We are pleased to present our first installment of PEA Soup's collaboration with Ethics, in which we host a discussion of one article from each volume of the journal. The article selected from Volume 120, Issue 1, is Cécile Fabre's "Guns, Food, and Liability to Attack in War" (open access copy here). We are very grateful to Jeff McMahan for starting our discussion; his commentary follows beneath the fold…
Commentary on Cecile
Fabre, “Guns, Food, and Liability to
Attack in War”
The common or orthodox view of civilians who provide certain
essential forms of support for combatants in war – such as weapons, food, and
medical care – is that none of them acts impermissibly, even if the support
they provide enables unjust combatants to pursue an unjust cause with greater
effectiveness. Those who provide for the
combatants’ welfare needs (for food, clothes, shelter, medicine, and so on) are
never legitimate targets of attack. Yet
those who provide what combatants need specifically for military purposes are legitimate targets of attack, at
least during those times when they are actively engaged in the provision of
such support. These civilians are
morally vulnerable to attack even if they provide support only for just
combatants who fight by permissible means for a just cause.
Cecile Fabre’s “Guns, Food, and Liability to Attack in War”
offers a careful, sustained, and rigorous critique of this orthodox view of the
status of civilian suppliers to the military in war. Her arguments challenge two of the key
elements of this view. She argues,
first, that neither the permissibility of their action nor whether they are
liable to attack depends on whether civilians are suppliers of military
resources (civiliansm) or suppliers of welfare resources (civiliansw). And, second, she argues that in the case of
both civiliansm and civiliansw, whether they act
permissibly does depend on whether
the combatants they support are fighting for a just or unjust cause. It’s compatible with these claims for her to
take the view that while civilian suppliers on the just side act permissibly
and are not liable to attack, most or all of those on the unjust side act
impermissibly and are liable to attack.
But the position she defends is different. It is that although civilian suppliers on the
just side act permissibly while most or all of those on the unjust side act
impermissibly, none on either side are liable to attack.
Fabre identifies and criticizes two views that have been
thought to support the claim that the civiliansm are liable to
attack while civiliansw are not.
One is the “functionalist” view, according to which providing combatants
with what they need to fight constitutes a form of direct participation in war,
while providing them with what they need merely to live is not, and it’s direct
participation in war that grounds liability to attack. The other is the “moralized functionalist” view,
which holds that civiliansw, whether on the just or the unjust side,
fulfill a duty that their state has to minister to the well-being of all its
citizens and that this justification for their action exempts them from
liability to attack. By contrast, civiliansm whose side is pursuing
an unjust cause have no moral justification for providing their combatants with
the means of fighting and hence are not exempted from liability for their
contribution to the achievement of an unjust cause.
I will not discuss her objections to the two forms of
functionalism, which I find largely convincing – though Massimo Renzo of the
University of Stirling has some forceful criticisms of her arguments that I
hope to convince him to post on Pea Soup once the discussion is in progress. Instead I will offer some observations about
other elements of her argument as well as a few reflections of my own about the
issues she discusses.
It’s an implication of Fabre’s view that civiliansm
are in general not liable to attack and thus not legitimate targets of
intentional attack. She can, of course,
still justify attacks on munitions factories and similar targets, though the
harms that such attacks cause to civiliansm will have to be
justified on grounds of necessity or as the lesser evil, and will have to
satisfy the condition of proportionality, which they will do more readily if
they are unintended. But since the
standards of justification on grounds of necessity are more demanding than
those on grounds of liability, Fabre’s view imposes greater constraints on
attacks on munitions factories and similar targets than the orthodox view
does.
Suppose, however, that despite Fabre’s best efforts, we
continue to believe that civiliansm, at least on the unjust side,
are legitimate targets. And suppose
further that, because of Fabre’s best
efforts, we have become convinced that there is no difference between civiliansm
and civiliansw that is relevant to liability. In that case we will believe that civiliansw
on the unjust side are also legitimate targets of intentional attack, provided
the contribution they make to an unjust war is significant enough that killing
them to eliminate it would be proportionate.
In effect, the combatant status that the orthodox view assigns to
civiliansm would be extended to civiliansw as well. The category of combatants would thus be
expanded to include a significant number of civilians, who would then be
considered legitimate targets. It’s
important to note, however, that even if we are repelled by the thought of
endorsing such extensive killing of civilians, the killings would not
constitute terrorism. For the goal of
such action would not be to coerce surrender by means of terror and
intimidation. Rather, the action would
be literally defensive: it would be aimed at the goal of eliminating the threat
posed by combatants by disabling or incapacitating them through depriving them
of essential resources. Such action
would be permissible, for example, according to the view of killing in war
advanced by Scanlon in Moral Dimensions. On his view, the killing of civilians in war
is permissible if it is necessary, either as a means or as a side effect, for
the achievement of a proportionate military advantage.
A case that might be relevantly similar to one involving
civiliansw is that of a person who provides food for a kidnapper who
has no other means of obtaining it.
Suppose that killing this person would weaken the kidnapper by depriving
him of food, thereby enabling his victims to escape. Would it be permissible for the victims to
kill the food provider? Intuitively it
seems that it would be, at least given certain assumptions about the provider’s
beliefs and intentions. It seems that it
would be permissible if the food provider knew that the kidnapper was holding
victims in confinement and if he intended to assist the kidnapper’s action or
even if he was simply acting to make money.
On some accounts of the morality of defensive action, however, the food
provider would not be liable to be killed if he was nonculpably unaware that
the person he was feeding was a kidnapper.
The relation between the food provider and the kidnapper is
not, however, relevantly similar to that between civilian suppliers and unjust
combatants in war, according to Fabre.
She would concede that the food provider is liable to be killed because
his individual contribution to the ability of the kidnapper to continue to
confine his victims is substantial and because he is culpable for making that
contribution. Fabre’s view is that athe
reason civilian suppliers aren’t liable to attack has nothing to do with
whether the contribution they make to the action of unjust combatants takes the
form of military or welfare support; rather, the reason they aren’t liable is
that their causal contribution is marginal and the degree of their moral
responsibility for it is low. For
civilian suppliers to be liable to attack, their causal contribution and degree
of moral responsibility would have to pass a certain threshold that in fact
they almost never pass.
In advancing this position, Fabre is in effect defending
what she calls the “traditional account of liability in war” against the charge
that it entails an expansive doctrine of civilian liability. This traditional account is that the proper
criterion of liability to attack in war is moral responsibility, which tends to
presuppose causal responsibility, for a threat of wrongful harm that is
sufficiently serious that killing those responsible for it could be
proportionate as a means of averting it.
She calls this account “traditional” because its earliest
roots are in the writings of classical just war theorists such as Aquinas,
Vitoria, Suárez, and Grotius. Yet
because this account has been in abeyance for several centuries, I think the account
that displaced it and has subsequently reigned as the orthodox view deserves to
be called the “traditional” view.
According to this view, all combatants are legitimate targets,
irrespective of whether they serve a just or unjust cause, while all noncombatants
are immune from attack simply because they pose no direct threat to enemy
combatants. It is true, as Fabre points
out, that in recent years the classical approach to discrimination in war has
been revived and now presents a challenge to the reigning orthodox view. But it would, I think, be less confusing if
we called this revised reincarnation of the classical approach the
“neoclassical” account, and reserved the label “traditional” for the account it
seeks to overthrow. I will therefore use
“neoclassical” to refer to what she calls the “traditional” account, and
“orthodox” to refer to the rival account that has been dominant for several
centuries.
As a proponent of the neoclassical account, I have good
reason to hope that Fabre is right when she argues at length that the contributions
by both individual civiliansm and individual civiliansw
to the action of unjust combatants are causally insignificant and that the
moral responsibility these civilians bear for whatever causal contributions
they make is also negligible because of their unavoidable ignorance of relevant
facts. For it would certainly count
heavily against the intuitive plausibility of any view of the ethics of war if
it implied that civilians can be extensively liable to attack in war. That would be too close for comfort to an
embrace of terrorism, even if it wouldn’t literally be an endorsement of
terrorism.
But Fabre’s argument offers even more than a promise that
the neoclassical view has the resources to give terrorism a wide berth. For she claims that, when properly understood
and applied, the neoclassical account implies that few, if any, civilians are
liable to intentional attack in war, at least in the conditions in which wars
are actually fought. Since the orthodox
account, which embraces a functionalist criterion of liability, holds that
civiliansm are liable to
attack, Fabre’s argument, if correct, would show that the neoclassical account
actually offers greater protection to
civilians than the orthodox account.
That would be a gratifying, albeit rather surprising, conclusion for
proponents of the neoclassical account, and a discomfiting one for defenders of
the orthodox account.
As I’ve indicated, Fabre makes three basic claims in support
of her contention that the neoclassical account exempts civilian suppliers of
the military from liability: that the causal contributions that individual
civilians make to the action of unjust combatants are slight, that the degree
of their individual moral responsibility is very low, and that there is a
threshold below which certain forms of causal contribution and certain degrees
of moral responsibility are insufficient for liability to attack. I will say only very little about her claims
about causal and moral responsibility.
But I do have serious doubts about her claim that there’s a threshold
for liability that the contributions of individual civilian suppliers rarely,
if ever, exceed.
There’s one point that Fabre doesn’t make explicit that
helps to support her contention that the degree of an individual’s causal contribution
to an unjust threat is relevant to the degree of his liability. This is that when his causal contribution is
slight, attacking or killing him can do very little to avert the threat. According to what I think is the best
understanding of liability, a person can be liable to suffer a harm only if
harming him is instrumental to, or an unavoidable side effect of, the
prevention or correction of a wrongful harm for which he is or would be
responsible. If this is right, military
attacks against civilians are likely to inflict harms on them that exceed what
they could be liable to suffer as a means of eliminating their marginal
contribution to their side’s unjust war.
That said, I do find one of Fabre’s claims about causal
contribution puzzling. She writes that
“more often than not, welfare infrastructures and their agents will, in
wartime, help both just and unjust combatants. … The same can be said, mutatis
mutandis, of munitions factory workers.” (58-59) There are undoubtedly cases in which weapons
manufacturers supply weapons to both sides in a war. In those cases, the acts of munitions workers
may contribute equally to the action of just and unjust combatants. But I suspect that this is true in only a
small minority of cases. In World War
II, for example, military production in the US was wholly separate from
military production in Germany. And I
don’t know of any cases in which the suppliers of food or medical care to one
army have also been the suppliers to the opposing army as well. (Since I wrote this paragraph, Fabre has
explained to me the misunderstanding of her text that underlies it. In case some other readers have made the same
mistake, I’m leaving the paragraph in to allow her the opportunity to clarify
her position in the subsequent discussion.)
