Scheffler’s paradox: Persons vs. animals

“Scheffler’s paradox” is a puzzling feature of the moral beliefs of most deontologists.
According to these beliefs, it is wrong for you to kill an innocent person even
if your killing the innocent person is the only way to prevent five killings of innocent persons from
being perpetrated by someone else. What could possibly explain this?

This feature of deontologists’ beliefs cannot be explained purely by their commitment to the moral importance of such distinctions as doing vs. allowing, intending vs.
foreseeing, or the like. Most deontologists think that there are stronger reasons
against doing harm to non-human
animals than against allowing harm to
animals; many also think that there are stronger reasons against acting with
the intention of harming an animal
than against causing such harm to an animal without intending it. But surely you
could permissibly kill one bear if
that is the only way for you to save five other bears from being killed by someone
else. In general, it seems that Scheffler’s paradox does not really arise for killing
non-human animals.

So what is the special
feature of persons that lies behind Scheffler’s paradox?

When no persons are involved, if you take the “interventionist” option of killing one bear to save five, you both actively do
and intend harm to one bear; but you also both actively do and intend good to the five other bears. So it seems that in the case involving bears, the reason for the interventionist option and
the reason against it are in the end fairly evenly balanced.

What is
different in the case involving persons? The pattern of doing vs. allowing and intending vs. foreseeing is just the same as in
the case involving bears. So I propose that the relevant difference is that when
persons are involved, the values of
the relevant consequences are
different, on the grounds that these consequences involve relationships between persons.

I believe
that there are certain intrinsic values and disvalues exemplified by
relationships between persons. For example, other things equal, when one person
saves another person’s life, he puts himself into a good relationship with that
person; when one person fails to save another person’s life, he puts himself
into a fairly bad relationship with that person; and when one person kills another,
he puts himself into an unspeakably terrible relationship with that person.

However, if
you take the interventionist option of killing one person in order to save five,
your relationship with the five is affected by the fact that you saved them by means
of sacrificing the one. In effect, you have put them into a relationship with
the one that is almost as horrible as the unspeakably terrible relationship into which
you have put yourself with the one. So overall the relationship with the five into
which you put yourself is not clearly a good relationship – you save them,
but only by making them the beneficiaries and cause of the involuntary
sacrifice of the one.

Thus, I propose, the way in which the interventionist option saves the five is sufficiently tarnished that it does not count as a clearly “good consequence” of the option at all. On the other hand, the bad consequences are intensified – they involve not just a person’s losing his life, but also your putting yourself and the five into various horrible relationships with him. So the reasons against the interventionist option are now significantly more powerful than the reasons in favour.

Of course, the “non-interventionist” option has some pretty bad consequences too. By not saving the five, you put yourself into a fairly bad relationship with them; while by not killing the one you put yourself in only a very modestly good relationship with him. Still, you did not actively bring about these bad consequences; you only allowed them to happen. For a deontologist, the fact that the act involves only foreseeing and allowing harm (not intending or actively doing harm) significantly weakens the reason against the act that arises from these bad consequences.

Overall, then, the
reasons against the non-interventionist option are markedly less strong than
the reasons against the interventionist option; the balance of
reasons tells decisively against the interventionist option. The special feature of persons that lie behind Scheffler's paradox is the special range of values and disvalues exemplified by relationships between persons.

32 Replies to “Scheffler’s paradox: Persons vs. animals

  1. Ralph,
    Do you have the same reaction if we stipulate into your case that the would-be victims never come to know that their lives are in danger in the first place?

  2. Hi Ralph,
    I’m not sure I see how your view distinguishes the human cases from the animal cases. The way in which your action in the human cases affects the relevant relationships seems entirely determined by factors like whether you do or allow harm, or intend or merely foresee harm. The thought seems to be that we have reasons not to place ourselves into certain sorts of relationships with other humans, and that when we do or intend harm to them we place ourselves in very bad sorts of relationships, relationships that are much worse than the ones that result from our allowing or merely foreseeing harm. But this only makes sense, as far as I can tell, if we assume that doing and intending harm are worse than allowing and foreseeing harm. What else could explain why the resulting relationships are so much worse in the cases in which you do or intend harm?
    If we think instead that the facts about the sorts of relationships to others that we put ourselves in by doing/intending vs. allowing/foreseeing harm explain why, in the human cases, the reasons against the interventionist option are stronger than the reasons against the non-interventionist option, it’s not clear to me why that wouldn’t also be the case in the animal cases.

