Perry Hendricks’ original ‘impairment argument’ against abortion relied on ‘the impairment principle’ (TIP): ‘if it is immoral to impair an organism O to the nth degree, then, ceteris paribus, it is immoral to impair O to the n+1 degree.’ Since death is a bigger impairment than fetal alcohol syndrome (FAS), Hendricks reasons that, by TIP, if causing FAS is immoral, then, ceteris paribus, abortion is immoral. Several authors have argued that this conclusion is uninteresting, since the ceteris paribus clause is not satisfied in actual cases of abortion: women have reasons for wanting abortions which do not apply to drinking during pregnancy, so all else is not equal, and the conclusion is irrelevant to the morality of actual abortions. In a recent article in this journal, Hendricks and Bruce Blackshaw try to evade this criticism by replacing TIP with the ‘modified impairment principle’ (MIP): ‘if it is immoral to impair an organism O to the nth degree for reason R, then, provided R continues to hold (or is present), it is immoral to impair O to the n+1 degree.’ MIP allows us to derive the ultima facie wrongness of abortion (not just its ceteris paribus wrongness) because MIP lacks a ceteris paribus clause. But I argue that this lack also renders MIP false: MIP faces counterexamples and implausibly produces genuine moral dilemmas. Since the moral principle on which it relies is false, the modified impairment argument fails. I close by considering what a principle would need to do for the impairment argument to succeed.
- applied and professional ethics
- political philosophy
- public health ethics
Statistics from Altmetric.com
Perry Hendricks and Bruce Blackshaw defend a novel argument against abortion: the ‘impairment argument’.1 2 Hendricks’ original version invoked the impairment principle (TIP):
If it is immoral to impair an organism O to the nth degree, then, ceteris paribus, it is immoral to impair O to the n+1 degree.3
Causing fetal alcohol syndrome (FAS) is immoral, and Hendricks argues that death is a greater impairment than FAS. Given this, he invokes TIP to argue that, ceteris paribus, abortion is also immoral.2 3
However, as several authors have noted, the ceteris paribus clause is not satisfied in the case of FAS and abortion, so this has little practical import.3–6 Women have reasons for getting abortions which do not apply to drinking during pregnancy, so all else is not equal between the two cases, and the conclusion tells us little about the morality of actual abortions.1 In an updated version of the argument, Blackshaw and Hendricks acknowledge this criticism.2 They replace TIP with the modified impairment principle (MIP), which lacks TIP’s ceteris paribus clause:
If it is immoral to impair an organism O to the nth degree for reason R, then, provided R continues to hold (or is present), it is immoral to impair O to the n+1 degree.
They suggest its immoral to inflict FAS on a fetus because it deprives the fetus of a ‘future like ours’, that killing the fetus also deprives it of such a future, and that killing the fetus impairs it more severely than giving it FAS. So, by MIP, abortion is immoral (not just ceteris paribus immoral).2
Here, I argue that Blackshaw and Hendricks’ response fails. The lack of a ceteris paribus clause allows MIP to yield the conclusion that abortion is ultima facie immoral, but it also makes MIP false. Since MIP is false, the modified impairment argument is unsound. I present two arguments against MIP.
The argument from cases
MIP faces counterexamples in any situation where (1) it would be wrong to inflict a given impairment, I, without a sufficiently good reason, but (2) it would be permissible to inflict an impairment I* of a slightly higher degree with a sufficiently good reason and (3) the feature which made inflicting I wrong is still present in I*. Consider:
(A) I flood a room you’re in with gas. The gas permanently hinders your ability to roll your tongue, which you enjoyed doing. I have no particular reason for doing this.
This wrongfully impairs you. (Blackshaw and Hendricks say that ‘to impair an organism is to limit one or more of its abilities’; I limit your ability to roll your tongue.)2 Now consider:
(B) I flood a room you’re in with gas. The gas permanently removes your ability to roll your tongue, which you enjoyed doing. I do this because otherwise the gas would flood into a room with me and several other people who are allergic to it, killing us.
In this case, I act permissibly. (In fact, I might act permissibly even if much less than saving several lives was at stake.) But MIP, together with our judgement in (A), implies that my action in (B) is impermissible. My action in (A) was immoral because I limited your ability to roll your tongue (which you enjoyed doing), I also limit your ability to roll your tongue (which you enjoy doing) in (B),2 and I impair you more severely in (B) (removing, rather than merely hindering, your ability to roll your tongue). Since MIP implies that my action in (B) is immoral, and since my action in (B) is not immoral, (B) is a counterexample to MIP.
TIP avoids this counterexample. The ceteris paribus clause is not satisfied for (A) and (B), since I have stronger reasons for action in (B). Removing the clause in MIP allows one to avoid the criticism of the original impairment argument, but at the cost of facing counterexamples.
