1 Introduction

There have been a wide array of criteria proposed by philosophers, legal theorists and political activists to answer the question—when, if ever, is abortion permissible? Entering into this moral and political fray means one must generally be prepared to answer extremely thorny questions about the ethics of killing, autonomy and the development of moral status. In this article, I first defend the claim that abortion cannot be wrong until the fetus is conscious (has experiences of any kind) because until that point the act has no victim. In Sects. 2 and 3 I argue for a view of personal identity and Derek Parfit’s “Relation R” based upon psychological criteria in which the continuity of consciousness defines numerical identity and the other forms of psychological connectedness/continuity define qualitative identity. Coupled with neurological evidence showing that fetuses cannot be conscious before at least 22 weeks of pregnancy, I conclude that abortion is permissible at least until that point in gestation. In Sect. 4 I consider and reject alternatives to my “Victim Requirement” for the impermissibility of killing including the “valuable future like ours” formulation, comparisons to Fetal Alcohol Syndrome and consequentialist objections. Finally, in Sect. 5, I discuss the role of autonomy and consent and the implications of my view in a world where fetal consciousness began much earlier.

My positive thesis is the claim that a fetus cannot gain moral status at least until it develops consciousness (experiences of any kind)—the best neuroscience places this somewhere between 22 weeks and birth (Burgess and Tawia 1996; Lagercrantz 2014). Here I am appealing to the following intuition:

The Victim Requirement (VR).Footnote 1

The act of killing is wrong only if it has a victim. A victim in this sense is (1) an identifiable being which (2) is, temporally speaking, somewhere in the midst of a chain of psychological connections in the sense of Derek Parfit’s “Relation R”—“psychological connectedness and/or continuity,Footnote 2 with the right kind of cause” (Parfit 1986).

2 Teletransportation and the Corporeal Criterion

2.1 Teletransportation

It is often claimed in the abortion debate that the entire question turns on the development of “personhood.” I appeal instead to what Derek Parfit calls “Relation R.” Consider what Parfit calls “Simple Teletransportation”—“The Scanner here on Earth will destroy my brain and body, while recording the exact states of all of my cells. It will then transmit this information by radio… This will then create, out of new matter, a brain and body exactly like mine. It will be in this body that I shall wake up” (Parfit 1986). If we have the intuition that killing the replicated body is just as bad for the person as killing the original body before teletransportation, and that there is nothing morally bad about destroying the original body after teletransportation, then we cannot appeal merely to facts about the body to explain the badness or wrongness of killing human beings. What matters to us in survival then, if one shares this intuition, is something beyond the mere survival of the body. We care about the continuity of consciousness, memory and personality.

This seems to refute:

The Corporeal Criterion.

What makes the infliction of death wrong is damage to some physical body B.

Some may object: “But I do not share this intuition. I believe that teletransportation is a form of murder and that it is just as wrong to inflict it upon someone as it is to kill them.” It is difficult to make ethical arguments to people whose intuitions radically differ from one’s own, but I doubt the sincerity of people who claim to hold this belief. If such people were offered the choice between (1) Simple Teletransportation and (2) Ordinary Death I suspect almost all would choose (1). They would do so precisely because in the case of (1) they would wake up on Mars whereas in the case of (2) they would never wake up again.

2.2 The Materialist Objection

Some materialists may resist Simple Teletransportation. They would believe that what it describes is not possible since consciousness just is or supervenes on the holding of particular physical facts. One of these may be the numerical identity of the material in a human brain. So it is not possible to preserve one and the same stream of consciousness across even microphysically identical but numerically distinct brains. To fully respond to this objection would be beyond the scope of an article on abortion, but I will make some brief remarks. It seems to me that whatever the outcome of the materialist/dualist debate may be, for the purposes of invoking ethical intuitions, speaking of Simple Teletransportation as if it were possible is good enough and does not commit me to the truth of dualism. But for those materialist hardliners who claim that the intension of “consciousness” is necessarily physical this will not do since my claim is semantically incoherent. For such objectors I offer the analogous case of:

Brain Death.

