1. Introduction
With few exceptions (Marquis Reference Marquis2008), most positions in the abortion debate implicitly assume that potential personhood is synonymous with counterfactual claims about individuals – or at least most don't explicitly appreciate the difference between the two.Footnote 1 In this article, I challenge this assumption by demonstrating how counterfactual claims are importantly different from claims about potential personhood. There is an important subset of potential persons that, I argue, have the same moral status as normal adult human beings but that are importantly distinct from merely potential persons: counterfactual persons. I then argue for the following claim: it is prima facie wrong to kill counterfactual persons, namely, those individuals who, were they not killed, would have been persons.
1.1 An objection to the pro-life position
Consider the following problem that has been brought against many pro-life views: embryo rescue cases (Stretton Reference Stretton2008; Lovering Reference Lovering2013; Räsänen Reference Räsänen2016, Reference Räsänen2020). Imagine that there is a young girl trapped in a building fire. She is unconscious and unable to save herself. She will die a painless death if no one comes to save her. There are also 10 frozen embryos in the building that will be destroyed if not saved. Jones enters the burning building, but he only has the ability to save either the girl or the embryos.
On some popular pro-life theories of the wrongness of killing, it turns out that, at best, there is no compelling moral reason to save the girl over the embryos. The sort of accounts that lead to this result are substance views (Finnis Reference Finnis and Gillon1994; Lee Reference Lee1996; Beckwith Reference Beckwith2004; Eberl and Brown Reference Eberl, Brown and Napier2011; Pruss Reference Pruss and Napier2011; Eberl Reference Eberl and Lizza2014; Hershenov and Hershenov Reference Hershenov and Hershenov2015; Blackshaw Reference Blackshaw2022; Playford Reference Playford, Colgrove, Blackshaw and Rodger2022). These views claim that the moral value of an individual is determined by its membership in a particular kind, typically the kind “rational agents.” In other words, an individual has the same moral status as normal adult humans just in case it is the sort of thing that can exercise rational agency. Beckwith summarizes the view nicely:
According to the substance view, the human being is a particular type of living organism – a rational moral agent – that remains identical to herself as long as she exists, even if she is not presently exhibiting the functions, behaviors, or current ability to immediately engage the activities that we typically attribute to active and mature rational moral agents. Because the human being is a rational moral agent, she is a person of intrinsic moral value as long as she exists. (Beckwith Reference Beckwith and Napier2011, 67)
Given this view, the objection goes roughly like this: if embryos have the same moral status as the unconscious girl, there is no compelling moral reason to save the girl as opposed to the frozen embryos. Yet, it seems clear that one ought to save the girl. I will not canvass the literature on responses to embryo rescue cases; instead, I offer another possible response to the problem.Footnote 2 It is the goal of this article to set up an account of the wrongness of killing that does not automatically entail that killing embryos or allowing them to die is impermissible.
2. Defining counterfactual persons
The thrust of the account is this: the wrong-making feature of killing is that killing deprives one of personhood. The goal of this article is to fill in this simple explanation in explicitly counterfactual terms. Intuitively, what makes it wrong to kill an individual is the fact that she is deprived of personhood. Unlike Marquis (Reference Marquis1989, Reference Marquis2008), the present account does not attribute the wrongness of killing to the value of one's future. Instead, I attempt to provide an abductive argument for the claim that depriving one of personhood is one of the worst things that can happen to her, and that is what makes killing us wrong: were we not killed, we would have been persons.