On the individual moral responsibility of unjust combatants
and their civilian suppliers, Fabre says that “unavoidable ignorance either of
the fact that they are posing a threat or of the threat’s wrongfulness negates
moral responsibility, from which it follows that combatants who are unavoidably
ignorant of either fact are not liable to being killed. Similarly, civilians who are unavoidably
ignorant either of the fact that they are contributing to lethal threats or of
the wrongfulness of those threats are not liable either. … [And] the epistemic
burdens faced by civilians are absolutely enormous – so much so, in fact, that
those civilians can plausibly be regarded, in many cases, as unavoidably
ignorant of the relevant facts.” (58)
The problem here is in understanding how literally we’re to take the
notion of unavoidability, both in relation to ignorance of empirical facts and
in relation to ignorance of normative truths.
It’s hard to see how unjust combatants who pose an unjust threat could
ever be unavoidably ignorant of the fact that they pose a threat. Their ignorance must therefore be of the fact
that the threat they pose is unjust.
That form of ignorance is seldom literally unavoidable in a way that
would absolve them of all moral responsibility for their action. Even more obviously, unjust combatants can
surely know that there’s a significant risk
that the threat they pose is unjust, and that alone seems a sufficient basis
for the attribution to them of some responsibility for the wrongful harms they
cause. This is so even if the epistemic
obstacles they face are sufficient to absolve them of all culpability – though
when what they’re doing is intentionally killing people who are in fact innocent,
I find it hard to believe that they’re not culpable even to the slightest
degree. Parallel points apply to
civiliansm and civiliansw who supply unjust combatants
with the resources that enable them to fight.
Fabre concedes that some unjust combatants are responsible
for the wrongs they commit and indeed that some are even culpable. And she concedes that the same may be true,
though more rarely, of some civilians on the unjust side. But she also argues that these civilians,
even if they are culpable, are nevertheless liable to attack only if their
“contributions pass a certain threshold of causal and moral responsibility.” (61) She notes that I have criticized the view
that there is a threshold for liability and cites an example I give of a person
who decides to drive for reasons of self-interest knowing that in doing so she
will be imposing on others a tiny risk of hitting them with her car. Although she drives carefully, aberrant
conditions cause her car to go out of control and threaten a pedestrian. I have claimed that she is liable to be
killed if that is necessary to save the life of the pedestrian, even though the
degree of her moral responsibility for the threat she poses is certainly well
below any threshold that Fabre would regard as reasonable.
Fabre argues that rather than being a case in which there is
a significant moral asymmetry between the driver and the pedestrian, this is an
instance of what I have called a “symmetrical defense case,” in which each
party is morally permitted to attack the other.
If that is right, she says, then “the conflict between the conscientious
driver and the pedestrian is relevantly similar to that between the just
tactical bomber and the civilians who live near his target: neither party is
liable to direct attack at the hands of the other, and both are permitted to
infringe the other’s right not to be killed.” (61) I think, however, that there is a major
difference between the two cases.
Although the conscientious driver acts permissibly in driving, she has
no positive moral reason or justification
for imposing the risks she in fact imposes.
The tactical bomber, by contrast, has a moral justification for dropping
his bomb. He may even be morally
required to drop it. His justification
exempts him from liability, but a mere permission to impose risks does not,
when the choice to do so is voluntary.
In the book in which I discuss these issues, to which Fabre
refers, I suggested that the case of the conscientious driver challenges the
idea of a threshold but conceded that people may think there are no analogues
of this case in war. So I offered a
different hypothetical example to challenge the threshold view. In that example, however, the killing of
civilians operated by terrorist means to achieve a just cause rather than by
eliminating the contributions that individual civilians make to an unjust
war. So I will here give a revised
version of that example that involves killing civilians as a means of
eliminating their causal contributions to an unjust war.
Imagine a society in which decisions about the resort to war
and the continuation of war are made by plebiscite. The entire civilian population votes about
whether to go to war, or to continue to fight a war. If a majority votes in favor of war, a signal
is automatically sent to activate a large number of robots that are
preprogrammed to conduct attacks without any human involvement. Suppose that an unjust war has been initiated
in this way but that the population votes again every week to determine whether
the war will be continued. The people
who are the victims of this unjust war cannot defeat the robots but they know
that the initial vote in favor of the war in the enemy country was won by only
a tiny margin. Suppose that they can
identify roughly 1000 ardent proponents of the war, who they can be sure voted
for it, and that they can kill these 1000 civilians selectively, thereby
reversing the outcome of the vote in the next plebiscite. In that way they could bring the unjust war
to an end by eliminating these civilians’ contributions to it. I think it’s not implausible to suppose that
it would be permissible for them intentionally to kill those civilians and that
their justification would not be one of necessity or lesser evil but would instead
appeal to the claim that those civilians were liable to be killed by virtue of
their causal and moral responsibility for the unjust war. But all these people have done is to advocate
and vote in favor of the war, which means that their causal contribution and
degree of moral responsibility are surely below Fabre’s threshold, since they
are below those of typical civilian suppliers of the resources that unjust
combatants need to fight. It’s worth
stressing, moreover, that what’s true of the 1000 voters who are selected,
perhaps randomly, for targeting is equally true of all the other citizens, of
whom there may be many millions, who initially voted in favor of the war and
would do so again.
I think this shows the implausibility of the idea of a
threshold. As I indicated earlier, I
have reasons for wishing that the threshold view were right. I have been tempted to defend it myself. But I think we have to find another way to
defend our intuitions about civilian immunity in war.
Jeff McMahan
Rutgers
14 December 2009
Notes on Fabre: Contribution to the Good
Cecile Fabre, in her excellent and provocative paper, claims that there is no significant difference between the liability of civilians who contribute military equipment to the unjust side of a war (civilians-m) and the liability of civilians who contribute to the welfare of soldiers on the unjust side (civilians-w). I think that Fabre has overlooked a significant difference between the two groups.
Suppose that for every civilian-m and civilian-w on the unjust side in a war that is killed one wrongful killing of a person on the just side is prevented. If that is the case, we might then think it permissible to kill both civilians-w and civilians-m.
Compare
Cure – B is making a firearm for a villain, F, to kill V. B lives in a remote place. We could save V from being killed by F only by carrying out a missile strike on B, which will lead to B being killed. B knows that F will kill V and knows or ought to know that it would be wrong for F to do so.
It seems permissible to kill D to save V.
Weapon – C is working in a hospital caring only for a villain, G, who has a broken leg. If C cures G, as C knows, G will kill V and D knows or ought to know that the killing of V is wrong.
In this case, as Cecile Fabre suggests, C ought not to cure G and, further, it is permissible to kill C to save V.
That conclusion seems right. Perhaps things are different where there are groups assisters and groups of victims. I’m not sure.
Fabre makes claims about the degree of contribution required for liability, arguing that both civilians-w and civilians-m are not liable to be harmed. Suppose that we reject those threshold claims (as Jeff McMahan suggests we should, strongly supported by his robot army example).
Given the permissibility of killing in both Cure and Weapon does it follow that there is no difference in the permissibility of killing civilians-m and civilians-w?
Here is a distinction we might draw between civilians-m and civilians-w: civilians-w have a better chance of contributing to the good than civilians-m.
Many soldiers who are assisted during war (by curing them, or by providing them with weapons) will not in fact end up making any significant causal contribution to wrongful killings. They will not shoot anyone, or make any contribution to the shooting of anyone.
Suppose that 50% of soldiers on the unjust side in a war wrongfully kill (or make a significant causal contribution to a killing). Now compare these two cases:
Cure 2. Suppose that X, a soldier, is sick and will die unless D cures him. D cures X. X might wrongfully kill V. If X wrongfully kills V D has contributed to the wrongful death of V. But if X does not wrongfully kill V but rather goes on to live a valuable life D has contributed to X living a valuable life.
An action that has 50% chance of contributing to the bad and 50% chance of contributing to the good may be permissible and hence D’s action may be permissible.
Weapon 2. E provides Y with a weapon. If Y wrongfully kills V, E has contributed to the wrongful death of V. If Y does not kill V, E will have done neither harm nor good.
E’s action is clearly wrong. An action that has 50% chance of contributing to the bad and 50% chance of doing neither harm nor good is wrong.
Because many soldiers on the unjust side will not kill (or otherwise cause deaths) doctors might be justified in saving the lives of soldiers. Many soldiers will lead valuable lives and not commit harmful wrongs. But that fact does not warrant providing them with munitions, for providing them with munitions is only likely to contribute to the bad and not to the good.
This distinction between civilians-m and civilians-w need not depend on the intentions that these civilians might have, though that might be important. We might think that it is wrong to kill D in Cure 2 even if he intends to contribute to the war effort on the grounds that there is a chance that he will contribute to X’s valuable life. Or we might think that D’s intentions in Cure 2 make some difference to liability. If that is true, only doctors who intend to cure but who do not intend to contribute to the killing of those on the just side are not liable to be killed.
Against this distinction between civilians-w and civilians-m, it might be argued that it is wrong to cure a soldier on the unjust side who will try to kill even if he will not succeed. But that seems wrong. It is permissible (and perhaps even morally required) for D to cure X knowing that X will try but fail to kill V if X will then go on to live a valuable life not harming anyone else. Those on the just side need to be protected from being killed, not from attempts to kill them. That conclusion is even more plausible in this context where soldiers on the unjust side are often at least partially excused.
Furthermore, even if what civilians-w do in war is wrong all things considered given the probability that they will contribute to the good or contribute to the bad, we might be more inclined to excuse them. The closer a person is to being justified, the more we are inclined to excuse them. If a doctor is in error about the probability of contributing to the good by curing the soldier, his error is easier to justify than the error that munitions workers make.
Civilians-m can be excused only with respect to their mistaken judgement about the justification of the war; civilians-w can also be excused with respect to their assessment of the chance that their actions were justified even if they are on the unjust side – on the grounds that they are more likely to contribute to the good than the bad in curing soldiers. We should conclude that civilians-w are more likely to have partial or full excuses than civilians-m.
If the permissibility of killing in part turns on whether the person to be killed is excused (and I think it does, at least as far as proportionality is concerned), we should conclude that it is easier to justify killing civilians-m than civilians-w.
That is not to say that it is always permissible to kill civilians-m or that it is always wrong to kill civilians-w. There are many considerations that are relevant to answering these questions. But there is nevertheless an important difference between civilians-m and civilians-w: the latter group, unlike the former, often cause the good as well as the bad.
As McMahan notes, there is, according to Fabre, “a threshold below which certain forms of causal contribution and certain degrees of moral responsibility are insufficient for liability to attack.” It’s hard to assess Fabre’s claim and McMahan’s objections to it without having a good sense of how the threshold is supposed to work. I could use some help here. I would think that such a threshold would imply that the severity of the sorts of harms to which civilian suppliers are liable would depend on the degree of causal contribution and moral responsibility they have for the military threats that they make possible. So although, in typical wars, civilian suppliers may not exceed the threshold for the degree of causal contribution and moral responsibility necessary to be liable to lethal attack, they would be liable to some less harmful forms of attack: for instance, to non-lethal bombs that released the cold or flu virus into the civilian population via some aerosol device, thereby rendering them too sick to supply the combatants.