  3. Hello Ralph
    Suppose A sacrifices B as a means to save 5. We might think: A uses B and hence treats B as a tool for the sake of the 5. The relationship between A and B is man-tool.
    You also suggest that there is a problematic relationship between B and the 5. I wonder if you could clarify that. The 5 are beneficiaries of A’s man-tool relationship with B. But without something further, say endorsement, encouragement or something of A’s action, why should the relationship of the 5 and B be tarnished in the same kind of way as the relationship between A and B?
    Does the mere fact that harm to a person is a cause of my survival render my relationship with that person defective? I doubt it. Imagine that you and I are in a cave. A bear enters the cave and eats you and there is nothing I could do about it. As he’s no longer hungry he leaves me alone. Harm to you is the cause of my survival. But I don’t think I have a defective relationship with you.
    Of course the 5, in the previous case, are beneficiaries of wrongdoing, and they ought not to be glad about that fact. But the defective relationship, if that is the right way of understanding these cases, seems to be between A and B, not between B and the 5.

  4. Christian — Yes, I have basically the same reaction to these cases regardless of whether or not the victims know that they are coming to harm. (So I am using the term ‘relationship’ in a very broad sense, rather like the broad sense that Scanlon uses.)
    Brian — I certainly think that the doing/ allowing and intending/ not-intending distinctions are crucial to understanding all reasons for action. So I just wanted to argue that these distinctions can’t give the whole story about Scheffler’s paradox. If they did, then surely it would be impermissible to kill one bear in order to save five!
    So I agree that “the way in which my action in the human case affects the relevant relationships” is partly explained by the significance of intention and doing vs. allowing, etc. But it is also explained by the fact that persons are capable of genuinely cooperative relationships based on mutual consent — whereas such relationships are not really possible with non-human animals. The distinctive values and disvalues of relationships between persons are also part of the story.
    Victor — I agree that the 5 shouldn’t feel guilty or blame themselves for A’s killing B unless they endorsed or encouraged the act. But the 5 are still the cause of A’s killing B (A would never have killed B unless the 5 needed to be saved), and they are also the beneficiaries of B’s being killed. Their relationship to B is a bit like the relationship of a slaveowner’s descendants to the slaveowner’s slaves: the descendants have inherited the wealth and advantages that originated from the unjust treatment of the slaves. (Compare what Scanlon calls an “objective stigma”.)
    Of course, I agree with you that the most defective relationship here is between A and B. But I wanted to explain why this defect isn’t outweighed by the wonderfully good relationship between A and the 5. My suggestion was that because A causes a certain defect to be present in the relationship between B and the 5, there is also a defect in the relationship between A and the 5.

  5. Ralph
    Thanks for the response.
    To test the significance of relationships between the five and B in explaining whether A’s action is wrong we might consider cases where the beneficiaries of the actions are animals.
    Compare
    1) A causes B to suffer pain as a means to prevent the m degree of pain of n animals.
    2) A averts the threat of m degree of pain to n animals and B is harmed as a side-effect.
    It seems easier to justify 2 than 1 (n or m must be greater in 1 than 2 to render the action permissible).
    Whereas our intuitions alter considerably when we substitute animals for humans with respect to the agent harmed (in that it is permissible to harm an animal but not a human as a means to a somewhat greater good) they do not alter considerably when we substitute animals for humans with respect to the agent benefited.
    In other words, when comparing the permissibility
    A harms B as a means to a greater good to C
    and
    A does a greater good to C harming B as a side-effect
    it matters that A and B are humans. It does not matter whether C is human.
    Shouldn’t we then conclude that the relationship that would be established between C and B does not contribute to, but is rather parasitic on, the wrongfulness of A’s conduct?

  6. Victor —
    Well, as I said to Brian above, I hold that the distinction between results that are intended and mere side-effects matters to all reasons for action. So even if B is not a person but an animal (or indeed a priceless ancient statue), it is worse to harm B as an intended means to some further goal than to achieve this goal by means that have the side-effect of harming B.
    Of course, if B is a person, and C is a non-person, then it will very often be wrong for A to harm B as a means to benefiting C; and I can’t appeal to any defect in the relationship between A and C to explain this. But obviously I can appeal to the fact that there is no interpersonal relationship between A and C at all! The problem that I wanted to solve arises in cases where there seems, at least prima facie, to be a good relationship between A and C. I wanted to solve this problem by arguing that where persons are involved, this relationship is not as good as it may at first appear.
    Of course, the defect in the relationship between C and B is parasitic on at least one of the defects in the relationship between A and B (and on the nature of the relationship between A and C). But I would deny that the latter defect can be identified with the “wrongfulness of A’s conduct”. There are cases in which the relationship between a killer and his victim has the very same defect that it normally has, but nonetheless the killing is not wrong (i.e., when the reason against killing is not silenced, but simply outweighed).