Now, it might not be clear that preventing an entity from becoming a person who, were she not killed, would have been a person, is wrong.Footnote 3 But there are clear cases in which depriving an individual of something she would have had in the absence of your interference is just as wrong as depriving an individual of something she actually has. To see why, compare the following two scenarios. In the first scenario, imagine that a temporarily comatose acquaintance of mine, Steve, has $6MM. Even though Steve is temporarily comatose, we know that he will wake up within a week. Out of envy, I go and steal Steve's $6MM. Clearly, I have wronged Steve by depriving him of $6MM. Now imagine a second scenario in which temporarily comatose Steve has a lottery ticket and it is true that, were I not to rip up the lottery ticket, Steve would have won $6MM. He gave me his ticket to keep safe before an accident rendered him temporarily comatose. After he became comatose, I find out that he would win the money in the absence of me destroying his ticket. Suppose, also, that I really do not like Steve and make sure to rip up his ticket. It seems clear that I have deprived him of $6MM and that it was wrong of me to do so. In both cases I deprive Steve of something extremely valuable. The only difference is that, in the first case, I deprive him of something valuable because he currently has $6MM; in the second case, I deprive him of something valuable because he would have had $6MM in the absence of my preventing it. In both cases, not only have I done something wrong to Steve by depriving him of something of great value, but the wrong I commit in the first case by depriving him of something he would have had seems just as wrong as depriving him of something that he currently has. The present lottery case, I propose, presents us with at least one case in which depriving an individual who would have had something of great value is just as wrong as depriving an individual who currently has something of equal value.
Before moving on to a more precise definition of what sorts of individuals can be deprived of personhood, I must make two preliminary points. First, the lottery case is not meant to show that all cases of depriving an individual who would have had something of value (in the absence of someone preventing her from obtaining it) are just as wrong as depriving her of something of value that she actually possesses. Rather, the goal of the lottery case is to demonstrate a clear instance in which both sorts of deprivation are equally wrong. It leaves the question open as to what other cases are morally similar to the lottery case. This means that other considerations must come into play when determining whether or not both sorts of deprivation are equally wrong in a given circumstance. One significant consideration will be the explanatory power of a principle concerning the wrongness of a particular type of action; as I will argue in section 4, there is significant explanatory power in supposing that individuals who, were they not killed, would have been persons, are the sorts of beings that are impermissible to kill. Thus, the explanatory power of the account presented herein heavily tips the weights in its favor.
Second, the simple and intuitive account sketched above does not depend on which account of personhood one adopts. When using the term “personhood” I am referring to the descriptive notion of personhood, not a moral notion such as “the right not to be killed unjustly.” Many accounts of personhood (excluding substance views) are compatible with the view sketched above, including accounts which have self-awareness as necessary conditions for personhood (Tooley Reference Tooley1972; Giubilini and Minerva Reference Giubilini and Minerva2013). The argument presented below will still follow given any of these accounts. I need not commit to anything more than this, since nothing that follows will depend on one's descriptive account of personhood.
Having provided a sketch of the account, the next step is to define what sort of being is one which can be deprived of personhood, what I will call a counterfactual person. I define a counterfactual person as follows:
x is a C-person =df. were x not killed, then x would have been a person at a later time.
To make this all a bit more precise, this picture ought to be supplemented by an account of counterfactuals since there must be a fact of the matter as to which counterfactual statements are true. For the purposes of this article, I will utilize Lewis's (Reference Lewis1973) classic account, which requires a brief sketch of the closeness of possible worlds. Take three possible worlds: the actual world, W1, and W2. To say that W1 is closer to the actual world than W2 is to say, roughly, that the history of W1 up to and including the present is more like the actual world than W2 (see Lewis (Reference Lewis1979) for more specific criteria). Given this (all too brief) sketch of the closeness of possible worlds, a counterfactual ϕ > ψ is true for Lewis if the conditional ϕ → ψ is true in every nearest ϕ-permitting possible world. I will use these Lewisian truth conditions to determine the truth value of the counterfactual in the present definition of C-people, although it is certainly possible that other accounts of counterfactuals can also be used to develop the account.