Until I hear more about how the threshold works and in particular how we’re supposed to assess one’s degree of causal contribution and moral responsibility, I can’t adequately assess McMahan’s claim that, in his robot army example, the 1000 selectively killed civilians’ “causal contribution and degree of moral responsibility are surely below Fabre’s threshold.” Assuming that their votes made possible the continuation of the unjust war and that the continuation of that war would lead to much more than a 1000 unjust deaths, then isn’t their causal contribution great enough to exceed the threshold?
I suppose it depends on how we are to assess causal contributions. If people had in the past been doing something that contaminated the ground water with lead but that the levels had never gotten to the point of causing any serious medical problems, but then I add 10% more lead to the ground water knowing that this will put it over the threshold where it becomes a serious health concern, do I bear only one-tenth of the responsibility or much more than that for the ensuing deleterious health effects on the population. I would think more. After all, it was my contribution that caused things to move from a harmless situation to a harmful one.
I wonder if something like the (non-moralized) functionalist view can be supported in the following way. It’s not that civilians-m more directly participate in (or contribute more to or hold more responsibility for) the war than civilians-w. It’s simply that we want to stop the delivery of the means of killing. To butcher a slogan, food and blankets don’t kill people, guns kill people. So, on this view, since civilians-m (roughly) provide what kills people and civilians-w do not, civilians-m are liable in a way that civilians-w are not. (This is similar to what is considered and rejected on pp. 42-44, but it doesn’t make use of the concepts of direct participation or contribution to war or enabling soldiers to fight, which seem to be the targets of the grounds for rejection.)
Hi Josh,
What constitutes “the delivery of the means of killing”? If providing combatants with food and blankets enables them to kill more people than they would otherwise be able to kill, does providing combatants with food and blankets constitute the delivery of the means of killing? If not, can you explain what you take to be the distinction between the means of killing and everything else that ensures that there will be more killing and explain why you think that it’s morally relevant.
It’s simply that we want to stop the delivery of the means of killing.
Surely, we want to stop the killing (presuming we’re talking about unjust killing) and not just the delivery of the means of killing. By stopping the delivery of food and blankets to combatants, one could, in principle, stop some killing.
I too have been focused mainly on the causal threshold requirement, which I find to be problematic for essentially the reasons that Jeff raises.
Doug, as to your first question about this requirement, that is, whether the requirement is intended to also imply lesser, non-lethal harms for civilian suppliers whose causal contribution and moral responsibility do not rise to the level of justifying lethal retaliation… I think Cecile’s answer, which she briefly addresses on p. 61, is that it does. (She says there, for example, that such civilians could be liable for nonlethal harms, such as higher reconstruction costs.)
About your second question, whether the “last water contaminator” meets this causal threshhold, despite being only one of several contaminators, I agree with your intuition that such a person would actually meet the threshhold. But there are of course several differences in the two cases, perhaps the most important being that, without the contaminator’s contamination, no unjustifiable harm would be forthcoming to those who need a clean river. However, harm would still be forthcoming to the victims of an unjust war even without any individual fervent voter’s vote. I’m not sure how much to make of this and other differences in the case, though.
Josh, I’m not sure I agree with your point about stopping the “means of killing.” It is true that guns kill people, and so guns are a means. But surely people can be used as a means also, right? And so I might be entitled to prevent the use of means (people) to use other means (guns) to cause me harm? Can you deny this without reverting to either the “direct participation” or the “enabling soldiers” points?
Hi Doug,
Let’s start by clarifying: on your second question, the proposed functionalist view that I had in mind is not about enabling killing, but about the means of killing.
As to why this is morally relevant, maybe it could just be basic that since killing is morally relevant, the means of killing are morally relevant as well. We might also go with a conventionalist account, where we have to draw the liability line somewhere to limit the damage of war, and the means-of-killing criterion seems like a sensible place to do that. (The moral relevance of means-of-killing allows that civilians-w and civilians-m are morally analogous in other ways, such as that by stopping them we stop some killing (as your last sentence notes). I’m just looking for some morally relevant difference between them, even if they are similar in morally relevant ways.) Relatedly, I think it would be significant if we could just come up with some criterion that actually does distinguish between civilians-m and civilians-w, since Fabre’s arguments seem to show that many existing candidates don’t successfully do that. If the functionalist view is widely endorsed as intuitive, the means-of-killing distinction might not be so implausible.
Finally, on your first question, obviously I don’t have any great way of carving out “means of killing” that includes something like tight necessary-and-sufficient conditions. But what about a know-it-when-you-see-it approach? So, if it were a war in which a widely used weapon is blankets–say, it is a war in which the chief killing technique is suffocation–this functionalist account would make blanket-deliverers uniquely liable. But if it is a war in which blankets are not (routinely?) used to kill but guns are, then gun-deliverers would be uniquely liable.
Dan,
That might be right: it looks like an implication of the functionalist view I put out there that since people are a means of killing in some cases, those who deliver them are liable in those cases. So soldiers in combat zones and those who deliver them are going to be liable on this account. But that’s probably not going to trouble any functionalists, I’m guessing. Are there other cases that they might find more unpalatable?
(By the way, before we go any further, I just want to go on the record as saying that I’m not committed to functionalism, I just want to entertain the idea that it might have an escape route here.)
Josh, I was thinking that doctors, food-providers, etc. were all “delivering” the means (people) of killing. No?
Ah, got it, Dan–I misunderstood the first time. I don’t think that doctors or food-providers provide (or deliver) people; I think that they provide health care or food. (Hence your scare quotes.) But (to anticipate Doug!) I don’t have a nice analysis of the concept PROVIDE.
It’s hard to believe that the means of killing can make a significant difference. In the law of complicity, there is no difference between driving D to V’s house where D will kill V and giving D a gun to kill V. But only the latter contributes to the means of killing.
Also, if I am trying to kill you and I can’t because my leg is trapped in a vice, it would be wrong for another person to release me from the vice, enabling me to kill you. But that person would not contribute to the means of killing.
I thought a better explanation is the one in my post above – that doctors, in curing soldiers, might contribute to the good by curing soldiers who do not kill anyone and who go on to live valuable lives. That seems a simpler explanation of the difference.
Dan,
Good points. So make the lead example one in which I along with ten others each add an additional 1% more lead to the already contaminated ground water knowing that our actions will collectively result in 20 deaths even though no one of our actions is individually sufficient to result in any deaths. What’s the extent of my contribution and responsibility? Is it two deaths? More? Less? And how do we figure out the threshold? McMahan and Fabre both talk about what does and doesn’t meet the relevant threshold, but how are we to assess these claims if we don’t know how the threshold is calculated or how the extent of one’s contribution and responsibility is calculated. To be honest, it was over a week ago that I read Fabre’s paper so I may just be misremembering. But I don’t remeber her saying anything precise about how these things are to be calculated. So I wonder why we should be so sure that in McMahan’s case the threshold isn’t met.
Josh: Fair enough.
Victor,
I didn’t mean for the means-of-killing approach to constitute a rejection of your explanation. It’s just another idea. That said, I do suspect that part of the appeal of functionalism is that it recognizes the moral relevance of weapons. Also, I don’t think that the trapped-leg case is going to do any work here, as it shows that enabling killing is wrong, which is compatible with the idea that supplying the means of killing is also independently wrong. On that front, that’s an interesting point about the laws of complicity. Unfortunately, I don’t know enough about that area, or about the relation between law and the morality of war, to say much about that (other than to reiterate that the goal was just to come up with a possible way of rendering functionalism more sensible).
Josh,
I thought that you wanted to show that civilians-m are liable to be killed where civilians-w are not because civilians-m provide the means of killing where civilians-w do not. We can at least sometimes see that a person can permissibly be killed to prevent them contributing to a wrongful killing in other ways (see Cure in the original post, as well as the vice case). What support is there that the fact that a person would otherwise contribute to the means makes a special difference to their liability?
I’m open to be persuaded that there is support for this, but perhaps we can find an alternative explanation why functionalism seems intuitively attractive. What makes functionalism seem attractive is that when we provide the means of killing our action seems more closely associated with killing. But that might simply be because the only morally salient contribution that we might make to the world by providing soldiers with weapons is by those soldiers using the weapons. In contrast, if we cure sick soldiers we contribute to their ability to do all sorts of other things beyond killing. For this reason, doctors can claim that they potentially contribute to the good as well as to the harm where munitions workers can only claim that they potentially contribute to the harm.
But when a person will otherwise do an action which we know does contribute to the harm (as in the vice case) we think that the person making the contribution is liable to be harmed to avert the threat. This seems to suggest that what matters is not whether a person’s contribution could only contribute to the threat but whether a person’s contribution will certainly contribute to the threat. Releasing the person from the vice could contribute to other things, but will in fact certainly contribute to the threat, so it is permissible to kill the person who will release me from the vice.
Dan and Doug on causal contribution: perhaps Cecile thinks that the greater the number of people whose action is necessary to create a threat the less each person is liable to be harmed to avert the threat. I don’t think that this is true though for reasons similar to Jeff and others.
Imagine a game show with 1000 contestants. Each contestant gets a £10 reward for pressing a button. If all 1000 press a button V, an innocent person, will be killed. If I know that all 1000 will press the button I can kill any of the 1000 to save V. I think that I can kill a contestant who I know is about to press the button even if I am not sure that the other 999 will press the button. Is there a significant difference between this case and the munitions worker?
In the contamination case, if killing is overdetermined it is wrong to kill one of the contributors, but that is because killing them would do no good. That is true even if there are very few people involved. Suppose that two people are each about to put sufficient poison into the water to kill V. I can stop one person by killing them but not both, so whatever I do V will be killed. It would be wrong to kill either contaminator in this case (other than as punishment).
There might also be cases where many people together contribute to the killing of one but where killing all of those people is necessary to avert the threat. Killing those who cause the threat, in that case, might be disproportionate depending on degree of culpability (it would be wrong to kill 1000 negligent people who together will cause one death but it would probably not be wrong to kill 1000 people who act in consort to kill one). Is there any way in which causal contribution makes a difference independently of proportionality considerations like this? I can’t see it right now.
Victor,
You seem to be suggesting that your defense of functionalism is superior to the means-of-killing defense of it. That may be — again, I was only putting forward an account of functionalism that seems to be plausible and evade Fabre’s criticisms.
I’ll have to think more about the merits and specifics of your view, but as a separate matter, when you ask why it makes a difference that an agent provides the means of killing, I’m not sure I can say much more than what I said to Doug above (comment at 11:58). But again I do agree that enabling people to kill might make the agent liable, as in your trapped-leg case (although that is independent of whether providing means also makes one liable).