  7. Hi Ralph,
    A devil’s advocate argument that your view does not dispel Scheffler-like worries:
    (1) A knows that B, C, and D will jointly torture and then kill T over several months. A kills T to prevent B, C, and D from getting into such horrible relationships with T and to prevent T from suffering through the torture.
    (2) Admittedly, A thereby gets himself into a very bad relationship with T. And A also gets B, C, and D, into bad relationships with T (they are, as you say, like descendants of slaveholders).
    (3) But, A reasons, [One killer and user relationship and three bad-beneficiary relationships] is more choice-worthy than [One innocent bystander relation and three torturer and killer relationships]

  8. I should add this to A’s reasoning:
    Intentionally entering the relationship of mercy-killer with someone is better than intentionally entering the relationship of instrumental-killer with him. I nonetheless grant that intentionally entering a relationship of mercy-killer with T counts against my killing him (it is a reason-against), but it counts much more weakly than would my intentionally killing him for merely instrumental purposes.
    And I think that that weaker (con) reason is outweighed by the considerations relating to the resulting relationships.

  9. I thought that the Sheffler paradox was supposed to be structural problem that arise whenever the option not to intervene arises from some value of the consequences be it loss of life, relationships, or whatever.
    So, suppose I step into a situation in which five other people are in the intervention situation. They all are willing to intervene – to shoot one person so that five others do not shoot further five people. Imagine that by shooting one person I could make each of the five people not to intervene. Should I intervene? If I intervene and shoot the one guy, the five will not intervene and so they will each be in a good relationship to one person each whereas they haven’t lost as valuable relationships to further five. If I don’t intervene, I will be in a good relationship to one person – the one I didn’t shoot, but five people will not be in good relationships but rather in bad one’s to the five they saved by shooting one person each. If I maximize valuable relationships, I should kill the one guy because there will be more valuable relationships in that way. [I know there are problems with case but it has the right structure]
    So, something else than agent-neutral, intrinsic value of certain kind of relationships have to be doing the work. Perhaps the idea is that I should be only concerned about my relationships and their agent-relative value. But, it was just this point that was supposed to be paradoxical.

  10. Brad —
    As I said in reply to Brian and Victor above, I’m assuming that the deontologists will continue to rely heavily on distinctions such as doing vs. allowing, intending vs. not intending, and so on. So of course the impersonal value of the consequences are better if we have [one killer-and-user relationship and three bad-beneficiary relationships] rather than [One-innocent-bystander relationship and three torturer-and-killer relationships].
    But for a deontologist, the strength of the relevant reasons for action is determined, not merely by the impersonal value of the consequences, but also by (i) whether or not the agent intends those consequences, and (ii) whether these consequences are something that the agent does or merely allows. This is why there are on balance stronger reasons for the course of action that involves merely allowing and not intending the consequence [One innocent-bystander relationship and three torturer-and-killer relationships] than for the course of action that involves actively doing and intending the killing of T and the creation of these bad-beneficiary relationships between B, C, D, and T.
    You’re probably right that the fact that A’s killing T is a mercy killing means that the reasons against A’s killing T are a lot less strong than if A were killing T to save someone else (say, U). But this is something that a “relationship-focused” approach seems well placed to explain.
    Jussi —
    I’m not quite sure what you’re saying here. (Of course, the terminological question of what is the standard interpretation for the phrase ‘Scheffler’s paradox’ doesn’t really matter. What really matters is whether we can justify and explain these deontological intuitions….)
    1. Are you saying that the very idea of agent-relative value is paradoxical? I wouldn’t accept that. I think that reasons for action are always agent-relative (they have to be, if they are to be sensitive to the doing/allowing distinction and the like); and reasons for action can be expressed by talking about what is a “good thing for me to do” — and so in a sense these reasons define a kind of (agent-relative) value.
    2. Are you saying that it is paradoxical to claim that a prohibition on maximizing an impersonal value could be grounded purely in that very value? Then I would agree that it is paradoxical — indeed absurd — to claim this. But I wasn’t trying to ground a prohibition on maximizing an impersonal value purely on that very value itself. I was suggesting that this prohibition could be grounded partly (i) in the intrinsic agent-neutral value of relationships, and partly (ii) in the agent-relative significance of distinctions like doing vs. allowing, intending vs. not intending, etc.

  11. Ralph
    I found your post, and the ensuing discussion, fascinating.
    A couple of comments:

    1. In connection with Christian’s comment (does the fact that victims know, or not, that they are coming to harm, make a difference ): I agree with you that it makes no difference and see no difficulty there – anymore than I see a difficulty in averring that one enters in a pretty horrible relationship with one’s spouse if one is unfaithful even though he/she will never be aware of it.
    2. BUT: the notion of relationship warrants a bit more elaboration there. Suppose I have six cats, and have to kill one in order to prevent the other five from being mauled to death by some other agent. I believe that I may so act, but that intuition is less strong (to me) in that case than in the case of the bear (or in the case of 1 v. 5 cats whom I do not own/care for etc). I would actually feel pretty bad about that precisely because I have a relationship with all six (though I would not hold the view that I put the surviving 5 in a bad relationship with the 1.)
    3. The foregoing leads me to wonder about the relative weight, in your view, of the bad relationship between the agent who kills and the victims/beneficiaries of the killing, and the bad relationship which the agent who kills creates between victims and beneficiaries.
    4. Finally, I wonder how you would extend your argument (if at all) to human non-persons (and, I assume, who have no prospects to become persons: the permanently comatose, the incurably senile, etc.) Thus: may we kill one such human being to harvest his organs, with the view of saving five? And so on.