The argument from dilemmas
A second argument against MIP is that it leads to the existence of genuine moral dilemmas—situations in which I act wrongly, no matter what I do.7 The existence of genuine moral dilemmas is rejected by most ethicists, and there are numerous arguments against it. For instance, it seems plausible that (1) if I am obligated to do something, then I can do it (‘ought implies can’), and that (2) if I am obligated to do X and I am obligated to do Y, then I am obligated to do both X and Y. But these two claims together rule out the possibility of genuine moral dilemmas.7
MIP will produce genuine moral dilemmas in any situation in which (1) it would be wrong to inflict a given impairment, I, without a sufficiently good reason for doing so, but (2) all of my options involve inflicting an impairment of a higher degree than is involved in I and (3) all of my options contain the feature which made inflicting I wrong. Consider:
(C) I start my trolley and begin driving it down the track. I see a person in front of me. If I continue driving straight, I’ll hit them, severing their right leg. I can turn the trolley onto a side track. If I turn, nothing bad will happen.
Here, if I continue going straight, I wrongfully impair the person I hit. I limit their ability to walk (among other things), and have no good reason for doing so. But consider:
(D) I start my trolley and begin driving it down the track. I see a person in front of me. If I continue driving straight, I’ll hit them, severing both of their legs. I can turn the trolley onto a side track. If I turn, the track will loop back around and I’ll hit them anyway, still severing both of their legs.
Together with our judgement in (C), MIP implies that (D) is a genuine moral dilemma. Going straight in (C) was immoral because I removed the person’s right leg, limiting their ability to walk. I do the same thing if I go straight, and I impair them more severely than in (C) by going straight (since I remove both legs). So, by MIP, going straight is immoral. But the same things are true of turning. So, by MIP, turning is immoral. So going straight is immoral and turning is immoral. But those are my only options. So all my options are immoral. So I am in a genuine moral dilemma.3
Again, TIP avoids this. The ceteris paribus clause is not satisfied between (C) and (D), since I have a reason for not turning (namely, that I will hit the person anyway) which does not apply to not turning in (C). Removing the clause in MIP allows one to avoid the criticisms of the original impairment argument, but at the cost of producing genuine moral dilemmas.
Due to its ceteris paribus clause, TIP was sensitive to the fact that the morality of impairing an organism depends, not just on the impairment, but also on the reasons one has for inflicting the impairment. This also rendered it incapable of generating a conclusion about actual abortions, since all is not equal between abortion and causing FAS. MIP is not sensitive to this. This means it can generate the conclusion that actual abortions are wrong, but it also means that MIP is false. It causes MIP to render implausible judgments and to incorrectly generate genuine moral dilemmas.
I suggest that the principle which the defender of the impairment argument is looking for needs to balance two things. On the one hand, it must recognise the fact that, in some cases where inflicting an impairment is wrong, inflicting a relevantly similar, more severe impairment is permissible, if one has stronger reasons for inflicting the latter. On the other hand, it will allow us to reason from the wrongness of inflicting FAS to the wrongness of actual abortions, even though women have stronger reasons for getting an abortion. I await attempts to develop such a principle.
I am grateful to Bruce Blackshaw and Philip Swenson for discussion of relevant issues.
Contributors I am the sole author of the paper.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
↵This is one of two major lines of possible criticism of the impairment argument of which I am aware. The other focuses on Hendricks’ handling of the ‘fate objection’.1 The impairment argument grants (for purposes of argument) that the fetus is not a person. But this apparently leaves it open to the objection that causing FAS is wrong because it impairs a person (later on, once the fetus develops into one), whereas abortion is permissible because it only prevents a person from coming to exist. I defend this objection in a (currently unpublished) manuscript.
↵Of course, that I limit your ability to roll your tongue makes the action wrong in (A), and not in (B). But in stating MIP, Blackshaw and Hendricks cannot intend that ‘R continues to hold (or is present)’ only if R makes the action wrong in both cases. That would make the modified impairment argument question-begging: we could know that MIP applies to the FAS/abortion pairing only if we first knew that the fact that it deprives the fetus of a future like ours makes abortion wrong. But that abortion is wrong is what the argument is trying to show.
↵A proponent of MIP might claim that keeping the trolley straight is an omission, not an action.They might say that MIP applies only to causing impairment, not allowing impairment, and that going straight merely allows the trolley to do what it would do anyway. MIP would then be irrelevant to both (C) and (D). But notice that, unlike in ordinary trolley cases where I am a bystander, in (C) and (D), I started the trolley. I am causally responsible for the fact that it’s heading towards the person in the first place, so if I let it go, I will have caused their impairment. We could also alter the case so that, to keep the trolley going straight, I have to actually take an action—I must manipulate the steering wheel, say. However, if I do nothing, the trolley will derail, killing hundreds of passengers. Then I have three options: (1) do nothing, (2) go straight and (3) turn. Surely it’s wrong to do (1), killing hundreds of passengers. But MIP implies that (2) and (3) are impermissible. So every option is impermissible. So I am in a genuine moral dilemma.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.