Imagine two potential acts: (1) a lethal dose of morphine is injected into the heart of a brain dead patient—a sufficient amount of brain tissue has died to permanently destroy consciousness (2) a lethal dose of morphine is injected into the heart of the same person with a healthy brain

Are (1) and (2) equivalently bad? Even for those who believe (1) is somewhat wrong it is unlikely they find it to be just as wrong as (2). It also seems that if (1) is wrong it cannot be because of altruistic concerns for the patient. On some preference theories of wellbeing, however, (1) is perhaps bad for the patient. If the patient previously formed a desire to be kept alive in such cases (declared, perhaps, in an advanced directive health plan) then it may be wrong to kill them. But, I do not see how we can explain this wrongness by recourse to the harm we thereby do to the patient. This is because I doubt the preference theory of wellbeing. Again, I do not have the space to present a full argument on this point but consider the following cases:

The Wretched Man and the Spectrum Of Preference

  1. (1)

    John is an exceptionally unlucky man. He has nothing that he wants out of life, including happiness. He is not only psychologically and physically miserable, but has no wealth, no meaningful relationships, no meaningful degree of physical health or fitness, and so on.

  2. (2)

    John is granted, by degrees, one or another of these things that he desires until he has them all, and by any plausible account of preference satisfaction has everything a rational person would desire, except for physical and psychological wellbeing. He is still miserable.

  3. (3)

    The John from (2) is granted, by degrees, progressively more physical and psychological wellbeing so that his psychological and physical pain is gradually erased and eventually replaced with pleasure.

It is not obvious to me that there is a meaningful difference between the wellbeing of John in (1) and (2). But there is clearly a difference between (2) and (3). If there is a difference between (1) and (2) it doesn’t seem to me that it is nearly as great as the difference between (2) and (3). We might also consider a fourth scenario:

  1. (4)

    The John from (3) suddenly suffers a series of terrible misfortunes that reduce him back to the state of John from (1); however, they do not affect his mental state.

Is (4) worse than (3)? In other words, if one is equivalently happy with one’s circumstances, do the circumstances themselves have intrinsic prudential value? I am inclined to answer No. If that is the case then psychological wellbeing—specifically pleasure and pain—are all that make a life go well. Believing this view is, for all practical purposes, indistinguishable from hedonism.Footnote 3

If we are hedonists, then it is clear that the acts described in scenario (1) in Brain Death are far less bad than those described in (2). Most importantly, they are permissible, all else being equal, whereas the acts described in (2) are impermissible. If that is the case, then even on the hardline materialist view, the Victim Requirement, and not the Corporeal Criteron, makes the difference between the permissibility and impermissibility of killing.

3 Consciousness and Personal Identity

3.1 What Is Consciousness?

Here I should pause to clarify some potential points of confusion. Firstly, “consciousness” can be defined in multiple ways, some of which are not what I intend when I use the term. One can be conscious in the sense of being awake, being self-aware or having complex thoughts including the ability to reflect on one’s own mental processes (meta-cognition). The sense of consciousness I mean is best understood using Thomas Nagel’s definition in “What Is It Like To Be A Bat?”—“fundamentally an organism has conscious mental states if and only if there is something that it is like to be that organism—something it is like for the organism” (Nagel, 1974). This view of consciousness goes by many names including “qualia”, “experience”, and so on.

But this distinction has obvious implications for the abortion debate since different forms of consciousness develop at quite different times in early human life. Self-awareness, for instance, may not develop until months or years after birth (Anderson, 1984). But the minimal form of awareness I refer to—the fact that it is “something it is like” to be a fetus—probably develops significantly earlier, somewhere between the 22nd week of gestation and birth (Burgess and Tawia 1996; Lagercrantz 2014). To be clear, I do not intend the Victim Requirement to permit infantacide up until the development of self awareness—months or even years after birth. The sense of “psychological continuity” I am using here relies primarily on the development of minimal awareness, or the fact that it is like anything to be an organism.