Causal contribution is still the sticky wicket for me. It’s just unclear what the nature of causal contribution is and how precisely it’s related to moral responsibility and liability for being killed. For one thing, full moral responsibility isn’t zero-sum, so it could be attached to agents who nevertheless causally contribute to varying degrees.
More importantly, though, I think there are difficulties here with respect to omissions. Consider a variation on Jeff’s voters/robots case. Suppose the initial plebescite approving an unjust war passes because 1000 agents simply omitted to vote against it (and they were ardently against the war and would have voted had they not negligently forgotten to). I think they’re morally responsible for what they’ve omitted to do, but it’s difficult to say that they provided a causal contribution to the unjust war simply because they didn’t cause anything.
It’s less clear that they’d be liable to be killed, however, so consider a variation in which one of the weekly votes fails, but only because 1000 ardent supporters of the war failed to vote for it. Now it seems that those 1000 could well be liable to being killed by the potential victims of the country’s aggression, simply to prevent them from voting in the next round. But here their liability isn’t a matter of their causal contribution.
Now this may still not be a compelling counterexample, given that one could say their liability isn’t because of their omission: it’s because of their potential causal contribution. But then suppose there are roving military psychologists who could prevent various crazy marines from continuing to commit wartime atrocities by discharging them, but they don’t even bother to do any field work to identify these crazies. Seems to me they’re liable to being killed, despite their lack of any causal contribution, and precisely because of their omission.
David
In your military psychologists case we need to show that the psychologists would be liable to be harmed as a means to avert the threat created by the crazy marines (or other threats that they are not involved in creating or failing to prevent). Killing them would otherwise be useless in averting any threats. That might be true, though, if the psychologists are sufficiently culpable. They might be sufficiently culpable because of their special responsibility to avert the threat created by the crazy marines.
But must we say that they do not cause the threats of the crazy marines? We might think that omissions can cause – for example if I starve my baby to death by failing to feed her do I not cause her death?
At any rate, I’m not sure that Cecile is committed to saying that causal contribution for a threat is a necessary condition of being harmed to avert the threat. If a person has special responsibility to avert a threat and fails to do so, she might endorse the view that the person is liable to be harmed as a means to avert the threat. That would leave her view about civilians-m untouched though. IF she is right (which I doubt) that they have insufficient causal contribution for the creation of a threat they would also not have special responsibility to avert the threat.
“ Much ado about nothing” posted by Pintoo Shome
It is a nice and mutually satisfying intellectual discussion for ethicists to debate on whether killing civilians is right or wrong based on whether they are directly involved in supplying arms (Cm) or are indirectly involved in helping the military with other provisions or support (Cw) or their involvement lies somewhere in between these two.
First, I tender my apologies for making such a strong remark. Next, I agree with the author that we can not make a distinction. My view is that it is unethical to even start making a distinction between the two types of civilians on whatsoever basis. In today’s world, more often than not, the full circle and trajectory of arms supply travel back to the attacking nation in one form or the other. The attacking country, at a point of time in history, may have directly or at least indirectly contributed to the arms supply or in its development to the nation it is now attacking in a ‘just war’. So, extending the logic a little further one should attack oneself first before even starting to make a distinction between enemy country civilians. In most cases the attacking nation may find, ‘I am the arms supplier, I am my own enemy’, should I attack myself first and how exactly do I do it? The answer to this of course is simple. Follow the guidelines of the normative theories of ethics in a comprehensive fashion. The interesting thing about the normative theories of ethics is that in isolation each may be somewhat indefensible under certain circumstances but together they stand formidable and serve as an authoritative guide for deciding what is ethical and what is not. The key lies in systematically running through all of them using a simple logical structure and putting these jewels (normative theories) on our crown, and together.
Hi, Victor.
First, yes, I was thinking of the psychologists as being culpable given the special role they have in the military. (Incidentally, the very same reasoning could apply to chaplains who fail to “counsel down” the crazies, say.) And I don’t think it’s appropriate in their case to say that they “cause” the threats of the crazy marines; rather, the marines cause those threats. The psychologists simply fail to do what they could to identify and stop those threats.
This last point also might serve to distinguish this sort of case from the baby case you mention, given that in that case I’m in a special relation with a specific baby. I’m more inclined to say in that case that I’ve caused the baby’s death by omitting to feed her, but because of the lack of such a specific relation in the psychologist case, it’s much harder for me to see it as case of causation at all. After all, the psychologists don’t have a special responsibility to any specific marine.
So how does this relate to Fabre’s article? Her thought is that “a contribution must, on its own individual terms, meet a threshold of causal significance in order for its author to be liable” (61). Causal contribution thus sounds pretty necessary to the account to me (at least given her qualified allegiance to the “traditional account,” what McMahan calls the “neoclassical account”). What I suppose I’m wondering, though, is if one might avoid the worries about causal contribution noted here by trying to adopt a straightforward “moral responsibility” account, one which could incorporate both causal contributions as well as omissions. There remains a major difficulty in doing so, of course: how to articulate a threshold in terms of moral responsibility? On that question, I have nothing to say. Indeed, I wonder whether talk of thresholds is worth preserving at all in that case.
(On a complete side note, Victor, I’m going to be teaching your book Criminal Responsibility this semester, so I’m glad to have made this connection with you.)
Hello David
I think I agree with your intuitions about the difference between the baby case and the marine case. Causal language seems to get a bit messy here (why should the specificity of the relationship matter?)
A moral responsibility account is plausible. Whether the idea that liability to avert a threat is dependent on moral responsibility for the threat is attractive depends on the view that one takes on innocent threats. I tend to think that innocent threats can be harmed, and so a person can be harmed to avert a threat that they are not responsible for creating. So I tend to think that the causal role that one’s body has in the existence of a threat can ground liability to be harmed to avert the threat. But that is admittedly controversial (McMahan and Fabre reject the idea).
Even on a moral responsibility account, causal contribution (were we able to make sense of it) could be a component of moral responsibility (it might be argued that a person is not morally responsible for a threat if their causal contribution is too modest). But we might just think that causation is all or nothing and that it is impossible to explicate a coherent idea of a causal contribution that is too minimal. That seems a concern that we all share.
On McMahan’s moral responsibility account of liability if a person acts in a way that the risk of threat creation is negligible, the person is not responsible for the threat if it is realised, and she cannot be harmed to avert it (for example, if my phone is wired up to a bomb, McMahan thinks that it would be wrong to kill me to prevent the bomb going off to save one other person on the grounds that making a phone call is negligibly risky). So with respect to risk level, McMahan sets a low threshold for responsibility. At least we can coherently talk about a threshold in that context. But of course, that view would not help to vindicate Fabre’s claim that civilians-m and civilians-w are not liable to be killed in self-defence.
As I think that we can kill innocent threats I think that it is permissible to kill the person with the phone to stop the bomb going off. So I am even less inclined to think that it is wrong to kill civilians-m. For the reasons outlined in my first post, though, I think that civilians-w might be another matter.
Sorry that you’re having to put up with the book! Let me know any thoughts you have about it, however abusive!
This is really interesting, Victor. I think I might actually share your thought that moral responsibility isn’t what matters here, given some more consideration about the killing of innocents. Borrowing from Jane English’s old article on abortion, it would seem permissible to kill innocent hypnotized attackers if that was the only way to stop the threat, so why wouldn’t it be permissible to kill innocent civilians who aren’t morally responsible for their contributions to the unjust war effort? I haven’t thought very much about these issues before, but this strikes me as initially plausible.
Just a small point on Victor’s last comment regarding liability and innocent threats. Victor you suggest that whether some version of the moral responsibility account of liability is attractive/plausible depends on one’s view about the permissibility of harming innocent threats. The implication is that if (as you do) one thinks it’s permissible to harm innocent threats then this provides a reason to reject the responsibility approach to liability. But this isn’t necessarily the case. One might believe that innocent threats are not liable to be killed, that is, they retain their right not to be killed, but that it may nevertheless be permissible to kill them because of the mode of agency involved – i.e. because the killing is merely eliminative and not opportunistic, to use Warren Quinn’s distinction. I know you and I agree on the importance of Quinn’s eliminative/opportunistic distinction, but I just wanted to point out that we should not confuse liability with permissibility here. I think it’s permissible to kill innocent threats, but it’s clear these innocent threats are not liable: they haven’t done anything to waive or forfeit their rights, and so if we transgress those rights, I would follow Thomson and say we have infringed (all things considered permissible wrong), but not violated (impermissible wrong) their rights.
Thanks to Cecile’s for a terrific paper and to Jeff for involving me in this exchange. Here are 3 objections to the way Cecile deals with the pure functionalist account and the moralized functionalist view.
1) The PURE FUNCTIONALIST ACCOUNT (PFA). According to this view the reason why civilians-M are liable to attack, while civilians-W are not, is that the former’s contribution counts as ‘direct participation’ in the unjust war. The key distinction here is between “those who make what combatants need to fight and those who make what they need to live” (Walzer, Just and Unjust Wars, p. 146). Since Civilians-W give unjust combatants what they need to function not as combatants, but as human beings they are immune from attack. In rejecting this view, Cecile (correctly in my opinion) objects that this is an inadequate understanding of what counts as “direct participation”. Given that civilians-W “make something which is designed to enable combatants to fight, and which is made so that they can fight” (p. 42), they provide combatants with what they need to fight no less than civilians-M. Thus the functionalist distinction doesn’t work.
Notice however that this objection does not target the main intuition of the PFA, namely that (at least in principle) only those who “directly participate” in an unjust war are liable to attack. Cecile seems to accept the distinction between those needs that combatants have qua human beings and those needs that they have qua combatants (i.e. those needs that arise out of their being engaged in the war). What is challenged is the way in which PFA understands the notion of “direct participation”. Cecile argues that some Civilians-W can be said to be “direct participants” because, like civilians-M, they provide combatants with what they need to fight. The point is that civilians-W typically don’t provide combatants with regular food or with regular medical treatment, but with specialized food (prepared and packed to be consumed in war conditions) and specialized medical treatment (which strengthens their physical condition and enables them to quickly get back to the battle). For this reason, civilians-W are making a “contribution which is clearly directed to the war” (p. 45).
I think that Cecile is right in claiming that those civilians-W who provide combatants with specialized food and specialized medical treatments designed to enhance their capacity to fight should be considered as “direct participant” to the war. However, we should distinguish between two kinds of needs that combatant have because of their being engaged in lethal fighting:
A) Those needs that combatants have in virtue of the fact that their aim is to keep fighting at their best. Meeting these needs serves the end of enhancing combatants’ capacity to survive the fight and kill their enemies in more effective ways.
