A Puzzle

Today in my Metaphysics seminar we discussed Derek Parfit’s “Non-Identity Problem.”  It’s a neat problem–here’s what I hope is a faithful version of that problem.

A 14-year-old girl is told by medical professionals that if she has a child before her 18th birthday, the child will have a 75% chance of being born with a debilitating disease (dd), but that the child’s life will still be worth living (the dd will not be so bad that it burdens her with little or no quality of life).  On the other hand, if she has a child after her 18th birthday, there is a 0% chance that her child will be born with a dd, and consequently her child will likely have a much better quality of life.  Soon after receiving this information the 14-year-old girl has consensual unprotected sex (with a 14-year-old male, say, to avoid issues of genuine consent) and becomes pregnant and nine months later has a child with a dd.  Was it immoral of the girl, knowing the risks, to have a child before her 18th birthday?

Most people think: yes.  At any rate, I’m inclined to say ‘yes.’  Under the circumstances, the morally right thing for the girl to do was wait until her 18th birthday.  But I’m also inclined to accept the following two assumptions, each of which strikes me as plausible:

Explanatory Assumption:  The girl’s act is morally wrong because it harms her child.

Analysis Assumption:  P harms P2 by X-ing only if P2 would have been better off if P had not X-ed.  (If P’s X-ing in no way detracts from P2’s well-being, then P’s x-ing has done P2 no harm.)

But it seems that I cannot coherently believe that (i) the girl acted immorally by having a child before her 18th birthday, (ii) EA, and (iii) AA.  Why not?

If we accept AA, then the girl cannot be said to have harmed the child born with a dd, since it is not true that that child would have been better off had the girl waited until her 18th birthday.  Rather, that girl would not have existed.

So if we accept AA, then we must reject EA.  But if we reject EA, we need a different explanation of why the girl’s decision was immoral.   Alternatively, we could keep EA, and reject AA.  But AA seems really solid. Alternatively, we could keep EA and AA, and reject the claim that the girl’s decision was immoral.

Which option should we choose?

This entry was posted in Uncategorized and tagged , . Bookmark the permalink.

20 Responses to A Puzzle

  1. juan says:

    another problem for the analysis assumption,very much related to the one in the text,is the case of killing someone.would X have been better off if i hadn’t killed her?well, the worlds we are considering when saying ‘had i not killed her’ are pretty much like any other world where she goes on living as usual,without me or anyone having attempted anything on her life.
    i guess we actually have to compare a world where she exists(goes on living) with one where she doesn’t (where she’s killed),and that’s ok as far as it goes, but how can we say which one is better for her,since in one she doesn’t exist?
    we could ask her which one she’d prefer,but then she might end up preferring to die.so if we include this kind of subjectivity in the analysis assumption,for ex,then we need to say that i’m not harming anyone if they prefer to die anyway.

    maybe AA is wrong.

    • David says:

      Hi Juan,

      This is interesting. I take it your suggestion is that there might be a non-identity problem with respect to ENDING a person’s existence that parallels the problem with respect to BEGINNING a person’s existence. However, I’m inclined to disagree.

      If I kill you, then I deprive you of a valuable future, or all the value that your future would have contained. Thus I clearly make you worse-off than you would have been had I not killed you. We cannot say, however, that I deprive an individual of a valuable future by not bringing him/her into existence, because in that case there’s no identifiable subject of harm. (It’s not like potential people are in a waiting room somewhere hoping to be let in.)

      • juan says:

        i see no reason to apply two standards to the two cases.if you say that it is me you are harming by taking my life(and you point at me, the real person), then i can point as well to Jennifer,say,who exists now,and say it is her i harmed by bringing her into existence.
        if you are correct that the two cases are not analogous,then you can only be correct(at least so far as your argument above goes) by rejecting AA,or reading the counterfactuals in it in a very non-standard way, which i don’t know what it would look like.
        perhaps it’s just better to abandon the formulation of the harm principle in terms of counterfactuals,otherwise the standard analysis of counterfactuals will lead to the analogy with killing

  2. Nates says:

    This is a great puzzle — it’s really got my head in a knot.

    It seems that, in measuring the harm done, we want to compare the situation of one possible child to the situation of the actual child, and the problem is that there is no one person who exists in both situations.

    I wonder if we could make any sense out of the idea that the harm is done to the set of her possible children? If that worked (and I’m really not sure it does), then I guess this would involve a rejection of EA (moving the italics from “harms” to “her child”)

    • David says:

      Hi Nates,

      –“I wonder if we could make any sense out of the idea that the harm is done to the set of her possible children?”