3.2 Relation R and Personal Identity

This view is not precisely what Parfit meant by Relation R, and the way I use this concept differs subtly, but importantly,Footnote 4 from his. Parfit believed that two beings are R-related if they share the following relationships: “Psychological connectedness is the holding of particular direct psychological connections. Psychological continuity is the holding of overlapping chains of strong connectedness” (Parfit 1986). It does appear to me that these connections can determine the strength of our reasons to care about some being in the future in the strong sense of self-interest. However, it also appears to me that Parfit misses a necessary condition of those reasons:

The continuity of consciousness: the fact that experiences in fact follow each other in a temporal chain from time T to time T + n.

Parfit seems to worry that it is impossible to describe such a thing without referring to some dubious metaphysical entity like a soul or cartesian ego. I doubt this view. We need not even refer to the bearer of some experience to make the claim that in fact that experience follows another. There does seem to be something potentially improper about this definition in that it makes no distinction between the experience of person P at time T and person P1 at time T. But, in the sense that I mean “follow” we can distinguish these two streams of experience without reference to persons P or P1.

When I experience eating an apple, the feeling of biting into it follows the feeling of picking it up in a way that it wouldn’t follow the feeling of another person, at the exact moment I picked up the apple, tripping down a flight of stairs. These are a series of temporally connected mental events. If I “switched bodies” with my unfortunate counterpart, I would experience tripping down the stairs after walking up to them but I could not experience tripping down the stairs after picking up the apple. With the concept of “the continuity of consciousness” we can disentangle many of Parfit’s personal identity puzzles.

3.2.1 Numerical and Qualitative Identity

It seems to me that the continuity of consciousness is the criterion defining the numerical identity of persons whereas the other forms of psychological continuity and connectedness define the degrees of qualitative identity:

There are two kinds of sameness, or identity. I and my Replica are qualitatively identical, or exactly alike. But we may not be numerically identical, or one and the same person. Similarly, two white billiard balls are not numerically but may be qualitatively identical. If I paint one of these balls red, it will cease to be qualitatively identical with itself as it was. But the red ball that I later see and the white ball that I painted red are numerically identical. They are one and the same ball (Parfit 1986)

I claim that two beings are numerically identical if they are R-related in the sense of sharing a continuity of consciousness and they are qualitatively identical to the degree that they share the other forms of psychological connectedness and continuity.

To illustrate this point consider the following examples from Reasons And Persons:

  1. (1)

    William’s Example I am the prisoner of some callous neuro-surgeon, who intends to disrupt my psychological continuity by tampering with my brain. I shall be conscious while he operates, and in pain. I therefore dread what is coming. The surgeon tells me that, while I am in pain, he will do several things. He will first activate some neurodes that will give me amnesia. I shall suddenly lose all of my memories of my life up to the start of my pain. Does this give me less reason to dread what is coming? (Parfit 1986)

  2. (2)

    The Branch Line Case Several years pass, during which I am often Teletransported. I am now back in the cubicle, ready for another trip to Mars. But this time, when I press the green button, I do not lose consciousness. There is a whirring sound, then silence. I leave the cubicle, and say to the attendant: ‘It’s not working. What did I do wrong?’… Someone politely coughs, a white-coated man who asks to speak to me in private. We go to his office, where he tells me to sit down, and pauses. Then he says: ‘I'm afraid that we're having problems with the New Scanner. It records your blueprint just as accurately, as you will see when you talk to yourself on Mars. But it seems to be damaging the cardiac systems which it scans. Judging from the results so far, though you will be quite healthy on Mars, here on Earth you must expect cardiac failure within the next few days’ (Parfit 1986)

What we can learn from William’s Example is that the continuity of consciousness provides us sufficient reasons to care about some person in the future in the sense of rational self-interest. It indeed makes sense for one to dread the pain the neurosurgeon shall inflict upon him even if the person who experiences it is not qualitatively identical to himself in the present. What we can learn from the Branch Line Case is that qualitative identity is not sufficient for rational self-interest, and that numerical identity is necessary for rational self interest. It does not seem to me that the fact that my replica will continue to live after I am gone (despite his qualitative identity with myself) gives me any less reason to fear my impending demise. It also gives me reason not to care about what happens to him in the sense of rational self-interest. I may regret his death, but not fear it in the sense that I believe I shall die if he does. This is because he and I are not one and the same person.