B) Those needs that combatants have because they have been engaged in lethal fighting, but that do not aim to enhance their capacity to kill in a more effective way. Take for example a soldier who needs to have his broken leg fixed. Even if he broke his leg while fighting, fixing his leg does not serve the end of enhancing his capacity to fight.
The former are needs that only combatants have, as ultimately they have to do with enhancing combatants’ capacity to kill. The latter are needs that every human being has (every human being needs to have a broken leg fixed). Meeting these needs does not serve the end of potentiating combatants’ capacity to fight.
Does Cecile claim that only those civilians-W who provide resources that meet the first kind of needs (i.e. resources enhancing the capacity to kill) count as direct participants, or is it part of her view that those who provide resources that meet the second kind of needs count as direct participants as well? She seems to go for the former view. For example, she claims that a doctor who treats a soldier who will be discharged does not count as a direct participant (p. 45). But what if a doctor can provide some non-specialized treatment to a soldier who will get back to the battlefield? Does this doctor count as a direct participant? Typically manufacturers of specialized food and clothing do not also work to provide non-specialized food and clothing, but the case of doctors is interesting in my opinion because the same doctor usually will have to provide both. Sometimes she will be in a position to provide specialized medical treatment. Sometimes she will be only able to fix a broken leg or heal the fever. Would Cecile agree that when providing this latter kind of treatment doctors do not count as direct participants, even if soldiers cured by them will get back to the battlefield? If so, the gist of the functionalist position is not affected by Cecile’s criticism. The functionalist position still holds; we only need to replace the distinction between civilians-M and civilians-W with the distinction between:
– civilians-M and civilians-W providing specialized resources,
– civilians-W providing non-specialized resources.
This of course has important practical implications as to the question of civilians’ liability to attack. Cecile concludes the section devoted to the PFA saying that “if direct participation is a sufficient condition for liability to attack, then both civilians-W and civilians-M are liable”. If I am right, this is too strong. At least in the case of those civilians-W who provide medical treatment to combatants, before we can establish that they are liable we need to consider their specific contribution on a case by case basis.
2) This conclusion plays an important role in relation to the second target of Cecile’s paper: the moralized functionalist view. According to this view, citizens of any state, including those who are fighting an unjust war, are owed basic material resources by their states. Civilians-W who provide them with these resources might well be directly contributing to infringing the enemy’s right not to be killed, but they have a justification for doing this; i.e. they are helping their state discharge its duty of justice towards its combatants. For this reason civilians-W are not liable to attack.
Of course the crucial question here is whether civilians-W are actually justified in providing unjust combatants with these resources. Cecile considers (and rejects) three possible justifications. Here I will focus only on one of them, namely the view that grounds such a justification on patriotic partiality – i.e. on special relationships existing between civilians-W and their combatants. This argument according to Cecile proves too much, for it would justify not only civilians-W in providing unjust combatants with material resources, but also unjust combatants in killing just enemy combatants in defence of their comrades. If patriotic partiality does justify the act of enabling an unjust threat why doesn’t it also justify the act of directly posing an unjust threat?
Here the difference sketched above between those needs that combatants have qua human beings and those needs that they have qua combatants becomes relevant. For if it makes sense to distinguish between welfare resources that meet the former and welfare resources that meet the latter, the patriotic partiality argument might work better than Cecile suggests. It could be plausibly argued that special responsibilities to compatriots are weighty enough to ground an overall duty to provide them with non specialized material resources that they need qua human beings (even when they are fighting an unjust war), but not weighty enough to provide them with specialized resources that will help them to fight an unjust war. Suppose that someone who is in a special relationship to me (say my sister) is unjustly attacking someone. It seems to me that my special relationship does ground a duty to cure my sister if she gets injured in the fight, though it does not ground a duty to attack the other person myself, or to provide my sister with any weapons that might help her in attacking the other person. Simply my special relationship is weighty enough to ground a duty to rescue my sister, in spite of the fact that by keeping her alive I am indirectly contributing to an unjust threat (I discuss this problem below), but not weighty enough to ground a duty to directly attack someone else. Of course I also have a duty to try and stop my sister, but letting her die is not one of the means available to this end.
Similarly, if we grant (as Cecile does) the existence of patriotic and role-based obligations, we should conclude that doctors often have a duty to save their compatriots if they can (in spite of the fact that this is a way of contributing to the unjust threat that they are posing), though they also have:
– a duty to boycott the war in other ways;
– a duty not to enhance their capacity to fight by providing specialized material resources,
– a duty not to kill just combatants, even in this is the only way to save their compatriots
Once defended on these grounds, the view that there is a moral difference between the act of providing material resources which are necessary to survive and the act of providing material resources that enable to kill seems to me plausible.
3) This point is important because the main objection that Cecile moves to the moralized functionalist view is precisely that the various obligations of civilians-W need to be balanced with each other. For example, doctors’ role obligations to provide medical treatment and their special obligations to compatriots must be balanced with their general obligation not to wrongfully harm (or contributing to harm) others. And the latter, Cecile argues, is weightier because wrongfully failing to help is morally preferable to harming wrongfully and to significantly contributing to the imposition of a wrongful harm.
But even if we agree that failing to help is morally preferable to harming wrongfully, it is by no means obvious that it is also morally preferable to contributing to the imposition of a wrongful harm. Whether this is the case will presumably depend on the significance of the contribution. It might be argued that a small, marginal contribution to the imposition of a wrongful harm is sometimes preferable to failing to help. If a doctor doesn’t save a combatant dying in front of her she bears full responsibility for letting him die. If a doctor provides regular medical treatment to the combatant and the combatant will get back to fight an unjust war, the doctor might be said to have made a contribution to the unjust deaths caused by the latter, but her responsibility for these deaths will be limited. The responsibility for these deaths will fall primarily on the combatant, on those who declared the war etc., and only in minimal part on the doctor who provided him with regular medical treatment. Once we compare the doctor’s full responsibility for letting the combatant die with her limited responsibility for indirectly contributing to the unjust threat it is not obvious that the general obligation will always outweigh the role-based one.
Notice also that whether this is the case will presumably depend on the kind of contribution that the combatant to be treated is likely to make to the unjust war. For the contribution that the doctor will make to the unjust threat ultimately depends on the contribution that her patient will make to the unjust threat. If the patient is a general that will cause the death of thousands of enemies, the doctor’s general duty might outweigh her role-based duty. If the patient is a soldier who is likely to make a small contribution to the unjust threat, the role-based obligation of the doctor is likely to be weightier.
Curiously enough Cecile acknowledges the importance of this element in the last part of her paper, when she argues that only those civilians who make a significant contribution to the unjust war can be deemed liable, while those who make a marginal contribution cannot (in spite of their acting unjustly). My point here is that the significance of the contribution made by civilians ultimately matters not only to establish whether they are liable, but also to establish what their overall duty is. Doctors might have an overall duty to provide (regular) medical care to those unjust combatant who make a small, marginal contribution to the unjust threat.
In other words, if Cecile is right that “the mere fact of providing medical care to combatants does not provide the doctor with a moral justification for so acting irrespective of what those combatants will thereby be enabled to do” (p. 55), it is also true that the mere fact of indirectly contributing to an unjust threat does not provide the doctor with a moral justification for not providing medical care to unjust combatants irrespective of what those combatant will be likely to do. Once we pay attention to the distinction between specialized medical treatment and regular medical treatment, as well as to the nature of the indirect contribution made by doctors and other civilians to the unjust threat posed by combatants, both patriotic and role-based obligations seem to be stronger that Cecile suggests.
Jon
It’s a tempting proposal but I’m not sure.
First, I think that it is sometimes wrong to harm a person to avert an objectively threat which they are morally responsible for creating. For example, suppose that I attempt to rescue you at great risk to myself. In doing so, through no fault of my own, I create an objectively unjustified threat to you. It might be wrong, then, for you to impose on me an equivalent harm to avert the threat.
Secondly, the reason that is most often given to distinguish between the liable and the non-liable, as well as between those whose rights have been infringed and those whose rights have not been infringed is the obligation to compensate the person harmed. But I suspect that the justice of compensation will not always track the distinction between those who are morally responsible for the creation of a threat and those who are not. For example, in some cases like that in the previous para, if I attempt to rescue you at some risk to myself but happen to do create an objectively unjustified threat you might sometimes be permitted to harm me, but you might also owe me compensation for having harmed me.
Finally, in law we do sometimes talk about liability in cases of where there is no moral responsibility, for example if a person is (very) strictly liable. So the language of liability itself would not necessarily be out of place in cases of non-responsible threat creation.
Sometimes I’m tempted by the following thought. Wouldn’t it be simpler just to determine a) who can permissibly be harmed; and b) who is owed compensation for being harmed. We could then drop the language of liability as well as the infringed/violated distinction with all of the extra complications that language brings.
Hi Victor,
Thanks for the reply. On your first point – it might be true that McMahan’s version of the moral responsibility account of liability won’t yield the result you want in this case, but that doesn’t mean that some suitably modified version of the responsibility approach couldn’t do so, i.e. one where we don’t hold people responsible for faultless harms arising from supererogatory behaviour.
Second, although you are right that one reason that’s given to care about distinguishing liable from non-liable is the compensation issue, that’s not the only reason to care about liability. Another important reason to care about liability is this: it seems relevant to determining who is permitted to ‘fight back’. If X is liable to be killed in self-defense by Y, then we might plausibly suppose that means that X cannot permissibly fight back when Y attacks. This is because Y is not wronging X when Y acts in self-defense, and X has no justified complaint against Y. If, however, X is *not* liable to be killed by Y, but Y (let’s say permissibly) acts in self-defense against X, then it makes sense to think that X should be permitted to fight back, that is, that this should be a case of symmetrical self-defense, where each party has a permission to act in self-defense.
It also might be true (I think it is true) that an innocent victim who might act in self-defense is morally required to bear a greater burden to avoid harming a non-liable threat as opposed to a liable threat (e.g. I might be morally required to suffer five minutes of severe pain to avoid harming a non-liable threat, but I might not be required to suffer an equivalent harm to avoid harming a fully liable threat).
In short, we should not treat non-responsible threats in exactly the same way as responsible or culpable threats, and I think this indicates that they are differentially liable. So, I think an independent account of liability is important partly because it’s necessary to determine what is permissible in various self-defense scenarios.
C. Fabre. Reply to Peasoup comments made between Jan 11 and Jan 13 2010-01-14
First of all, let me thank the editors of Peasoup for selecting the article: it is an honour for me to help start what promises to be a fruitful and stimulating collaborative enterprise between the blog and Ethics. I am also deeply grateful to Jeff, and to those of you who have posted a blog entry on the paper here, for such constructive and thought-provoking comments. The latter, taken together, currently run at 20 pages of single-spaced text (!), which is above and beyond what I could have expected. In this and the next two or three posts, I make some points about Jeff’s critical introduction and comments made by others which directly touch on the issues which Jeff raises. My points divide into responses to specific comments, and programmatic thoughts as to where more work is required. I will move on to other issues (esp. discussions of functionalist arguments) in a separate entry later today or tomorrow. I will try to address all the issues raised in the posts so far, but apologise in advance if I miss some of your points.