      That’s a clever suggestion, but as I understand the NIP it runs into the following fatal objection. If the girl’s act is immoral because it decreases the overall well-being of the set of all her possible children, then every time she declines to have intercourse she is at risk of doing something seriously immoral! (After all, she could be adding members to the ‘set’ that have lives of at least marginal well-being.)

      • Nates says:

        Hmm, you’re right. Have to think about this some more…

      • juan says:

        i don’t think this is fatal: why is anyone under the obligation of adding members with good lives to the set of possible kids?if she refuses to add to the set, she hasn’t done anything wrong.moreover,how would not adding members to the set harm the possible beings already in that set?the possible children are just fine without new members.

        however, i’m afraid i’m not clear what possible child we are talking about in this case and comparing to the actual one.if she hadn’t had intercourse,there would have been no child at all.that is what generates the problem. i don’t see any possible child in the picture.when we say that she doesn’t exist in that world,she actually doesn’t exist there,it’s not that she exists in a possible way, whatever that might mean.

  3. David says:

    I might modify AA as follows: “P harms P2 by X-ing only if, for some event Y which is a direct causal result of X-ing, Y deprives P2 of some benefit or enjoyment that is rightfully theirs.”

    I would then modify EA as follows: “Action X is morally wrong if it harms another person, and that the harm can be reasonably anticipated as a direct consequence of X.”

    These modifications address some issues I find problematic with the originally stated EA/AA:

    – Some things that I do are harmful to another person, but not morally wrong, such as competing with another person for a job (without which other job applicants would be ‘better off’), or driving on the roads (thus causing traffic, without which other drivers would be ‘better off’).

    – Unintended consequences that could not have been rationally anticipated generally shouldn’t make an action morally wrong, should they?

    – If X-ing doesn’t harm anyone directly, but X directly causes event Y, and event Y is catastrophic, shouldn’t the person who did X be held responsible, given that both Y and its catastrophic consequences could be reasonably anticipated from X?

    Given these modifications, I would accept an argument that the mother did in fact directly harm the child. Her actions caused pregnancy at an early age, which caused direct harm to her fetus, which caused the birth of a child with DD. All of these consequences could be reasonably anticipated. Just because the sex act didn’t directly cause the harm shouldn’t let her off the hook. Rather, some direct consequence of the sex act caused the harm, at a time when the child did exist (as a fetus) to receive the harm. Therefore she should be responsible.


    • David says:

      Hi David,

      Interesting comments. (By the way, is this David Sch., or David from my Metaphysics class?) In any case…

      You write:

      –“These modifications address some issues I find problematic with the originally stated EA/AA: Some things that I do are harmful to another person, but not morally wrong, such as competing with another person for a job (without which other job applicants would be ‘better off’), or driving on the roads (thus causing traffic, without which other drivers would be ‘better off’).”–

      Yes, but that’s why there’s no sufficiency claim in AA, only a claim about a necessary condition of harmful action. On to more important matters. You write:

      “I would accept an argument that the mother did in fact directly harm the child. Her actions caused pregnancy at an early age, which caused direct harm to her fetus, which caused the birth of a child with DD.”

      But the child in question would not have been better off if the girl had acted differently, nor would the child in question have received benefits that are rightly hers. THE CHILD IN QUESTION WOULD NOT HAVE EXISTED. So I don’t see how, according to your modified AA, you can claim that the particular girl in question was harmed. And hence I don’t see how modified EA is going to help, either.

      So it seems to me that the puzzle remains.

      • David says:

        This is David from class. In situations like these, I sometime go by David #2 to avoid confusion.

        You wrote, “the child in question would not have been better off if the girl had acted differently.” True, if you mean that she would not be better off had the mother not done X. But the child in question would have been better off if it had not contracted the DD, event Y. I can’t see a case for saying that X did not cause Y, and that I think is what underlies the intuition saying that X was wrong.

        Here’s the original AA: “P harms P2 by X-ing only if P2 would have been better off if P had not X-ed”

        Here’s a much simplified version of my modified AA: “P harms P2 by X-ing only if X-ing causes event Y, and P2 would be better off if Y did not occur.”

        By disregarding what would have occurred if P had not X-ed, don’t we get around the puzzle? I am basically claiming that the following as an argument is FALSE:

        (A) Action X causes event Y
        (B) Event Y harms P2
        (C) P2 would have been better off had X not been performed

        I can’t actually think of a case where the first two attain, but the third fails. Except for the one we’re considering =)

  4. sarah says:

    If the child were born with out dd would the girls act still have been immoral?