    Cecile

  12. Ralph,
    Thanks for the response – I now see that it was a mistake to think you aimed to be dispelling the paradox for persons & that your story about that would depend on independent assumptions about the existence of deontic constraints. Thanks for your patience.
    Like Cecile, I now want to hear more about the relationships in question. Specifically, I want to hear more about why you think only person-to-person relationships can generate Scheffler-type effects.
    Consider a case involving a blind woman with a seeing eye dog. She thinks it is impermissible for her to intentionally sacrifice her dog to save three “stranger” dogs. And she defends that claim by appeal to the fact that if she were to do so she would be in a bad relationship (betrayer?) with her trusted companion.

  13. A follow-up on Brad’s comment: I think it would be really fruitful to consider the following cases systematically, and separately:
    – person-to-person
    – (nonhuman) animal to (nonhuman) animal
    – (nonhuman) animal to person
    – human nonperson to human nonperson
    Etc…
    Cecile

  14. Cecile and Brad —
    I completely agree with Cecile’s intuitions about cats, and with Brad’s intuitions about guide dogs (as seeing-eye dogs are called here in the UK). When one takes a cat into one’s home and starts treating it as a pet, one is assuming a kind of responsibility for it. Admittedly, the cat is incapable of reciprocating in the way that a person could, but the cat does come to regard its owner differently from the way in which regards strangers. So it is indeed an ethically significant relationship.
    Although I believe that I can have an ethically significant “relationship” (in the Scanlonian sense) with persons who are complete strangers to me — say, with the Haitians who are helped by my donations to OXFAM — I doubt that I can have an ethically significant relationship with a wild animal. This is why my original example concerned bears, and not cats or guide dogs.
    At all events, I’m very far from having a complete theory of the value of relationships. Relationships that involve genuine cooperation in shared activities (with mutual consent etc.) seem to me the most important species of relationship, but they are certainly not the only kind that matters. (E.g. my relationship with future generations with whom I will never interact is almost important, especially for understanding issues like global warming and climate change.) I completely agree with Cecile that it would be really fruitful to explore these different kinds of relationship systematically.
    Finally, let me say something quickly about Cecile’s question about the “relative weight” (in my view) of “the bad relationship between the agent who kills and the victims/ beneficiaries of the killing, and the bad relationship which the agent who kills creates between victims and beneficiaries.”
    I actually didn’t see these as being weighed against each other at all. My idea was that there is an explanatory dependence here. The fundamental factor is (i) the badness of the relationship between the killer and his victim; this underlies and helps to explain (ii) the badness of the relationship that the killer creates between the victim and the beneficiaries of the killing; this in turn explains (iii) why the relationship between the agent and the beneficiaries is not the unmitigatedly good relationship that one might initially have thought.

  15. Ralph,
    sorry about not being clear. I was trying to say a couple of things. First, if you think that agent-relative values and reasons are not paradoxical, then there is no paradox to talk about in the first place. I agree with you that there are agent-relative values and reasons. That’s why it is really easy to formulate a non-paradoxical theory that creates the right sort of agent-centred constraints (in fact, I did this in one way in my JESP paper). Of course, there is a further question of what is the bearer of the agent-relative values – be it personal relationships, agential involvement of the agent, the patients and their uniqueness and dignity, or whatever. But, that’s a substantial issue.
    In any case, the paradox only gets going if one tries to create the permission or requirement not to intervene from an agent-neutral value. You seem to be doing this and that’s why your solution will lead back to the paradox. There will be cases (like the one I described) where by violating the deontic constraint to kill you can bring about more of the agent-neutrally valuable relationships to other people because they don’t then kill to save even more people. If there is a requirement to maximise agent-neutral value, then we ought to kill in this situation ourselves (one relationship is lost but five gained). This means that the agent-neutral value of the relationships cannot create the deontic constraint. This work has to be done completely by something else.