Finally, the posts will be rather long, so I refer to respondents by name specifically throughout, in case you are interested in responses I have to your own specific points and want to do a name-search.
1. Specific points on Jeff’s reply
1.1. Clearing up some misunderstanding
Jeff brings out, more clearly and sharply than I did, the implications of my argument for the relative merits, and demerit, of the orthodox and neo-classical accounts of liability to intentional attack (incidentally I agree with his observation that the label ‘neo-classical account’ is less confusing than my own ‘traditional account’, so from now that is the label I will use.) To wit: that if my argument is correct, then civilians enjoy greater protection at the bar of the latter than of the former. The crucial question, thus, is whether the argument is correct – and in particular, whether the threshold conception of liability which I defend in section V can bear the weight of the argument.
Jeff is sceptical, to say the least – and I myself am acutely aware of the fact that this is the most problematic part of the paper (though Massimo, whose comments I will attempt to tackle presently, has interesting doubts about my criticisms of the functionalist arguments.) But let me first clear up a misunderstanding pertaining causal contributions, and make a point about moral responsibility.
In the paper, I make the following claim: ‘more often than not, welfare infrastructures and their agents will, in wartime, help both just and unjust combatants. … The same can be said, mutatis mutandis, of munitions factory workers. (58-59) Jeff confesses to finding this puzzling, on the following grounds: ‘There are undoubtedly cases in which weapons manufacturers supply weapons to both sides in a war. In those cases, the acts of munitions workers may contribute equally to the action of just and unjust combatants. But I suspect that this is true in only a small minority of cases. In World War II, for example, military production in the US was wholly separate from military production in Germany. And I don’t know of any cases in which the suppliers of food or medical care to one army have also been the suppliers to the opposing army as well.’
Here is the misunderstanding. My point was not that civilian workers provide both belligerents with military and/or welfare resources. Rather, my point was that a war which is unjust in toto is likely to have just phases – a point which Jeff himself makes in some of his writings, and which I stress on p. 57. Workers who work for the in-toto unjust side might thus end up contributing to just missions/phases of war within that in-toto-unjust war. I raise that issue for two reasons. First, Jeff’s misunderstanding prompted me to think about further examples of the point he attributed to me. Here is an interesting case, that of suppliers who work in a country who is occupied by an unjust enemy, and where a section of the population mounts military actions against the occupants (During WWII many French farmers supplied both German soldiers and French Resistants (the latter, from 1942 onwards, were organised into an army of a kind, with a chain of command, etc.) Much more work needs to be done, I think, on the ethics of war under the circumstances of a military occupation. It is a relatively overlooked issue, though one which, as a French and thus for very obvious reasons, has exercised me for some time. Second, proponents of the neo-classical account (of whom I am) must do more work on the implications for the account of the multifaceted character of war. As some of them have noted (Jeff himself, myself in this paper, Rodin too), it is not appropriate to speak of just and unjust combatants by referring solely to the moral status of the war ad bellum. In addition, more work needs to be done, I think, on the in-toto/in parte distinction (when does a war which has a just cause but unjust phases become unjust in toto, etc.?)
A further thought on Pinto Shoome’s point to the effect that the defending nation is likely to have supplied the attacking nation with weapons in the past: true enough. But I have in mind civilians’ continuing, ongoing contributions to a war as the war unfolds – in which case it is unlikely that members of belligerent A would supply arms to belligerent B (though they might do so, of course, in which case their situation would be relevantly analogous to civilians from B who supply weapons to B. In attacking (or more likely prosecuting, one would hope) those individuals A, belligerent A would not necessarily be attacking itself (as opposed to some of its members.)
1.1 (Ced)
So much, for now, on causal responsibility. With regard to moral responsibility, Jeff argues that unjust combatants and unjust non-combatants can hardly ever plead (ex post) unavoidable ignorance of the moral status of the threat to which they are contributing. At the very least, he writes, they ‘can surely know that there’s a significant risk that the threat they pose is unjust, and that alone seems a sufficient basis for the attribution to them of some responsibility for the wrongful harms they cause.’ Now, I try to address this objection on p. 59, and offer a reply at the top of p. 60, as pertains to the case of civilians. There, I do not deny that the latter bear some degree of responsibility and culpability for the unjust threats which they enable (though reading that para again I can see why one might think that I do deny that.) But the point at issue at this juncture of the paper is the different one of whether they are liable to attack (notwithstanding their responsibility.) Having said that, readers will have noticed of course that I simply assert, and do not provide an argument for, the claim that civilians are not liable even if they are aware of running a risk of contributing to unjust threats. (See the (long) sentence at the conclusion of the first para on p. 60.) That’s because I have no considered argument for it – just an unshakable intuition. And I think that it would have been entirely fair of Jeff to point out, here, that an argument is needed. Accordingly: any suggestions more than welcome!
Some technical difficulties seem to have cropped up. The rest of Cecile’s replies follow:
1.2. On the liability threshold
Setting those various points aside, however, the main issue (at least to me) which arises from Jeff’s comments is that of the liability threshold. Here, we must distinguish between two questions: (a) does it make sense to endorse a threshold conception of liability, whereby whether individuals are liable to lethal harm depends on the location (as it were), in relation to a liability threshold, of their causal and moral responsibility for an unjust lethal threat? (b) Assuming that it is appropriate so to conceive of liability, were are civilians’ contributions located in relation to that threshold?
With respect to (a), as I note in the paper, I find it difficult to see how to avoid resorting to the notion of a threshold of causal and moral responsibility – at least, from the standpoint of a morality which posits that the harms which individuals are liable to suffer must be a function of what they do, and are morally responsible for doing, as individuals. Or, to put it in Doug’s words (Jan 11, 10:29am), ‘the severity of the sorts of harms to which civilian suppliers are liable would depend on the degree of causal contribution and moral responsibility they have for the military threats that they make possible.’ This, I think, captures the point well.
In the paper, I try and show that Jeff’s objections to threshold-liability, via the example of the conscientious driver, fails. Jeff’s claim is that on a threshold view, a conscientious and careful driver who is about to hit a pedestrian as a result of factors beyond her control is not liable to being killed (for surely, he writes, she is below the moral responsibility threshold). Yet, there are good reasons to think that the pedestrian may kill her, from which it follows that we should reject the threshold argument. My reply takes the following form, to wit, that at the bar of partiality-based justifications for permissible killing (whereby agents are (sometimes) permitted to confer greater weight on their own life, goals, attachments, even at the cost of a Threat’s life), the pedestrian is permitted to kill the conscientious driver, though the latter is not liable. In other words, there is a way to justify such killing (in line with pre-theoretical intuitions in that case) which does not appeal to liability. Jeff is not persuaded: on his view, given that the driver is merely permitted to drive and thereby impose some risk on pedestrian (and lacks a positive justification for so doing: she could after all walk, instead of driving, as opposed to an ambulance driver), she is not exempt from liability. But I fail to see why. Rather, this is so only if there is no plausible justification for overriding the prohibition against intentional killing, other than liability or lesser evil. I believe that partiality does provide us with the best hope of just such a justification, at least in that case (though I admit that more work needs to be done on the scope and limits of partiality arguments for self defensive killings. Quong’s recent Ethics piece addresses it.)
With respect to (b), Jeff has ‘serious doubts about her claim that there’s a threshold for liability that the contributions of individual civilian suppliers rarely, if ever, exceed.’ In particular he asks us to think about the case of citizens who in week 1 vote (by a small margin) to start an unjust war (the war is fought by robots who are activated by the voting machine, depending on the outcome of the vote.) Those citizens have the option to decide every week whether or not to continue with the war. According to Jeff, if it were possible for the enemy to identify 1000 staunch supporters of the war (in the twofold sense that they advocate it and vote for it), and if killing them would reverse the outcome of the vote the following week, then killing them would be permissible. Yet, ‘their causal contribution and degree of moral responsibility are surely below Fabre’s threshold, since they are below those of typical civilian suppliers of the resources that unjust combatants need to fight.’ It’s worth stressing, moreover, that what’s true of the 1000 voters who are selected, perhaps randomly, for targeting is equally true of all the other citizens, of whom there may be many millions, who initially voted in favor of the war and would do so again.’
This is a fiendishly tricky problem, so let me simply offer a few thoughts. First, I take it that the reason why killing them would reverse the outcome of the subsequent vote is that, without those 1000 votes, the robots will not be activated (as opposed to it being the case that, as a result of their deaths, fewer citizens decide to vote in favour of the war than is required for a majority decision.) But if that is the case, then it is not clear to me that their causal contribution to the war falls far below that of a munitions factory worker who screws a rifle together. If killing those 1000 citizens suffices to stop the war (for the reason just indicated), then it must be because their contribution is causally highly significant. The contribution of an individual factory worker is usually not, hence no liability to lethal harm (although, as Dan Boisvert notes in his posting of Jan 11, I also say that they might be liable to lesser harm. (Contrastingly, the contribution of an individual engineer who designs a super-dirty bomb is significant, and so he might well be liable.) Here, I agree with Doug Portmore’s first posting (Jan 11 at 10:29am).
Second, there is an important difference between the citizens and the civilians (and here, we have to return to the earlier point about the multifaceted character of the war.) Jeff seems to suppose that those citizens vote for a war which is unjust in toto. And it is that, I think, which might provide support for the intuition that they are liable (together with the significance of their contribution.) As I noted above, however, typical civilian workers in typical wars contribute to both just and unjust phases of a war, which I think goes some way to explain why they are not liable (for it is not clear that their contributions are contribution to unjust acts of killing). To be sure, one could imagine a case where citizens decide to continue to vote for a war some phases of which are just, whilst others unjust (suppose, for example, that the enemy unjustly targets retirement homes and nursery schools in prosecution of the war, and that the robots can block such threats by killing the jet fighters.) In such cases, then it would seem that they are not liable.
1.3. More on the liability threshold
On what I think is the toughest issue of all), Doug writes that I do not provide a way of calculating the threshold (Jan 11, 01:41am). He is absolutely right: I don’t. This is largely because I do not see how one can arrive at a general, fixed, principle or rule which would tell us, in any given case, whether an individual is liable or not. Rather (and I concede that this is deeply unsatisfactory, though probably the best we can do), we might have to decide on a case by case basis, bearing in mind both causal and moral responsibility taken separately, and their relative weight. I do not really know how this might work. But I think that further work on this would have to address the following questions: (a) could there still be rough guidelines as to what counts as a causally significant contribution to a collective enterprise? There is a rich and growing literature on individual responsibility for, and liability to harm as a result of, collective wrongdoing.