Similarly, in William’s Example, he goes on to explain that the neurosurgeon will replace my memories and personality with those of Napoleon. Despite the fact that this person and I will no longer be qualitatively the same, we will nonetheless be one and the same person in the sense of numerical identity. The fact that I dread the pain he shall experience is evidence of this.

Qualitative identity may make a difference in other important ways and may allow us to make claims such as “He’s a different person when he drinks,” and these claims may have a bearing on the degree to which we can be self-interestedly attached to ourselves in these moments. If I now know that after consuming my next shot of whisky I will become dangerous to my wife and child, I have less reason to care about what happens to me in that state. I may even hope to be injured and incapacitated. The difference here seems to be explained by the radical disjoint in my goals and interests, and therefore in my qualitative psychological identity. Before the shot of whisky, I love my wife and would never injure her. Afterwards, I lose control of my emotions and become a violent menace. Though I may even despise myself in that state and wish harm upon myself, I still have reasons to fear my death while drunk. On Parfit’s account, this is difficult to explain since that person was not determinately me.

The fact, rather than the degree, of self-interest and of what matters in survival holding for some future person in relation to ourselves ultimately has to do with the continuity of consciousness. If the continuity of consciousness is necessary and sufficient for numerical identity then in its absence there is no way to identify the entity in a morally relevant way.

3.3 Degrees of Identity

One place where my view and Parfit’s align is in the assertion that there is a spectrum of identity, and therefore of rational self-interest in survival depending upon the holding of certain psychological connections. On one end of the spectrum is killing a bacterium. There is nothing wrong, in the general case, with killing a bacterium because it is not the kind of thing that can be a victim. I do not claim that this act of killing is not wrong because literally nothing is killed—merely that the bacterium cannot be the victim, in any morally relevant sense, of the act of killing. That is you have killed something but not someone. We have seen that what matters in survival, and what defines sameness across time, is the continuity of consciousness.

Towards the other end of the spectrum is ending the life of a sleeping or reversibly comatose person. This example is commonly invoked as a challenge for any account of the wrongness of killing that does not include early term fetuses/embryos. However, such people do satisfy the Victim Requirement. Once awakened, these people will continue to be the same locus of conscious experiences—they will bear Relation R to their past selves, the temporal chainFootnote 5 of which extends over the period of their unconsciousness. Importantly, they will be psychologically very similar upon waking up. In fact, on my view there may be something worse about killing them than if the extended unconsciousness had not occurred. Fewer psychological changes have resulted in the intervening time and so there is a minimal compromise of qualitative identity which defines the degree of self-interest and therefore the wrongness of killing.

There are various ways to cash out the wrongness of killing them—on consequentialist views we fail to maximize the good, on deontological ones we violate constraints on our actions, etc.—however, VR is meant only to be one necessary condition for that wrongness. I do not need to give a full account of the wrongness of killing to show that for much of a pregnancy abortion fails to satisfy one of its necessary conditions.

Some may object that my view is not plausible on this point. They may ask me to consider:

Momentary Consciousness.

At some point in a pregnancy, quite early on, the fetus develops consciousness for one second before falling back into unconsciousness until week 22.

Can this really make a difference to the morality of killing the fetus? I answer Yes, but only to a limited extent. Recall that VR is phrased as a necessary condition for the wrongness of killing. Therefore, to claim that VR commits me to the view that it would be wrong to abort the fetus described in Momentary Consciousness is to commit the fallacy of affirming the consequent:

Abortion in this case, while satisfying the Victim Requirement, may still be permissible because, even though the fetus can still be the victim of the act of killing, it bears an exceptionally weak R relation to its potential future self. Recall that on my view the degree of psychological connectedness and continuity establishes the degree of rational self-interest involved in survival.