  16. On Cecile and Brad’s point
    If I take (or otherwise have) special responsibility for the care of a person or animal, that can provide a constraint on harming that person or animal where without the special responsibility harming them would be permissible. For example, I think:
    It is permissible to divert a threat away from 5 people harming 1.
    It may be wrong to do this if the 1 is my child.
    It is permissible to divert a threat away from 5 cats harming 1.
    It may be wrong to do this if the 1 is my pet cat.
    Because it is normally wrong to harm a person as a means to the greater good there is no issue about the permissibility of harming my child in particular for the greater good. It is wrong to harm my child as a means to a greater good just as it is wrong to harm anyone else as a means to the greater good.
    It is often permissible to harm one animal as a means to the greater good (I can kill one sheep in order to save 5 from being killed). But as responsibility based constraints operate to constrain harming as a side-effect we shouldn’t be surprised to find that responsibility-based constraints might make it wrong to harm my pet or guide dog as a means to the greater good as well.
    Ralph’s original question is ‘What feature do humans have that explains why it is wrong for a person to kill another as a means to save five from being killed’. It might difficult to see how special responsibilities can help to explain that, for the reason that special responsibilities can provide a constraint on harming the person or animal one is responsible for either as a means or as a side-effect.
    Ralph’s thought is that if I kill one person to save five from being killed I have a bad relationship with the one, which would infect the relationship that I have with the 5 through saving them. What we need to know is why this is not also true of animals. Why is it that if I kill one animal as a means to save the five my good relationship with the five is not infected by my bad relationship with the one.
    We might think that the answer to this is that there are some relationships between humans and humans that it is wrong to form where it is permissible to form equivalent relationships with animals. We can form special relationships with both humans and animals which can generate agent-centered constraints on actions. But we cannot permissibly enslave humans for the greater good where we can enslave animals for the greater good. In other words the range of relationships that we can permissibly form between ourselves and animals is broader than the range of relationships that we can permissibly form between ourselves and other humans.
    Does this help?

  17. Jussi —
    You say: “if you think that agent-relative values and reasons are not paradoxical, then there is no paradox to talk about in the first place.” I basically agree with this. There is no real paradox, only an apparent paradox. (So perhaps the term ‘puzzle’ would be better than ‘paradox’?) Indeed, for me the really puzzling point is the disanalogy between persons and valuable non-persons, which is what I aimed to highlight in my original post. (So perhaps it was slightly misleading for me to entitle my post “Scheffler’s paradox”. Still, I hope that you see what I was getting at.)
    You also say: “the paradox only gets going if one tries to create the permission or requirement not to intervene from an agent-neutral value. You seem to be doing this and that’s why your solution will lead back to the paradox.” But I made it clear that I am not doing this: at least I am not “trying to create” the requirement not to intervene purely from an agent-neutral value. I am trying to create it from both (i) the agent-neutral value of relationships and (ii) the facts about how the doing/ allowing and intending/ not-intending distinctions affect agent-relative reasons for action. So I’m not vulnerable to the objection that you mention here.
    Victor —
    I agree with everything that you say here. In the end, of course, I would like to explain why it is true that “the range of relationships that we can permissibly form between ourselves and animals is broader than the range of relationships that we can permissibly form between ourselves and other humans.” Presumably it has something to do with the different capacities of persons and non-human animals. (If there were intelligent Martians, we would have almost all the same duties towards them that we have towards human beings.) But all this would require a great deal of investigation of course….

  18. P.S.: Jussi, I’ve tried to explain how I think we can get a non-consequentialist conception of reasons for action, in spite of having a view that gives a central place to agent-neutral intrinsic values, in my recent paper “Intrinsic Values and Reasons for Action”, Philosophical Issues 19 (2009): 342-363.

  19. Ralph,
    This is very interesting. Of course if you appeal to capacities there will be familiar worries about the non-person human cases Cecile brought up earlier.
    You can deal with some of them by applying the conclusion from the guide dog case – you can say you have a special, contingent relationship with the human non-person who is your newborn son, for example and that that relationship grounds restrictions.
    But this will get more troublesome in, e.g., the stranger-new-born case.
    Are you, like me, unsatisfied with Scanlon’s suggested solution to this problem in WWO?
    [In effect, he says that non-person humans count as persons in moral reasoning because they are “human born”, and then asserts that this is not tantamount to “a prejudice, called ‘speciesism'” (185 WWO)]
    Any thoughts about how you will go on this issue?

  20. Thanks Ralph – I’ll need to check that. Can you just explain why your view isn’t vulnerable to the objection? Why is it that we shouldn’t maximise the number of agent-neutrally valuable relationships when doing so would clash against a deontic constraint? If I can make others to have more of these relationships by terminating one of my own by killing someone, what reason do I have not to do so in the revised cases? Maybe I’m just being thick.

