency.
Suppose that people impermissibly and unjustifiably invade your house and violate your
privacy and property rights (because you are a quasi-celebrity or something like that). At
first, if you do more than shout at them and maybe call the police it might be thought that
you are acting excessively. But after the fourth invasion on the same day it seems quite
understandable that you may do something more extreme, like punch the invaders on
their nose. But if we think about it in terms of proportionality limits on self-defense, it is
not obvious why this should be the case. After all, if the conditions for proportional selfdefense were not met the first time, how were they met after several recurrences? But if
we consider the alternative, i.e. that there are no proportionality limitations on selfdefense and that you were entitled to punch the invaders the first time they entered your
home the story starts making sense. Whereas you were entitled to punch the first group of
invaders, it wouldn’t have been justified to do so. On the other hand, you seem on firmer
grounds when you punch the fourth group of invaders.
In this respect the so called proportionality requirement in self-defense is really a red
herring. It obstructs the real nature of excessive-use-of-force claims. When we claim that
someone exercised an excessive use of force in self-defense, what lies at the heart of the
claim is not that since the amount of force used was disproportionate to the harm averted,
the attack was not an instance of permissible self-defense. The claim is rather that even
though the attack was an instance of permissible self-defense, it was unjustifiable. This
- 11 -
may come across as a technical distinction, but it isn’t8. Here are a few practical
implications of this new understanding of ‘proportionality’ claims.
First (and least importantly), to say that one has a right (like the right of self-defense) is
to say that he is the one who decides whether to exercise the right or not. If I have the
right of way then I may decide to go first or to let the other vehicle proceed, but the right
is mine and the other vehicles must submit to my decision. The same goes for the right of
self-defense. If I have a right of self-defense, then it is my call whether to exercise it or
not, even though on occasions I may exercise it unjustifiably (just like I might exercise
my right of way even thought the other driver is obviously in a much greater hurry).
Second and related to the first point is the fact that ‘disproportional’ use of force in selfdefense doesn’t change the roles of the attacker and the defender. In some authors’
writing on self-defense9 once the former-defender reacts with an excessive amount of
force he becomes an aggressor whose attacks may rightfully be repelled by the formeraggressor, who is now defending himself against the excessive defensive measures, or by
third parties. So in the CD example, if I pull out my gun to shoot the guy who tries to
snatch my CD and his only means to save himself from being killed is by killing me, he
may do so. Since his violation of my rights did not amount to justifying the use of lethal
(disproportionate) force on my behalf, I am now unjustly attacking him and he may,
therefore, use force to defend himself against my attack. This makes some sense if
proportionality is a limit on the scope of the right of self-defense. But if, as I maintain,
Perhaps it may be useful to point out an equivalent distinction about property rights. That I’m not
fulfilling my imperfect duty to give to charity does not annul my property rights, nor does it make any
beggar entitled to my money.
9
Jeff McMahan, Cecile Fabre, David Rodin(?) etc. [REFERENCES]
8
- 12 -
there are no proportionality constraints on self-defense, only imperfect duties of
defenders towards aggressors, then the fact that I failed to act on my duty on some
occasion does not change the nature of the situation. I am still defending my rights from
being violated.
A third (again, related) point has to do with the outcomes of acting ‘disproportionately’.
Not all wrongful actions carry the same range of justified responses. Think about people
who abuse their right to freedom of speech by demonstrating for more money to the rich.
It may be appropriate to boo them, to scorn them, perhaps even to dissociate oneself from
such sociopaths, but it is by no means appropriate to send them to prison, to hit them, or
even to deny them the right to demonstrate again. More generally, only those who violate
other people’s rights may, normally, be disentitled from some of their rights. Other
wrongs normally call for responses that do not involve violating the wrongdoers’ rights.
If ‘disproportional’ self-defense is a violation of the aggressor’s rights then it may well
entail that the defender who acts ‘disproportionately’ thereby loses some of his rights. If,
on the other hand, excessive use of force in self-defense is a failure to fulfill an imperfect
duty then it has no such implications.
The fourth, and most important, point concerns the range of legitimate considerations to
be taken by the defender. I already mentioned in passing two legitimate considerations
which only apply if there is no ‘proportionality’ restriction on the scope of using force in
self-defense. The first has to do with my relationship with the aggressor. I have more
reason to use less than maximum force when the person trying to take my CD is my nice
neighbor, who is a good and supportive neighbor and often helps me when I’m in need
(alas, he’s obsessed with my Barbara Streisand CD and hence the conflict) than I have to
- 13 -
use less than maximum force when engaging with my obnoxious neighbor. The second
has to do with the number of attempts made at my CD. While it is excessive to kill
someone who tries to take my CD, it is quite understandable that when this happens fifty
times in a row, I may be less inclined to show restraint. None of these make sense if
proportionality is an intrinsic limit on the scope of right of self-defense. I may not violate
the rights of anyone, whether I like them or not, whether they’re nice or not. Even more
clearly, it is unthinkable that I may go and hit someone impermissibly just because I was
punched numerous times earlier. This only makes sense if using excessive force is a
violation of some imperfect duty which I only need to fulfill some of the time.
Moreover, if ‘proportionality’ really is ‘an imperfect duty to aggressors in disguise’ then
the range of legitimate considerations is even broader than that. Consider, again, the
imperfect duty of charity. One may justify his refusal to give charity on a certain occasion
by quite trivial considerations such as that he’s hungry and doesn’t want to be delayed for
lunch. If it is indeed the case that the person in question habitually fulfills his duty to give
to charity and if no overriding circumstances obtain (the beggar is not about to die of
starvation, there are other people around etc.) then not much more is needed to justify not
fulfilling an imperfect duty on this or other occasion.
Finally, proportionality adherents should concede that once the proportionality
requirements (as well as the other requirements of self-defense) are fulfilled the defender
may exercise full ‘proportional’ force to defend his rights. By contrast, on my view,
imperfect duties of beneficence towards aggressors exist even when the amount of force
used by the defender is not excessive. Of course, when the amount of force used is not
- 14 -
excessive it will often be the case that the reasons for employing force outweigh the
reasons against doing so, but that does not always have to be the case.
I could go on to discuss the explanatory advantages of my proposal vis-à-vis the
proportionality view10, but it is time to move on to discuss the implications of my view
on the ethics of warfare.
See for example Daniel Statman’s article in Ethics [REFERENCE] on the success condition in selfdefense. As Statman rightly points out this condition is sometimes taken to be overridden by relatively
trivial considerations like the defender’s honor. This can be best explained if there is no proportionality
requirement.
10
- 15 -
Download