But, my detractors may persist, what about the case of adults with exceptionally weak R relations to their future or past selves? Is it permissible, on your view, to kill people with Alzheimer’s disease or severe amnesia? My answer is a resounding No. The strength of the R-relation in such cases may be weaker than it is for adults without severe cognitive impairments but it is still far stronger than in the case of Momentary Consciousness. Even with profound disturbances in one’s memory or personality there are some forms of psychological connectedness and/or continuity. One’s memory is not entirely destroyed, and neither is one’s skillset, muscle memory and so on. So there is someone who is deprived of future life in this case whereas in the case of the early term fetus, the identity conditions are not met. It is conceivable to imagine an adult who experiences life in the form of Momentary Consciousness, with absolutely no continuity of psychological identity, though such cases must be exceptionally rare. It is possible that in a case like that euthenasia may be permissible but determining the strength of psychological identity that remains is a nearly impossible task and prudence dictates caution in such cases.

This is related to another potential criticism of my view. Since consciousness is notoriously difficult to identify how can it be used as a morally relevant criterion? How do we know for certain that a fetus is conscious? This criticism, technically speaking, is a non-sequitir since it does not address the moral content of my argument. However, it is an important question. Answering it with any degree of certainty would require solving the Other Minds Problem. If, however, we are willing to employ the same standard of proof we use when considering whether adults are conscious it seems to me that the neurological evidence underpinning my earlier assertion that consciousness develops between the 22nd week and birth is relatively sound (Burgess and Tawia 1996; Lagercrantz 2014).

4 Alternatives to the Victim Requirement

Some writers wish to claim that the wrongness of abortion can be demonstrated without appealing to facts about the psychological features of the fetus. I shall argue these views are either mistaken or deeply problematic.

4.1 “A Valuable Future like Ours”

Don Marquis defends the thesis that “The category that is morally central to this analysis is the category of having a valuable future like ours” (Marquis 1989). Marquis denies condition (2) in my account of victimhood. He therefore claims that in some cases a being can be the victim of an act even if Relation R is not terminated. It is not clear then how he would resist the claim that Simple Teletransportation is as morally serious as murder. In the counterfactual where there was never teletransportation, the original body would have gone on to have a valuable future like ours. Now it cannot. Much, of course, rests on how we understand “future” in Marquis’ formulation. To be fully charitable to Marquis we may wish to grant that the continuity of consciousness counts as the preservation of a future like ours. But Marquis makes it plain that what makes killing bad is the fact that we deprive some being of a valuable future. To claim that the fetus suffers a deprivation of its future, Marquis must establish that it is the same being as the one which would have that future.

Marquis, and others like him, argue that a being can be deprived of a valuable future even without present mentation. They argue that the being is harmed by this deprivation despite having no mental states. I have already granted that this type of harm is possible. For instance, a comatose person is harmed by their death. However, I would deny that there is anyone who is harmed by an abortion.

Suppose that we grant, arguendo, Marquis’ claim. A zygote, then, is harmed by abortion even though it has no psychological characteristics. This claim can be cashed out in one of three ways (assuming one subscribes to one of the major theories of wellbeing): on hedonist theories of wellbeing, on preference theories of wellbeing and on objective list theories of wellbeing.

In every case, Marquis’ claim runs up against a dilemma—either: (1) we were never zygotes or (2) the “Physical Criterion” of identity is true. If we were never zygotes, then death would not have been a harm to us. I cannot be deprived of someone else’s prudential goods. It is a basic element of harm that I am the one who suffers it. Thus, if the identity relation does not hold, there can be no harm. If a zygote is killed and it is not the same being as the one which would have gone on to be happy, have satisfied preferences or obtain objective list goods then it is not deprived of its prudential goods.

The response to this claim would of course be that what goods accrue to zygote Z at some point in the future has no bearing on its present wellbeing. Killing it would deprive it, in the present, of an interest or objective good it presently has in continuing to live. Though I have argued for a hedonist account of wellbeing, this response fails even if one accepts an alternative view. The problem with this argument is that it does not take into account the force of my original objection. Z cannot be deprived, even in the present, of someone else’s future.