  21. Brad —
    I’m very unsure about how to deal with the case of human non-persons. (Since I go for a timeless view of moral status, rather like Liz Harman’s, new-born babies who will eventually become persons are not a problem for me. My problem is with those who will never become persons, no matter what one does.) I’m a bit uncomfortable with the Scanlonian suggestion that the crucial fact is that these human non-persons belong to the “human family”, but I can’t actually see how to do any better…
    Jussi —
    The idea is that the way in which agent-neutral intrinsic values generate reasons for action is not, as the consequentialist believes, by generating reasons to promote the maximal exemplification of the value in the world as a whole.
    Very roughly, the idea is that the strength of the reason for action depends on the kind of relation in which the action stands to the intrinsic value. E.g. Is the action the execution of an intention to bring about a state of affairs that exemplifies the value (or disvalue) or not? Does the action actively create an exemplification of the value (or disvalue) or does it merely fail to prevent the exemplification of the value (or disvalue)? Etc. But the story is developed in much more detail in the published paper.

  22. I’ll need to see the paper but I remain sceptical. I can intervene in the modified case by violating the deontic constraint with the *intention* to bring about more instantiations of the agent-neutral value of the given relationships and I am doing so *actively*. I take it that this must make the reasons stronger rather than weaker. It would be odd if you had stronger reasons to bring about intrinsic value without intending to do so, or to fail to prevent the exemplifications of the value.

  23. Brad, Cecile and Ralph.
    The issue of human non-persons (or ‘never-to-be persons’) and the means principle seems pretty tough.
    We think
    1) It is permissible to harm an animal as a means to save five animals from harm.
    2) It is wrong to harm a human non-person as a means to save five human non-persons from harm.
    Why? One approach attempts to pick out qualities that human non-persons have in themselves that animals lack. On this approach it is difficult to resist the charge of speciesism.
    Another approach that might be worth exploring grounds the constraint in 2 on the duties that persons owe to non-persons. We might think that whereas there are animals that no person has a special responsibility for there are no human non-persons that no-one has a special responsiblity for. That is because parents owe a duty, a duty that extends beyond the parent’s death, to protect and promote the well-being of their children.
    To harm a human non-person as a means is to interfere with the duty of the parent as a means to a greater good. That sets back the value of the life of the parent.
    Now suppose that a parent doesn’t recognise their duty to look after her non-person child. The child is abandoned. This might seem to create a difficulty. Here is a very tentative and badly thought out answer. We might think that it is wrong to harm the child as a means on the grounds that the parent owes a duty to protect the child and hence when the child is harmed the value of the parent’s life (even if not his well-being) is diminished. The value of the parent’s life might depend on her achievements and not only on her well-being, and if her child is harmed her achievements are less.
    On this view, the human non-person matters in a special way in virtue of being a person’s child. There seems to me something intuitively right about that idea, though I can’t say I’m confident about it or about my explanation for it.
    Any good?

  24. Victor,
    That is an interesting suggestion. And I think it might actually be better than species-norm, genetic, or capacities accounts.
    Two comments:
    First, I suppose the main worry about the view is that it assigns the right deontic constraints, but that it does not get the ground of those constaints right: it (apparently) tells us that, strictly speaking, we do not owe it to the child to not treat him as a means – we owe it to the parent. As you say, “To harm a human non-person as a means is to interfere with the duty of the parent.”
    Put otherwise, I would guess that many who want to vindicate 1 and 2 would also want the following two further claims, which your approach (apparently) cannot vindicate:
    3) You wrong a human non-person if you kill him as a means to save five human non-persons from harm.
    4) You owe it to a human non-person to not kill him as a means to save five human non-persons from harm.
    But at least you get the extension of the constraints right, without raising worries about prejudice.
    Second, might not your approach apply to some non-human cases? Consider, for example, someone who intentionally breeds dogs.

  25. Brad,
    thanks. That’s helpful.
    I agree that we have a sense that the human non-person is wronged. But I wonder whether that might be accommodated within the view.
    Consider the following:
    A is attempting to rescue B through a supererogatory act. C interferes with A’s attempt in order to get A to rescue D instead.
    In this case, B has no right to be rescued by A. And yet C would surely wrong B as well as A by interfereing with A’s act. C owes it both to A and to B not to interfere with A’s act.
    The fact that B is wronged, in this case, seems to be derived from the wrongfulness of interfering with A’s autonomy.
    Perhaps we can say something similar in this case. The fact that parents have a duty towards their human non-person children makes it wrong to harm those children as a means. But both the child and the parent are wronged.
    On the second point, I agree that it might be wrong kill some animals as a means to the greater good. It would be wrong for you to kill my pet cat to prevent five pet cats from being killed.
    That being said, it does seem permissible to breed animals in order that they will be used as a means to a greater good (farming) but wrong to breed human non-person’s in order that they will be used as a means to a greater good. I’m not quite sure what to say about this. Why are we permitted to breed animals but not non-human persons without developing special relationships of protection towards them?