3. Potential persons
Before moving on, however, I must make one more important preliminary remark; I want to distinguish the notion of C-person from that of potential persons. Some authors seem to have in mind what I will refer to as bare potential personhood or B-personhood. The common theme among these authors is that an individual is a potential person just in case it is physically possible for that individual to become a person given some suitable environment (Tooley Reference Tooley1972; Singer and Dawson Reference Singer and Dawson1988; McMahan Reference McMahan2002). McMahan (Reference McMahan2002), for instance, supposes that by “potential” the pro-lifer means that a fetus normally becomes a person given appropriate environmental conditions. So, one might understand B-persons along the following lines:
x is a B-person =df. possibly, if some appropriate stimulus conditions are met, then x is a person at a later time.
where the kind of possibility at issue is physical possibility. The details will be filled out differently by different authors.
What is important for the present purpose is that B-personhood does not imply C-personhood. This is because the fact that it is physically possible for a non-person to become a person does not mean they become a person in all possible worlds closest to the actual. To see why, suppose that there could someday be a pill that changes the psychological characteristics of canines such that they become persons. Take any canine in the actual world and call him Fluffy. Suppose that, in the actual world, Fluffy finds a pill on the ground. Fluffy is a B-person but not a C-person. Even though, in the actual world, this pill he finds on the ground is not a people-pill, there is some world, very remote from our own, in which the pill that falls on the ground is a people-pill. Thus, in a world quite unlike the actual, Fluffy eats the pill and turns into a person. So, Fluffy is a B-person because the following material conditional is true: it is possible that, if some appropriate stimulus conditions are met, then Fluffy is a person at a later time.Footnote 4
Fair enough, Fluffy is a person in some world very much unlike our own. However, Fluffy is not a C-person. In order for Fluffy to be a C-person, he must become a person in all the worlds closest to the actual in which he is not killed. It simply isn't true, however, to say that he is a person in these nearest worlds for the simple reason that we do not have this imaginary people-pill. We don't have the technology to create such a pill, and the world would have to be very different from its current state in order for such a pill to exist. So, it is not true that Fluffy is a C-person because the following counterfactual is false: had Fluffy not been killed, then he would be a person at a later time.Footnote 5
4. C-People and the wrongness of killing
Having specified a subset of potential people, I now have the machinery to state a new sufficient condition for the wrongness of killing:
(C) If x is a C-person, then it is prima facie wrong to kill x.
In this section, I will argue that (C) is true by inference to the best explanation. In particular, (C) explains why it is wrong to kill the following individuals: normal newborns, normal adult humans who are temporarily comatose, and normal adult humans. (C), further, does not entail that it is wrong to kill the following individuals: frozen embryos, embryos killed by the morning-after pill, and anencephalic fetuses. I conclude that, because this account matches our intuitions in these cases, there is strong reason to suppose it is a sufficient condition for the wrongness of killing.
4.1. What the account entails
The first step in the inference to the best explanation is to consider what the account entails. In particular, it entails that it is impermissible to kill infants, normal adult humans who are temporarily comatose, and normal adult humans.
First, (C) rules out infanticide whereas other popular accounts of the wrongness of killing fail to do the same (e.g., Tooley Reference Tooley1972; McMahan Reference McMahan2002, Reference McMahan2007; Singer Reference Singer2011; Giubilini and Minerva Reference Giubilini and Minerva2013; Hershenov and Hershenov Reference Hershenov and Hershenov2017; Rodger, Blackshaw, Miller Reference Rodger, Blackshaw and Miller2018).Footnote 6 To see how, imagine that a typical newborn is killed. We then ask whether or not the counterfactual “had the newborn not been killed, then she would be a person at a later time” is true. This counterfactual is true because in all of the closest worlds in which the newborn is not killed, she is a person at a later time. Newborns’ internal developmental tendencies are such that, in the absence of actively ending their lives or depriving them of resources needed to live, they become persons.