    • David says:

      Intuitively, I want to say: ‘yes.’ Whether or not the child is actually born with DD, the girl acted recklessly. I just don’t know how to square my intuitive disapproval of the girl’s act with a plausible analysis of ‘harm.’

  5. David says:

    I thought of that too, Sarah, in context of what I said above.

    Considering that, I might be willing to relax my revised EA by dropping the requirement that the harm actually took place. It’s the reasonable anticipation of harm, or that harm is likely, that’s crucial. Surely firing a gun with the intention of killing someone is immoral, whether or not any harm results.

    Another analogy I thought of involves a landlord who does work on his building with dangerous materials – lead paint, asbestos, etc. If he knows the materials to be hazardous, then using them is immoral regardless of whether or not any actual harm is done to a person.

    Now, suppose 10 years later a family moves in with a 7 year old child and the child gets sick from the materials in the house. Surely the landlord is responsible for the child getting sick, assuming he knew the materials to be hazardous. The fact that the child didn’t exist when the landlord redid the building is irrelevant.

    • David says:

      The puzzle isn’t generated by cases involving harm to people who don’t yet exist, and nor is it generated by cases involving intended harm without any actual harmful effects. Rather, it’s generated by cases wherein the alleged ‘subject of harm’ WOULD NOT HAVE EXISTED had the allegedly harmful action not been performed. So the cases you’ve alluded to are importantly disanalogous.

      • David says:

        Yes, I see that in the cases I’ve alluded to:
        – P2 would have been better off had action X not been performed, and
        – P2 would have been better off had event Y not occurred

        In the mother/child case, only the second attains. So I’m claiming that the first is not necessary. Only the second, along with the fact that action X causes event Y.

      • juan says:

        I think there is another objection to both AA and David #2’s very interesting modified version:

        Take something like chemotherapy:that clearly harms people in important ways, but it’s not true they would have been better off without it.Actually,those people who need it would have been worse without it.But it still harms them.

        • juan says:

          quick addition, to take care of potential grumbling about chemotherapy being too impersonal a process, whereas AA is about persons harming persons, and so on.whatever

          think about the guy in Saw who had to cut off his leg in order to get out of that room and not die.he was harming himself, but he would not have been better off without doing it, cause he would have died.

  6. David says:

    Well, no–according to my intuitions about ‘harm.’ Chemotherapy does not harm a patient who would be worse off without it, and nor does chopping off one’s leg harm a person who would be worse off otherwise (whether we’re imagining a Saw scenario or a gangrenous limb). These things cause the person pain, of course, and have other bad effects, but if the net gain in well-being is positive then they do not ‘harm’ the person. (Compare: the dentist causes me some pain when he removes my infected wisdom teeth, but in doing so he’s not harming me, right?)

  7. juan says:

    To the claim ‘Chemotherapy does not harm a person’ I have nothing better to offer than an increduluous stare.If that is not harm, I don’t know what is.My counterexamples use the intuition that sometimes you need to do harm in order to do good overall.And I take it that a lot of people (probably most) agree that chemo has good effects, but also what would properly be called ‘bad’ or ‘harmful’ effects (apparently you think something can be bad for a person without causing the person harm, since you use the word ‘bad’ too,but do not take it to imply harm.That is strange ,but maybe I’m wrong).
    Of course, I claim no right to the word ‘harm’, and if you want to define it in terms of net well-being, and so exclude these counterexamples by definition, that is ok with me.

    I still think that many people might share my intuitions and feel bad about somebody undergoing chemo,because they know that she is caused to suffer,and I think implied in that attitude is the idea that she is harmed. .But again, if you choose to say ‘That is not really harm’,you can do as you wish wth the word, and we can call the effects of chemo something else instead of ‘harm’.

    We both agree on the facts:the overall situation is better, and there is some stuff in between which we seem to agree is bad, but we disagree whether to call that ‘harm’.Since this seems to be only verbal, I don’t see that AA offers much insight into the real nature of harm.According to me, AA fails, according to you it stands.But there is no real issue here

  8. David says:

    No, the issue is not merely ‘verbal.’ We’re not free to define ‘harm’ however we wish, especially since we want to deploy that concept in moral argument. Let me just say this, which you may or may not find persuasive. My construal of harm has the advantage of being consistent with the use of ‘harm’ in medical contexts. Doctor’s take the Hippocratic Oath, which enjoins: “Do no harm.” They also administer chemotherapy.

Leave a Reply

Your email address will not be published. Required fields are marked *