(b) Suppose that killing one citizen C would not suffice to stop the war, whereas killing 1000, including C, would. Is C liable just in virtue of the fact (together with his moral responsibility, etc.) that he belongs to a group of individuals whose collective contribution is significant? Or is his liability conditional on the significance of his contribution taken singly? This avenue is one which war ethicists in general (and not merely proponents of threshold conceptions of liability) must address. Victor writes (11/1, 2:54am): ‘Perhaps Cecile thinks that the greater the number of people whose action is necessary to create a threat, the less each person is liable to being harmed to avert the threat.’ I do not think that (but, further, I do not think that I am committing myself to this.) For if the contribution of each and every one of 100,000 people is necessary to create an (unjust) lethal threat, there is no difference really between the moral status of each of those individual actions on the one hand, and the moral status of a lone murderer. My point though is that the contributions of the overwhelming majority of civilians to a war, taken singly, do not have that degree of significance (the war would go on if any one of those civilians, taken singly, stopped working.) The question is whether that is enough to absolve them of liability. (Incidentally I agree with Victor (same posting) that killing 1000 (non culpably) negligent threats would seem disproportionate.)
(c)Is it a decisive objection to the threshold conception that there is no such fixed general principle?
(d) What plausible alternatives are there to the threshold conception?
Fourth, on David Shoemaker’s point re omissions (11/1, 5:23), and ensuing exchange with Victor. Very interesting issue. David imagines a case where 1000 opponents of the war negligently fail to vote against it, as a result of which the war goes ahead. He does not think that they are liable. Now, if one wants to allow for the possibility that allowing harm to happen might make one liable, then the label ‘causal responsibility’ might indeed be problematic (one might to go for something like ‘one’s role in the chain of events which lead to the war,’ etc.) The interesting issue of course is whether posing or contributing to a wrongful threat of lethal harm, where the contribution takes the form of an action, is a necessary condition for liability. Here is one context where settling the issue is crucial: if it is a necessary condition, then it would seem that failure on the part of the affluent to fulfil their minimal positive obligations of assistance towards the very poor could not count as a just cause for war (I have written a bit on this elsewhere and will not pursue this here. I assume of course that the affluent do have such obligations).
Let us assume negligence is not a bar, in principle, to liability. Now, if liability has an inbuilt instrumentality requirement (which I think it must do), whereby an agent P is liable only if killing him would achieve the relevant end, it would seem that one cannot kill agents merely for omitting to do x: killing them now clearly will not make them vote next week. Note however that there are two ways of construing the requirement: (1) P is liable only if (together with other conditions such as moral responsibility, etc) killing him will avert the threat which he poses or contribute to causing to V (in which case, for reasons just adduced, war opponents who omit to vote are not liable); or (2) P is liable only if killing him will serve to avert the threat to which V is subject. Construal (2) is compatible with holding those anti-war citizens liable to lethal harm, in the following case: suppose that the enemy kill 1000 anti-war citizens in week1, and makes it publicly known that they are killing them precisely for neglecting to vote, and does so on the reliable information that this will get other anti-war citizens to go and vote against the war in week 2, with the effect that the robots will not be activated. The instrumentality requirement as per (2) is satisfied (which of course does not suffice to show that those citizens are liable: it could still be that killing them is a disproportionate response their contribution-by-omission – either because the omission is seriously wrongful, or because omissions can never be a basis for liability even if the instrumentality requirement is satisfied.
On the very last point: intuitively I should think that omission can sometimes be a basis for liability (so I would have to revisit the various statements where I talk of causal significance as being necessary.). Ex: I can only move my hands (which rests on the trigger of my gun) and am starving. Two agents are each in a position to throw me food and a mobile phone with which I can call for help. Suppose that they are under a moral duty so to act, but refuse (say I am black, or gay, or whatever.) If killing one would get the other to help me, may I not do so? (Victor’s posting of Jan 12, 12:08am, is apposite here,.) Note, incidentally, that I make no mention of special responsibility for help here (See David and Victor’s exchange on this). Generally, it seems to me that a morally laden understanding of omission and causation (which would rest on the absence, or presence, of special responsibility, is problematic (as it would be hard to avoid begging the question). And more generally still, pace Victor’s point (12/1, 8:32am), I am not persuaded by the view that a person is not morally responsible for a threat if their causal contribution is too modest. On that view, none of us is morally responsible for global warming, which seems odd (granting certain relatively uncontroversial facts about global warming.)
1.4. Liability and permissibility.
David and Victor makes some points about the permissibility of killing morally innocent attackers and the notion of liability to attack in war. But note that the view that only those who are morally responsible for an unjust lethal threat are liable to being killed, is compatible with the claim that the morally innocent may be killed (on a narrow understanding of liability which I define (first preliminary rk of the introduction) as having lost one’s right not to be killed.) As I say there, there might be cases where someone is not liable (has not lost his right) but nevertheless may be killed (in which case one will talk of a permissible right infringement.) Accordingly, my claim that the overwhelming majority of civilians are not liable is compatible with the view (which I neither defend nor deny ) that killing them is (in some cases) morally permissible. In other words, I agree with Jon Quong Jan 12/1, 11:44am).
Victor has a reply (Jan 12, 02:51am): if you rescue me at great cost to yourself and in so doing pose a further, unjustified threat to me, then (V says) I may not harm you. Agreed: to cater for this would require refining the liability account and distinguish between cases where the victim of the unjustified threat has not consented to the occurrence of the event which leads to the threat, from cases where they have; we should also bring in factors such as the magnitude of the harms incurred by the rescuers, etc. Crucially important of course for the ethics of wars humanitarian intervention. In any event, these would be refinements of the account, and would not undermine it (again, I agree with Jon’s second post on this.) I also agree with Jon that judgements about who can and cannot fight back are crucial to the liability account. The very few points I make in the paper about cases of symmetrical self-defence speak to this issue.
Victor suggests an important difference between civiliansm and civiliansw: if the soldiers whom the latter help do not go on killing, then helping them is permissible (even mandatory.) Not so with providing weapons (it is always bad to provide weapons to the unjust side even if those soldiers do not use them.) This is an interesting point, which I only touch upon in the paper. Unlike Victor, though, I believe that it is wrong to cure someone Y who, we know, has the intention and the means to kill V – unless we also know for sure that he will fail. The tricky issue is that of deciding whether to help in the case where we do not know for sure whether Y will succeed. There, it seems to me, we must decide whose life should be given greater weight: Y, or the Victim whom he has a x/100 chance of killing. Again, this is one case where we may not be able to get to a general, fixed principle. But what we must never forget is the fact that V’s life is in the balance.
Incidentally, I also believe that the case of weapons is closer to the case of medical treatment than Victor suggests: this is because soldiers on the unjust side are quite often threatened unjustly by soldiers on the just side; providing them with weapons, in that case, might seem prima facie permissible. Finally, I have no stand yet on whether excuses affect permissibility (whether, that is, the permissibility of killing Y in part depends on the degree to which Y has an excuse (as opposed to a justification) for killing. But note that my focus in the paper is on liability, not excuses (see end of my previous long posting for more on this.)
On the discussion between Josh, Doug Dan and Victor re ‘providing the means of fighting’
I basically agree with Doug (11/1, 11:38) that, even if guns kill, not food, food is nevertheless a means of killing. Historians of WWI thus often say that severe and persistent hunger adversely affected the German army’s ability to fight in 1918. But Dan raises yet another interesting issue, to wit, whether soldiers themselves can be regarded as means of killing. In a sense, yes: without rank and file soldiers, the killings ordered by the government via the general staff simply cannot take place. This however is compatible with the thought that those who provide those means of killing are not doctors and food providers but, rather, recruitment personnel (of whom I have little difficulty to say that they are liable, in war, if the war which they help fight is unjust, etc.)
At this juncture in the exchange, Victor revisited his earlier point that, by curing soldiers, we enable them to do all sorts of valuable things: not so by providing them with weapons. But I think that the contrast is overdrawn. First, a soldier whom we cure might go on to do pretty horrible things; second, to reiterate, a soldier on the unjust side might nevertheless have the right to defend himself on occasions and, indeed, to defend others (eg: German soldiers defending German women from gangrape at the hands of the Soviet army in the winter and spring 1945). Third – after the war – those soldiers on the unjust side who survived may well make a valuable contribution to reconstruction, etc. (This point raises a very tricky issue: should the just belligerent take into account, when deciding whether to kill unjust soldiers, the post bellum costs (to the community unjustly engaged in war, and to itself) of a wholesale destruction of its enemy’s manpower? And so on.)
Reply to MAssimo
1. Pure functionalist argument (PFA) Massimo distinguishes between two needs which combatants incur whilst fighting: those the meeting of which serves to enhance their ability to fight and those the meetings of which is not directed to that end. Ex: if a soldier breaks his leg in battle, fixing his leg does not serve to enhance his fighting ability.
Well, perhaps it does. That said, Massimo zooms in on something which I did not make sufficiently clear in the paper, when criticising the pure functionalist view. To wit: I accept the PFA’s premise that there is a difference between those two kinds of needs, and then I seek to show that, on those terms, PFA cannot offer blanket protection to civiliansw. That view is compatible with the claim that doctors who only pure welfare needs (for lack of a better label) are not liable. The question, then, is whether, on my view of noncombatant immunity, they are liable – whether, that is, a doctor is liable if he treats a soldier’s broken leg in the knowledge that this soldier will thus be able to get back to the front and carry on killing unjustly. My view is that, under certain conditions (going back to threshold liability) he is liable, even if it is appropriate to regard his medical contribution to the soldier’s ability to fight as non specialised. Again, however, this is not something which I discuss in my rebutall of PFA. Though Massimo is right when he notes that my claim ‘if direct participation is a sufficient condition for liability then both civiliansm and civiliansw are liable’is too strong.
2. Moralised functionalist argument (MFA). My rejection of this view applies to both specialised and non specialised treatment. Massimo objects, via the claim that we may well have obligations to compatriots which outweigh obligations to foreigners. He uses the example of providing medical care to his sister who (he knows) will then go on attacking someone else unjustly, and claims that he is under a duty to save her (though not under a duty to provide her with weapons.) Now, it is not clear to me at all that he is under a duty to save his sister in that particular case, nor in fact that he is permitted to do so (though he is excused). But even if I am wrong on that front, the example of the sister does not readily translate to that of compatriots: for even if there is a special relationship between compatriots, it is rather different (one would hope!) to familial relationships. Given that, in the case of the sister, it is the familial nature of the bond which most plausibly explains the presence of a permission (or a duty), it is not clear at all that the case helps with compatriots.
Yet, here is a point against my own view, and in support of Massimo: soldiers from the same company/regiments are trained to build very intense bonds of trust and loyalty with one another (after all, their lives depend on this.) So perhaps the provision of medical care by one soldier to another, within that setting, most closely resemble the sister example. Even then, it is not clear at all that the argument would translate to less personalised relationships between food providers and soldiers.