If Marquis takes the other horn of the dilemma, he is forced to accept some bizarre conclusions. If Z is the same being as the one which would have resulted from its maturation, it must be by virtue of the Physical Criterion (for it has no present psychological states). The Physical Criterion is the claim that a being persists across time by virtue of the holding of physical facts about its body. This implies accepting that teletransportation is a form of murder since it disrupts physical continuity.

Marquis may object that the relevant physical criterion is genetic. Thus, I am my Replica in Simple Teletransportation and Z is the being that would result from its maturation. The genetic version of the Physical Criterion, however, implies that I am my identical twin. Any variation of the Physical Criterion yields similar results. Any version of the Physical Criterion fails to make sense of the Branch Line Case for it implies that I am my Replica even though there are two of us.

There is one promising version of the Physical Criterion, but it collapses into Relation R. It is this:

The Numerical Identity Physical Criterion.

P persists across time in virtue of the numerical identity of enough brain matter to support the same stream of conscious experiences.Footnote 6

The NIPC makes sense of the Branch-Line Case. My Replica’s brain is not numerically identical to mine. However, it is easy to see how this criterion is, for all intents and purposes, the same as Relation R. Some materialsts would argue that there are complex modal and semantic reasons why the two are distinct. I take no position on this claim. It is enough to note that there is no morally significant difference between the claims of materialists and dualists on this point insofar as the continuity of consciousness is what matters. Most importantly, the NIPC does not yield Marquis’ desired result. It establishes only that a fetus can be deprived of a valuable future like ours at the point when it develops consciousness, which is essentially the Victim Requirement.

4.2 Comparison to Fetal Alcohol Syndrome

Perry Hendricks argues that because it is immoral to give a fetus fetal alcohol syndrome, it must also be immoral to abort it. This argument fails for reasons similar to Marquis’. Hendricks claims, roughly, that if it is wrong to give a fetus FAS then, a fortiori, it is wrong to abort it. Using condition (2) in my definition of victimhood, I claim that in Hendricks’ motivating Impairment Principle the ceteris paribus clause is violated in the case of abortion. TIP is stated as: “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” (Hendricks 2018). The trouble here is that O is not the sort of being that can be a victim. The wrongmaking feature of alcohol consumption during pregnancy is the fact that O will one day be infant I which can claim psychological continuity under relation R. It is immoral to impair O iff one takes a substantial risk that it shall one day be infant I. No such risk is undertaken in the abortion procedure.

Now, Hendricks raises an interesting objection to my view:

To give a fetus FAS is immoral at the time of the consumption of alcohol. The reasoning for this is simple. Suppose that we stumbled upon a person who is 8 months pregnant. Suppose further that we witness her polishing off a bottle of liquor. In that moment (so I claim), the pregnant person has acted immorally—her action did not magically become immoral at some later time. To illustrate this, suppose that the pregnant person was run over by a car 5 minutes after finishing the bottle of liquor, and that this kills both her and her fetus. This tragic situation does not magically erase the immorality of her finishing off a bottle of liquor 5 minutes prior; she is not morally lucky for having been run over (Hendricks 2018)

This raises interesting questions about the modal status of moral claims. Can the moral status of an act depend upon what will happen in the future or what has happened in the past? It seems to me that we do not need to appeal to the controversial concept of moral luck to explain Hendricks’ hypothetical, nor do we need to rely on the metaphysically contentious claim that some act A (alcohol consumption) is not immoral at time T because of what happens at time T + n (when the woman is run over by the car). We can appeal to the epistemic concept of recklessness. The pregnant woman in Hendricks’ scenario has acted immorally when consuming alcohol because of what is likely to happen based on her knowledge at time T. Despite the fact that at T + n the consequences of A are negligible, the agent still took a substantial risk that the consequences of A would affect a future child. Abortion, however, does not do so since, unless the procedure fails, which is exceptionally rare,Footnote 7 there is no risk that a future person will suffer the consequences of the procedure. Thus, the ceteris paribus clause in TIP is violated.

4.3 The Consequentialist Objection

Some consequentialists would reject the Victim Requirement. They would respond to my discussion so far:

The Consequentialist Objection.