  26. Victor,
    Interesting; I will have to think about the first part.
    On the second part: Nice point about it seeming “permissible to breed animals in order that they will be used as a means to a greater good (farming) but wrong to breed human non-person’s in order that they will be used as a means to a greater good.” I agree. But I am less sure about this: “…we permitted to breed animals but not non-human persons without developing special relationships of protection towards them”
    I think there are restrictions indicative of relationships of protection in animal breeding cases: it is impermissible to breed animals in order kill them and make use of their organs – imagine someone breeding their dog who has a faulty organ in order to then raise and kill the offspring and do a transplant. And if you breed some animals I think you are also, for example, (pt) obligated to provide their basic needs and not cause them great suffering in order to maximize your profits.
    (a) If these claims are true, then on your account, I would wrong you, and perhaps your animal, if I were to cause our animal suffering to increase my profits, etc.
    I think this might actually be a nice feature of your account…if it is one!

  27. Thanks Brad. I think I agree that there are constraints on breeding animals in order that they can then be used in a way that harms them for a greater good. I doubt that constraint is as tight as it is in the case of human non-persons though.
    Perhaps we can say something like this: if we breed a human non-person for this reason we intentionally breed a person with a defect, and the human non-person that we breed has an interest in not having the defect. It would be wrong intentionally to give the human the defect when we could avoid that. There is no equivalent problem in the case of breeding animals. Again, I’m not sure though.

  28. Victor, how would you deal with the following two cases (pertaining to your point re duties owed to parents etc):
    – Orphaned children who (as far as we can tell) have no surviving relatives at all (not a fanciful example, as the sheer numbers of those children in war zones)? You might say of course that we owe a posthumous duty to the parents, etc. BUt there are serious worries about posthumous duties.
    – Children who have committed atrocities against their own relatives, incl their parents (again, cases abound in war) such that the latter do not (one might be tempted to argue) have a duty of care towards them. (It is thus not clear to me that a parent who has been repeatedly tortured by his child and has had to witness the latter hack their relatives with a machete, has a duty of care to that child.)
    On the point re supererogation, etc: if B lacks a claim against A to be rescued, then it is not clear to me why she has a claim against C that the latter not interfere with the rescue attempt.
    Here is a point, though, which you might perhaps find appealing, to wit: suppose that A starts rescuing B, thus creating some expectation on the latter’s part that she will be rescued and perhaps leading her to modify her own conduct (eg she would have jumped out of the window to escape the house fire but is now waiting for A to get the ladder, etc). It seems to me, in that case, that B might well *acquire* a claim against A that the latter continue, or, at least, that A not give up unless she has pretty good reasons for so doing (indeed, stronger reasons perhaps than she would have had to adduce in order to justify not starting the rescue in the first instance.) And if that is correct, then, I would agree with you that C owes it to B not to interfere with A. But note, here, that B does have a claim against A, which distinguishes that case from yours.
    Note: in that example, we have to suppose that B is cognisant of A’s effort. What about cases where she is not? There, I am not sure (to repeat) that the argument from supererogation will enable you to accommodate the surely correct intuition that the duty is owed to the child.

  29. Hi Cecile
    Thanks very much for these great comments.
    First, I should say that I only think the account is worth exploring, not that it is true.
    On the first point, the scope of the account does seem dependent on the existence of posthumous duties, such as the duty owed to van Gogh not to destroy his paintings. There is, of course, a lot of disagreement about whether there are such duties. I’m not sure, but I’m inclined to believe that there are.
    On the second case, the case you are imagining might be spelled out as follows. Suppose that a child tortures his parent. In consequence, no-one has a special duty towards him (assuming that the parent can lose parental responsibility in this case). The child then, through accident or disease, becomes a non-person. Would it be permissible to use the non-person as a means?
    The first thing to say is that I think that there are exceptions to the means principle in cases of serious wrongdoing, and that is important in justifying punishment (a little shameless plug for a forthcoming book called The Moral Foundations of Criminal Law seems justified here). But I don’t think I’m compelled to say that the non-human person must be available for use as a means in this case. We might say that the non-human child is no longer the same child as the person child, and yet is still the child of the parent.
    On A, B and C, I’m surprised that you don’t agree that B has a claim against C not to interfere. I thought it was a clear case.
    Compare: A attempts to give a charitable donation to B who is much worse off than A, after having given B all that justice requires, it would be wrong for C to prevent A giving the donation. Suppose that A attempts to give the donation, is frustrated by C. Now suppose that A thinks that the donation has gone to B. He leaves the remainder of his estate to D and then dies, ignorant of the fact that B never received the donation. A might not have a claim against D (we don’t have a right to unilateral gifts), but surely he has a claim against C for the money that he would have received but for C’s wrongful action. And he has that claim even were he not aware of A’s attempt.