3. Massimo’s third comment a tough one for my argument. As he notes, I make heavy weather of the point that a doctor who helps an unjust combatant must take into account the fact that someone will/is extremely likely to die, as a result of their contribution. And I justify my conclusion (that the doctor ought not to help) by appealing to the doctrine of acts and ommisssions: failing to help (the soldier) is morally preferable to contributing to harming (the victim). Massimo notes that there might be cases where a failure to help is worse than a contribution to the wrongful imposition of a harm, eg when the contribution is marginal.
I cannot but agree to this. I also agree that whether or not the doctor’s role-based obligation to her combatant-patient depends on the kind of contribution to an unjust war which the now-healthy combatant will make (is he a general, a rank and file soldier, etc.) So more work needs to be done on this. Note, however, that my conclusion at the end of section IV (where I discuss MFA) is weaker than Massimo’s objection seems to imply: for all I say there is that ‘the mere fact of providing medical care to combatants provides the doctor with a moral justification for so acting irrespective of what those combatants will thereby be enabled to do. The intuitive power of the Hippocratic Oath argument is much diminished once one brings into view the fact that, in the cases that occupy us here, the doctor’s decision to treat the unjust combatant proves highly costly for the latter’s enemy – namely, the cost of being subject to a lethal threat to which they are not liable’ (55 – ital added here). That conclusion does make space for Massimo’s worries about the relative moral status of severe failure to help v. making a small contribution to the unjust imposition of a harm. In particular, it does allow me (I think) to agree with his claim that the mere fact of contributing to an unjust threat is not enough to exempt the doctor from a special, role-based duty to help. It remains the case, however, that the MFA cannot serve to provide a blanket exemption to doctors (as I think Massimo would agree.)
And this concludes, for now, my replies to those many, very constructive comments.
Cecile
Thanks very much for your fascinating paper and contributions to the discussion here, Cecile. There’s obviously a lot to take in (which is what I suspect many are doing now), but hopefully the dialogue will continue once your thoughts have been digested.
I thank Cecile Fabre for her clarification that here we are concerned with the ethical aspects during an ongoing war only after it has unfolded and events preceding is not within the scope of our discussion. The discussion so far has veered around ‘functionalist’ approach and ‘threshold limit’ of the function of civilians in an ongoing war that makes them liable to be attacked and killed. On the functionalist approach my main objection is on a somewhat different premise. During the course of an ongoing war, just or otherwise, all kinds of rights of civilians (human rights, negative or positive rights) are the first casualty irrespective of which side of the war they are unfortunate to be in. At the least their movements and daily lives get disrupted, they start living under constant fear from getting caught in the crossfire, and at the worst they may be used as a pawn or human shield to protect the fighting military (an ironic twist of international human rights norms which seeks to protect civilians and the military uses to its advantage). They no longer enjoy a free will of their own to decide on the morality of their own actions. Even if they were in a position to do so it is practically impossible for the attacking side to know this for sure. Therefore, they are entitled to benefit of doubt and it is the moral duty of the attacking side to recognize it as such.
As for the threshold limit of the civilians function during a war that might invite liability to be attacked and killed I would resort to simple conventional dictionary distinctions between ‘military’ and ‘civilian’ rather than tread into dangerous territory that could lead us to practically justify any action and any amount of killing of civilians.
Military: A member of the armed forces .. (period) ( no distinction as to whether formally inducted by a nation or informally coerced by a belligerent group or self inducted through righteous indignation)
Civilian: A person whose primary occupation is civil or nonmilitary (that is, not a member of the armed forces)
The threshold word is ‘armed forces’. A military is a person who has been a part of the armed forces prior to outbreak of the war and continues to be so during the war until he is captured and disarmed where by he attains the status of a prisoner of war. A civilian by contrast was not a part of the armed forces prior to outbreak of the war and continues to be so during its tenure until the time he gets inducted into the armed forces and has the potential to kill the enemy military directly, not by any other stretch of imagination. Military doctors, nurses, and unarmed people who provide indirect support to the army, should be also considered in the same category and immune to liability. Unintentional killing of such categories of people is at par with the unintentional killing of civilians both of which constitute collateral damage and has to be kept well measured.
Therefore, my view is that civilians, the way defined above, can never invite the liability of being killed. Immunity from harm is their human right. However, for them, the unfortunate probability of being killed as a part of inevitable collateral damage in any war still remains. This is an issue that rests solely on the attacking side of a just war who has to decide on, “The permissible limits of collateral killing in a just war and liability for compensatory justice thereafter for collateral damage,” so that the war continues to remain a just one when judged years after it is over.
Pinto (if I may) – I have doubts about the usefulness of the distinction military/civilians. For a start, on the account you suggest, why would a military doctor (as opposed to a civilian doctor working behind the frontline) be immune from attack? Second, where would you put civilian contractors (esp those who actually fight alongside regular uniformed troops)? Third, what about so called irregular combatants in general? To put the point more widely it seems to me that the notion of direct participation, which can cut across the military/civilian distinction, is a more fruitful (though less neat) way of looking at the issue of liability.
Best
Cecile
Cecile – Thanks for raising your doubt and questions. First, a military doctor and a civil doctor both have the moral duty to treat all patients as equal humans who have a right to live and restore suffering humans to good health, very much akin to a civilian doctor treating a convicted murderer or suspect irrespective of what the murderer does after recovering. A military doctor by his proximity to the war zone is more susceptible to accidental or mistaken attack but is not morally liable to be attacked by virtue of being a military doctor ( as opposed to a civilian one) until he takes up arms to fight the enemy alongside other soldiers. Civilian contractors and all other general combatants who fight with arms alongside the regular soldiers are at par with mercenaries and get included in the definition of ‘armed forces’ in my earlier post.
Treating civilians and military as a continuum in an ongoing war runs a very grave danger of erring and landing at the wrong end of the infinite continuum of morality. Nevertheless, it is a worthwhile exercise in search of that elusive line, that ideal shade of grey, between pure white and pure black even if to finally come to a conclusion that a functionalist approach is neither here nor there and the threshold limit is that ideal shade of elusive grey.
Thanks for giving me this opportunity to express my views on your critical and commendable analysis on a very practical real life issue, the conclusion of which I am largely in agreement with though some my reasons may be at variance.
Best
Pintoo
Pintoo, thanks for this – esp for the clarification regarding your viewson mercenaries.
Let me respond to the very first claim you make in your post, as regards to doctors (whether civilian or military)’ moral duty to treat all patients equally. I try to show, in section IV of the paper, that this claim, precisely is too broad – or (better put perhaps) not sufficiently attentive to two important and related points: (a) the doctor’s treating an injured soldier will sometimes have as a consequence the wrongful death of others (at the hands of the now cured soldier.) And that, it seems to me, must be brought to bear on the second point, to wit, (b) whether the doctor’s special duty as a doctor always outweighs her general duty, qua individual, not to contribute to the imposition of a wrongful harm. I do not think that it does.
Best
Cecile
Cecile, may we consider these for the military doctor –
a) she is a person with a sense of duty above all else of treating human beings as end in themselves and restoring them to good health to the best of her ability
b) she has never attempted, at least consciously, to infringe on any humans negative right
c) she considers it her noble duty to respond to a positive right that is of utmost importance, the human right to live
d) she assumes and hopes that the person she is treating would realize what grave injury means and would be a reformed person hereafter and not only quit the ‘unjust combat’ but preach to others why they should not injure others
e) she had become a military doctor with the noble intention of saving lives of her brethrens gravely injured in a war while protecting their country but has now got caught on the wrong side because power has been captured by an unjust dictator who wages ‘unjust wars’
Would it be just for us to impose on her the liability of being killed just because she happens to continue as a military doctor and has refrained from making Utilitarian assessments of the non-medical consequences of her treatment?
A consequence by its very nature of lying in the future has infinite possibilities with varying probabilities, wherein lies the limitation of an otherwise very valuable and very useful ethical norm.
Let us put on the ‘veil of ignorance’ for a while and think whether we would like to be born a patient in a society where doctors also attempt to diagnose the moral orientation of a patient and consequences thereof before starting treatment, or would we like to be born a doctor in such a society?
Let us don the cap of a person of great virtue for a while and try to imagine whether the person would make moral judgments that may lead to societies that no one would like to be born in?
I hope I have been able to express my views on your first point (Sec IV). I agree with you that it is too broad a view, but hold that it may be the most ethical precisely for the same reason. A broad structure that has many legs to stand on, Kantianism, Justice, Human Rights and Virtue and possibly a weak Utilitarian leg.
Your second point I agree with only partially because I believe that any professional code ceases to be of any ethical value the moment it crosses from being a subset of ethics to an intersecting set. A doctors professional duty should never ever outweigh (as against ‘always outweigh’) his moral duty because so called ‘professional codes of ethics’ are very loose and insecure pegs to hang our moral judgments on.
Sorry, the post has gotten too long.
Best,
Pintoo
Pinto, thanks for your comments, which raise difficult, troubling issues. Regarding (d): well, that doctor may well nurture the hope that her patients would see sense, but I am troubled precisely by the case where she can reasonably be expected to know that they are not seeing sense. And, regarding (e): take the case of someone who enlists in the army with a view to defend his country from an unjust aggression, indeed even perhaps with a view to fight a war of humanitarian intervention abroad. Suppose however that a dictator seizes power, and presses the army to unjust ends. Or take munition factory workers, with a similar set of motives and constraints. Would you say that they too are immune from attack?
On reflection, it seems to me that you and I would perhaps disagree on specific cases (should *this* doctor be liable, or *that* one? etc) but not on the general principle that special, role based duties to help can sometimes (often?) be overriden by general obligations not to contribute to the imposition of a wrongful harm.)
Best Cecile
Cecile, thanks for your reply and sorry for the delay in expressing my views on the two points raised by you, d) and e).
My views on (d) is already there in my previous post of ‘a weak utilitarian leg’. On (e) I would like to maintain that as long as a person is a part of the ‘armed forces’ and has potential to kill directly his moral immunity gets automatically lost. However, a munition factory worker is not liable morally just because he works in a factory called ‘Ammunition Factory’. The supply chain for manufacture of arms and ammunitions not only extends right back to all the mines from where the metals and ingredients come but on its path lies a whole host of other factories that cover really a very large part of the countries manufacturing and processing units. The easiest and most convenient way of decapacitating any arms manufacture is to attack all the water resources in the country and its pipe lines. So, where do we draw the line? Here lies the problem of making ammunition factory workers liable. If one decides to draw the line at the boundaries of the factory called ‘ammunition factory’ still the workers who get killed fall in the category of collateral unintentional damage much the same as armed extremists in civilian buildings and civilian residents held hostage therein who are not morally liable but may get killed.
Best, Pintoo