(C) also explains why it is wrong to kill the temporarily comatose. In a case much like that presented by Beckwith (Reference Beckwith2004), suppose that your Uncle Jed is temporarily comatose due to injuries sustained in a car accident. We can add that Uncle Jed would wake up in the absence of being killed (i.e., Uncle Jed is not permanently comatose). (C) renders the intuitive result that it is impermissible to kill Uncle Jed. This is because, were he not killed, he would have been a person. Uncle Jed, then, is a C-person.
(C) also explains why it is wrong to kill normal adult humans. It is wrong to kill normal adult humans because, if they are killed by some event or set of events which are sufficient to bring about their death then, in the nearest worlds in which they are not killed, they are persons at a later time. Therefore, since normal adult humans are C-people, it is wrong to kill them.
Here I will briefly pause to consider an objection. One might worry that the explanatory work in these previous two cases outlined above is done by properties actually had by Uncle Jed and normal adult humans. So, maybe it's not wrong to kill them because they're C-persons. Rather, the wrong-making feature of killing these sorts of individuals is the properties they currently possess.Footnote 7
There are at least three ways to respond. First, it is important to note that, if what explains the wrongness of killing in these cases is the fact that Uncle Jed and normal adult humans are persons, one must appeal to some property or some set of properties that delineates what sort of entities it is permissible to kill. But proposals of this sort have led some to the conclusion that infanticide is permissible (e.g., Tooley Reference Tooley1972; McMahan Reference McMahan2002, Reference McMahan2007; Singer Reference Singer2011; Giubilini and Minerva Reference Giubilini and Minerva2013; Hershenov and Hershenov Reference Hershenov and Hershenov2017; Rodger, Blackshaw, Miller Reference Rodger, Blackshaw and Miller2018). So, I'm not inclined to think appealing only to properties individuals actually/presently have will explain why it is wrong to kill newborns; the present account therefore fares better than many alternatives in virtue of avoiding the permissibility of infanticide.
Second, it doesn't seem all that implausible to claim that it is wrong to kill normal adult humans because, were they not killed, they would have been persons; and I think this is plausible because the actual properties of individuals make the relevant counterfactuals true. This is important to note since my account does not do away with the importance of properties that individuals actually possess. It is entirely consistent with the account that properties individuals actually possess are morally relevant. It merely provides an alternative framework for the role of properties individuals actually possess: the relevant counterfactuals are true because of the properties that Uncle Jed and normal adult humans possess. For instance, developmental dispositions that newborns actually possess make it true that, were they not killed, they would have been persons at a later time.
Third, the account only provides one sufficient condition for the wrongness of killing. One might propose another sufficient condition for the wrongness of killing that appeals to actual personhood. While this would make any complete account of the wrongness of killing less parsimonious, we should not assume that a complete account of the necessary and sufficient conditions of the wrongness of killing is bound to be a simple one.Footnote 8
4.2. What the account does not entail
The second step in the inference to the best explanation is to consider what the account does not entail. In particular, it does not entail a prohibition against killing the following individuals: frozen embryos used for IVF, embryos killed by the morning-after pill, and anencephalic fetuses. To see why the account does not entail that it is wrong to kill frozen embryos or let them die, consider the embryo rescue case outlined in section 1. My account does not entail that the frozen embryos have a moral status equal to that of the girl. As a result, my account does not entail that the moral reasons to save the embryos are just as strong as the reasons to save the girl. This is because embryos are not C-people: it's not true that, had the embryos not been killed by the fire, that they would have become persons. Take the set of worlds closest to the actual in which the embryos are not killed in the fire. Among those worlds there are some in which those embryos do become persons, but there are also some just as close in which they do not.
There are at least two reasons why embryos, even in the absence of being killed, do not become persons in these nearby worlds. Take an individual embryo (or group of embryos), A. First, we can imagine that another embryo (or group of embryos), B, was used in the IVF procedure instead of A. As a result, if there are other embryos available, it is most likely not true of any individual embryo that it would become a person at a later time, even if it wasn't killed.