I share your intuitions about Teletransportation, Brain Death, William’s Example and The Branch Line Case, but I reject the Victim Requirement. On my view, what makes an act bad is the impersonal fact that it fails to maximize the total amount of wellbeing. In most cases, any fetus which was not aborted would go on to have a life which would be worth living. Therefore, killing it does not maximize the good and on my view is impermissible even without present mentation.

Consequentialists of this sort may hold a wide range of views on the correct theory of population ethics. Parfit famously pointed out that such consequentialists are committed to:

The Repugnant Conclusion.

For any possible population of at least ten billion people, all with a very high quality of life, there must be some much larger imaginable population whose existence, if other things are equal, would be better, even though its members have lives that are barely worth living (Parfit 1986).

Avoiding the Repugnant Conclusion is an infamously difficult unresolved issue in moral theory. Parfit calls the theory that could do so without implying equally unintuitive results “Theory X.” The literature on the search for Theory X suggests that no adequate candidate is forthcoming.Footnote 8

Supposing, for the sake of argument, that there were such a theory, it seems unlikely that a consistent consequentialist could accept it and also believe that abortion is impermissible. The cost of raising a child is substantial, both financially and temporally, and this represents a massive opportunity cost for people who could potentially do more good elsewhere. So even if we grant that Theory X gives us the axiological preference for a world with one more child, that does not establish that abortion is impermissible if it allows the woman the opportunity to do even more good.

I would grant, however, that under some conditions abortion may be impermissible on the consequentialist view. In cases where there is genuinely no greater good available for the agent than having a child, she is perhaps obligated to carry it to term. But such circumstances are likely to be quite rare. It would have to be the case that there is simply nothing that the woman could do with the saved time and resources that would outstrip the impersonal good of adding one additional life worth living into the world.

It seems to me, though, that this is a more compelling argument against consequentialism than it is a good argument against abortion. While it might be a very altruistic thing for such a woman to choose to carry her child to term I do not see a compelling reason why it is morally obligatory.

The Saintly Rock Pusher.

Suppose that it were possible to save one human life a day by pushing very heavy rocks across a courtyard for a period of twelve hours.

On the consequentialist view, it would be impermissible to decline to drag the rocks, so long as doing so wouldn’t compromise one’s ability to continue doing the most good. But most would balk at making it morally incumbent for him to do so. The Saintly Rock Pusher might be considered a kind of moral hero for keeping up this work but it does not seem as if he has a moral obligation to do so. Even for those who accept that he has this obligation, there is some point at which this result becomes deeply counterintuitive. If we do not believe that the Rock Pusher is obligated to continue his work for eight months to save a single life, then, a fortiori, a woman cannot be obligated to gestate for nine months simply because of the good she will do for some as-yet unidentified person.

5 Autonomy and Consent

I have neglected perhaps the most challenging and contentious issue raised by abortion until now. This is a special type of objection to VR in that those who make arguments on the grounds of autonomy and the right of consent typically agree with my conclusion that abortion ought to be permissible in most cases, but they disagree about why.

Strictly speaking, my view as stated so far is compatible with even the most extreme pro-choice positions on abortion. I argue that abortion is permissible at least until the development of consciousness.

In reality, however, I do not hold this view, and in fact my view would be quite different if the facts of human development were not what they are. If it were the case that, from conception, the unborn were mentally equivalent to their actual mental state at week 22 of gestation, I would be opposed to abortion in most cases.