  30. Victor, Brad, Cecile —
    I’m sorry that I haven’t commented on this thread for a while. In fact, I don’t have a very strong view about Victor’s suggested solution to the problem about human non-persons.
    I suppose that I’m weakly inclined to think making the human non-person’s parents so central to the story seems dubious. At all events, I don’t have the intuition that in using a human non-person as a means, we would thereby be wronging the human non-person’s parents: the wrong somehow really does primarily involve the relationship that we have to the human non-person himself. But this is puzzling, given that it seems prima facie that the range of ethically significant relationships that it is possible to have with the human non-person are not very different from those that are possible with a non-human animal like a cat or a dog….
    Can I ask your intuitions about the following example: Suppose that we discover an extraterrestrial species of persons; let us call them the rational Martians. Would it be equally wrong for you (a human agent) to use a non-person member of this rational Martian species as a means as to use a human non-person as a means, or would it somehow be less wrong? If it would be less wrong, would be it dramatically less wrong or only slightly?
    My intuition, for what it’s worth, is that it would be less wrong — but only very slightly. This seems to suggest that any duty that we owe to members of our “human family” as such is significantly less weighty than the duty that we owe to all family members of rational persons as such.
    If that’s right, then Victor would be right that the fact that the human non-person is a close family member (e.g. a child) of persons is the crucial fact somehow. I’m not sure that Victor has yet put his finger on exactly why this is the crucial fact. But if he is right that it is the crucial fact, that would already be an impressive contribution to our understanding of this difficult question!

  31. Dear Ralph, Victor, and Brad
    My intuition (to answer Ralph’s question) is that it would be *equally* wrong to use a Martian (as the word is used by Ralph) non-person, as a means, as it would be to use a human non-person. And for what it is worth, I am very drawn to that intuition – for if those Martians are persons, to wit, beings with the capacity for rational agency , then I do not see what relevant difference there might be between us, and them, which would affect the moral status of our using them as a means v. our using a human non-person as a means.
    Cecile

  32. I agree with Ralph and Cecile about ‘our’ treatment of martian non-persons. But I don’t think that this is enough on its own to establish that special relationships are what matter.
    An alternative is that X has a right not to be used as a means to a greater good only if X could have been a person.
    I don’t know much about conditionals, but we might think that a human non-person or a martian non-person could have been a person in that they would have been a person under some conditions near to the conditions that did obtain. In contrast a donkey could not have been a person in that the donkey could have been a person only under conditions that are distant from the conditions that did obtain.
    The reason to prefer a story based on special relationships are twofold. First, it is easier to motivate. Why should my status depend on what I could have been rather than on what I am? And second it seems to apply naturally to the cases of the harm to animals that we have a special relationship that we were discussing previously.
    Ralph is right, though, that we don’t yet have a clear articulation of how the special relationship story is to work.
    Let’s see if we can get the arugment so far clear at least.
    1) If a person, X, owes a special duty to a non-person, Y, it would be wrong for X to harm Y as a means to a greater good, such as preventing an equivalent harm to five non-persons.
    2) That is true even if the five are non-persons who are owed a special duty by five other persons
    3) That is true even if the five are non-persons who are owed a special duty by X.
    4) Just as it would be wrong for X to harm Y as a means to the greater good it would be wrong for others to harm Y as a means to the greater good.
    5) All of this is true whether or not X recognisees that she has a special duty to Y.
    6) Parents have duties to ensure that their children (whether or not they are persons)
    are protected from harm that extend over the whole life of the child. They must ensure that their children fare well even after the parents’ deaths.
    7) All humans are children of persons or stand to some person as children stand to parents.
    8) No human non-person can be harmed as a means to a greater good.
    What needs to be shown to verify this argument?
    a) We might doubt 6 (Cecile’s suggestion which we have discussed to some degree), but I think that 6 is clearly true.
    b) We might doubt whether this can explain that it is the non-person who is wronged (see the discussion between myself, Cecile and Brad above).
    c) Perhaps some might doubt 7. There might be humans who are non-person children of non-persons. I think, though, that the grandparents would have a duty of care to the non-person children of their non-person children.
    d)Perhaps we might doubt 5. Some people might think that the status conferred on a non-person depends not on a duty owed but on a duty recognised. I doubt that this is true though. In Cecile’s original post about the guide dog, she thought that the person with the dog would think it wrong to harm the dog as a means. I suspect that this intuition does not depend on whether she cares about the dog, but rather on objective features of the relationship between her and the dog. But I’m not sure.
    e) I think that Cecile doubted 3 earlier. She thought that it might be permissible for me to kill one of my pets as a means to save five of my pets. I think that this is probably wrong, but I’m not sure. I certainly think that it’s wrong for me to kill my non-person child as a means to save five non-person children of mine. If there is a disparity between these cases, perhaps that counts against the view.
    What else?

Comments are closed.