Second, there are nearby worlds in which the embryo does not implant at all. The IVF procedure has a relatively low success rate, where success is a live birth. According to the Society for Assisted Reproductive Technology (2022), around 37% of initial IVF procedures were successful in 2020. For reasons not fully understood, embryos often fail to implant or die shortly thereafter, that is, there is some nearby world in which A fails to implant or dies shortly after implantation. In either case, A either does not implant or dies shortly thereafter and does not become a person.
Before moving on, I must be very explicit in how I am using the percentages of successful IVF procedures. It is important for me to be clear that it is not, strictly speaking, the likelihood of becoming a person that is morally relevant. I assume that there is a fact of the matter as to whether or not a particular counterfactual is true. And I am using the relevant data as one piece of evidence as to whether or not a particular counterfactual is true. These percentages give us evidence that not much needs to go wrong for an embryo to fail to implant or die shortly thereafter. So, what I take the percentages to be showing us in the present example is this: for any successful IVF procedure, in order for it to have failed, very few things about the environment, development and genetic makeup of the embryo, etc. would have needed to be different. In other words, there is at least one nearby world in which it fails to implant or dies shortly thereafter.
A desirable consequence of the view, then, is that my account does not rule out the permissibility of the typical IVF process. During the typical IVF process, many more embryos are made than are used because it is common that more than one procedure is needed for a successful pregnancy to result from the IVF process. So, these embryos are created with the knowledge that many of them will either die or remain frozen indefinitely and, for some, this result is morally problematic (Tollefsen Reference Tollefsen2001). But that is no reason to abandon the typical IVF process. The frozen embryos used in the IVF process are not C-people. It is not true of them that, were they not killed, they would have been persons. As a result, it is not impermissible to create them knowing that many of them will either die or remain frozen indefinitely. Allowing them to die through the IVF process is, therefore, not necessarily impermissible given my account.
The same reasoning that was used to show that IVF embryos are not C-people can be extended to embryos in the womb but not yet implanted (or that die shortly after implantation), such as those killed by the morning-after pill. If it is possible for a perfectly healthy IVF embryo to fail to implant, it is reasonable to think that an embryo conceived via intercourse could, for the same natural reasons, fail to implant or die shortly after implantation. Assume that, as in the last case, these are perfectly healthy embryos. It is not quite clear what the mechanisms are that prevent implantation in the IVF case, and, therefore not quite clear why perfectly healthy embryos conceived via intercourse may fail to implant or die shortly after implantation. Given this uncertainty, I am happy to establish the following conditional claim: if there is some nearby world in which a perfectly healthy embryo, conceived via intercourse, is not killed and does not become a person because it fails to implant or dies shortly after implantation, then that normal embryo in the womb is not a C-person. If this conditional claim is true, then the impermissibility of the morning-after pill is not entailed by my account. Thus, even if it is true that the morning-after pill stops an embryo from implanting, it does not follow, on my account, that it is morally impermissible.
Finally, my account does not entail the impermissibility of aborting an anencephalic fetus. Anencephalic fetuses, if they are able to make it to term, are born without significant portions of the brain. Because of this deformity, these fetuses will never become people. It is not true of these fetuses that, were they not killed, they would have been persons. As a result, they are not C-people. My account does not entail that it is wrong to abort in this case.
Before moving on, it is important to note that there may be other reasons that killing frozen embryos (or allowing them to die during the IVF process), using the morning-after pill, and aborting anencephalic fetuses are impermissible. All I have established is that such killings are not wrong because they're C-people; there are no C-people killed in such cases. In short, I have only provided one sufficient condition for the wrongness of killing that accounts for common intuitions about important cases considered in this section. It is a feature, not a bug, of the account that the permissibility of killing in these last three cases surveyed is not automatically ruled out.
5. A new argument against abortion
This account of the wrongness of killing has clear implications for the abortion debate. The previous sections provide the basis for a novel argument against abortion, the Counterfactual Argument:
(C) If x is a C-person, then it is prima facie wrong to kill x.