Perhaps the strongest and most well known criticism of this view is given by Judith Jarvis Thomson in her “famous violinist” example:

You wake up in the morning and find yourself back to back in bed with an unconscious violinist. A famous unconscious violinist. He has been found to have a fatal kidney ailment, and the Society of Music Lovers has canvassed all the available medical records and found that you alone have the right blood type to help. They have therefore kidnapped you, and last night the violinist's circulatory system was plugged into yours, so that your kidneys can be used to extract poisons from his blood as well as your own. The director of the hospital now tells you, "Look, we're sorry the Society of Music Lovers did this to you-we would never have permitted it if we had known. But still, they did it, and the violinist now is plugged into you. To unplug you would be to kill him. But never mind, it's only for nine months. By then he will have recovered from his ailment, and can safely be unplugged from you." Is it morally incumbent on you to accede to this situation? (Thomson 1976)

This seems to militate against the conclusion that establishing the moral status of a fetus implies anything about the impermissibility of killing it. However, there are a number of confounding factors involved in this case and I am doubtful that our intuitions on it can be reasonably generalized. The issue of consent, for instance, seems critical. In this case you are kidnapped whereas in the case of pregnancy resulting from consensual sex the analogy is not so clear. If we lived in a world where it was generally known that from conception an embryo was fully conscious and aware, consensual sex would mean taking on the risk of being in a situation where to terminate the pregnancy is morally problematic. Under such conditions, abortion would be an act of callousness even if the woman regrets the consequences of consensual sex.

Some doubt this claim. Margaret Little, for instance, argues:

If I consent to sexual intercourse and I’m informed of the risk of impregnation… this doesn’t mean I then consent to gestate should I become pregnant. For one thing, it’s the wrong party: to consent to a man for him to have sexual intercourse with me doesn’t mean I consent to the fetus for it to occupy my body… To assume the risk of impregnation is not the same as consenting to gestate rather than abort if I do become pregnant, any more than assuming the risk of lung cancer by smoking means that I consent to surgery rather than palliative care should I get the disease (Little 1999)

But the problematic aspect of abortion in such cases is not that a woman cannot deny consent to use her body in such circumstances. The moral defect comes from acting recklessly. I may deny my consent for the use of my body, time or resources at any time, but doing so becomes problematic if I entered into the arrangement knowing that another’s life may well depend upon my continuing consent.

Some baby birds “imprint” upon a parental figure within the first few days of life. If this process occurs with a human who subsequently abandons the bird, it becomes incapable of establishing a bond with its natural parent and will die. If it were the case that human development followed a similar path, allowing a child to imprint upon you and then subsequently giving him up for adoption would be an act of extreme callousness.

The case of the violinist more closely parallels instances of pregnancy resulting from rape. In such cases, the violation of bodily autonomy is fully the result of forces outside of the woman’s control. But does this make the kind of difference that would allow someone to take an innocent life? The question is not an easy or comfortable one, and I will only note by way of reply that whatever answer one gives probably commits them to a particular stance on the case of the violinist.

6 Conclusion

My Victim Requirement can be used to motivate a general defense of abortion until the development of fetal consciousness. In Sect. 1 I introduced the Victim Requirement as a necessary condition for the wrongness of killing. The act of killing is wrong only if it has a victim in the sense of an identifiable being who is, temporally speaking, in the midst of a chain of psychological connections in the sense of Relation R.

In Sect. 2 I gave the example of teletransportation to refute the “corporeal criterion” of the wrongness of killing. Our intuitions on teletransportation seem to show that merely killing a human body is not sufficient to establish moral blame. I also responded to materialist objections to the coherence of teletransportation.

In Sect. 3 I refined Parfit’s view of Relation R with a special focus on consciousness. I argued using William’s Example and the Branch Line Case that the continuity of consciousness, or the fact that experiences in fact follow one another across time, is both necessary and sufficient for rational self-interest in survival and therefore for the numerical identity of persons. By contrast, qualitative identity is defined by the other kinds of psychological connectedness and imply the strength of rational self-interest in survival.

In Sect. 4 I responded to various alternatives to the Victim Requirement including Don Marquis’ “valuable future like ours” formulation, comparisons to Fetal Alcohol Syndrome and consequentialist objections.

Finally, in Sect. 5, I discussed the issues of autonomy and consent and what difference it would make to my view if fetal consciousness developed much earlier.

Opponents of my view must carry the burden of showing, in non-question begging terms, that killing can be wrong even if the being which is killed was never or will never be conscious. If this cannot be done then we can have no compelling moral or legal interest in preventing abortions until the development of fetal consciousness.