(1) Normal fetuses are C-people.
(2) Therefore, it is prima facie wrong to kill normal fetuses.
I have argued that (C) is true by inference to the best explanation; this account is one of the only on offer that entails neither the permissibility of infanticide nor the impermissibility of killing frozen embryos. These desiderata are significant because many accounts of the wrongness of killing fail to accommodate one or the other.Footnote 9 (C) also captures common intuitions about the impermissibility of killing in other cases discussed in section 4. (1) is true by the definition of C-people. Any normal fetus is such that, were it not killed, it would have been a person. (2) follows from (C) and (1).
6. An objection
Yet, someone may propose a modified version of the embryo rescue case. Suppose that, in every nearby world in which embryo A is not killed, it is true that it becomes a person at a later time. For instance, say that the embryo is attached to a machine that will guarantee that the embryo is implanted in an artificial womb the next day. Nothing (in the closest antecedent-permitting worlds) will stop this machine from succeeding, as it is a very well-designed and efficient machine. As a result, it is true of the frozen embryo that, were it not killed, it would have been a person; A is a C-person in this instance. Since A is a C-person, it seems as though my account runs into a similar embryo rescue case objection when it is guaranteed that the frozen embryo becomes a person. Given the account presented, there may be no more of a reason to save the unconscious child than the frozen embryo.Footnote 10
In response, I argue that the main differences between an embryo and a newborn – given both of them are C-persons – are morally irrelevant factors: for example, time, dependence, and stage of development. Hershenov (Reference Hershenov2020) makes the argument that, once triage considerations are accounted for, the intuitive force of the embryo rescue case is diminished. Say, for instance, that the likelihood of survival of the embryo and the child after they are removed from the fire is the same. Say also that the child needs to use another's body to support her kidney function, much like Thomson's (Reference Thomson1971) violinist. Hershenov contends that it's not clear which individual one ought to save. However, Hershenov's view still has the implausible implication that most or all embryos have the same moral status as the child. Even though his view still has this implausible implication, his general strategy can be used to eliminate morally irrelevant features in the modified embryo rescue case outlined above. Thus, a plausible way to respond is as follows: once we eliminate morally irrelevant differences between the child and the embryo, the intuition that one ought to save the child over the embryo loses much of its strength.
Imagine that, instead of a child, one has the choice to save either a frozen embryo or an infant. Suppose, also, that the time which must elapse before either becomes a person is the same; perhaps the newborn is temporarily unconscious for nine months while the embryo develops until they are at the same stage of development. Finally, following Hershenov (Reference Hershenov2020), stipulate that the newborn has a pair of failing kidneys and will need to be attached to someone for nine months. Given all this, both the newborn and the embryo would become persons at the same time. That counterfactual is key: it is true of both that, were they not killed, they would have been persons at a later time.
Faced with this scenario, it is not clear which individual one has an obligation to save. Many may have the intuition that it is permissible to save either. The fact that one is less developed than the other alone is not sufficient to make a moral difference. It is implausible, for instance, to say that an infant has a lower moral status than a toddler simply in virtue of their stages in development. The time difference is also morally irrelevant; both individuals, were they not killed, would become persons around the same time. And removing these morally irrelevant features mitigates the supposed bite of this modified embryo rescue case. There is no moral difference in the status of this very unique embryo and the newborn precisely because they are C-persons. In this world very remote from our own, the high moral status of the frozen embryo is not nearly as implausible as in the original embryo rescue cases. At the very least, my account comes out much better than the substance view because my account does not rule out the permissibility of killing frozen embryos or allowing them to die.
7. Conclusion
What makes killing us wrong is that it deprives us of personhood; were we not killed, we would have been persons. This position has a clear consequence for the abortion debate: the vast majority of abortions are impermissible because most fetuses are counterfactual persons.
Open access
