1. Introduction
Most of us believe that it would be wrong to create someone whose life would be deeply miserable. For instance, let’s say that if you created a child this year, she would be born with a genetic condition that would make her life very short and filled with little else but intense pain. Unless there were something important at stake for you or third parties, it seems it would be morally wrong to create such a child. Most of us also believe that it would not be wrong to fail to create someone whose life would be fairly happy. For instance, let’s say that if you created a child this year, he would go on to experience the sorts of joy and sorrow that, taken together, make for a worthwhile life. Unless there were something important at stake for you or third parties, it seems it would not be morally wrong not to create such a child. Perhaps surprisingly, it has proven difficult to reconcile these two commonsense intuitions about the morality of creation, to show how we can believe both of them without inconsistency. The goal of this article is to show how we can hold onto both of these moral intuitions without contradicting ourselves.
After making these moral intuitions explicit, I will construct an argument that shows why some philosophers have thought that endorsing both of them leads to inconsistency. I will then consider how some philosophers have attempted to resolve this inconsistency, and explain why their solutions are unsatisfactory. I will then argue that the common way of formulating the relevant moral intuitions deploys an overly broad notion of ‘creation,’ which in certain cases generates counterintuitive consequences. I then argue that restricting the sense of ‘creation’ in our formulations of these moral intuitions both eliminates these counterintuitive consequences and defuses the argument that we cannot consistently endorse both intuitions. After responding to some objections, I will conclude by suggesting how resolving this apparent conflict among our moral intuitions has ramifications for broader questions about the ethics of creation.
2. The Procreation Asymmetry
The commonsense moral intuitions mentioned above might be formulated as follows:
Creating Miserable People: The fact that a person would be miserable – i.e. his or her life would be not worth living – gives us a strong moral reason not to create that person.
Creating Happy People: The fact that a person would be happy – i.e. his or her life would be worth living – gives us no moral reason to create that person.Footnote 1
Many people find each of these propositions intuitively plausible, and so are inclined to endorse their conjunction. Jeff McMahan (Reference McMahan1981, 100) dubbed this conjunction ‘the Asymmetry,’ Footnote 2 as the first conjunct tells us that people’s (low) welfare can give us strong moral reasons not to create them, while the second conjunct tells us that people’s (high) welfare can give us no moral reasons to create them. To avoid confusion with other moral asymmetries, I will call the conjunction of the above propositions the Procreation Asymmetry. Before explaining why the Procreation Asymmetry seems problematic, let me clarify exactly what each of its conjuncts means.
First, ‘person,’ ‘happy,’ and ‘miserable’ are terms of art. By ‘person,’ I mean an individual with significant moral status, such that agents can have moral reasons regarding that individual’s welfare and rights. A ‘happy’ person is someone whose life is worth living, and a ‘miserable’ person is someone whose life is not worth living. There is considerable controversy about the necessary and sufficient conditions for a life’s being (not) worth living, and I will not settle that dispute here.Footnote 3 Rather, I will assume McMahan’s (Reference McMahan, Coleman and Morris1998, 215; Reference McMahan, Roberts and Wasserman2009, 50) plausible and well-known view, which is that a life is worth living if and only if the positive value of the intrinsic personal goods in the life (i.e. the things that are good-in-themselves for the person) outweighs the negative value of the intrinsic personal evils in the life (i.e. the things that are bad-in-themselves for the person). Conversely, if the negative value of the evils outweighs the positive value of the goods, then the person’s life is not worth living. In other words, a life worth living is, overall, intrinsically good for the person who lives it, while a life not worth living is, overall, intrinsically bad for the person who lives it.Footnote 4 Creating Miserable People maintains that we have a strong moral reason not to create someone just because his or her life would be overall bad for him or her, while Creating Happy People maintains that we have no moral reason to create someone just because his or her life would be overall good for him or her.
Second, the moral reasons referred to in the Procreation Asymmetry’s conjuncts are deontic or requiring moral reasons, rather than merely commendatory or justifying moral reasons. Deontic or requiring moral reasons are reasons that can serve as the ground of an obligation or moral requirement, such that acting against them can be morally impermissible. Commendatory or justifying moral reasons, however, can render an action morally good or even supererogatory, but do not ground moral requirements.Footnote 5 Creating Happy People, on this interpretation, is consistent with the proposition that the fact that a person would be happy makes it supererogatory or morally praiseworthy to create that person. In this article, however, I take no position on whether creating happy people is in fact supererogatory or morally praiseworthy.Footnote 6
The Procreation Asymmetry seems intuitive to many; each of its conjuncts appears to be not just true, but obviously true.Footnote 7 Despite the Procreation Asymmetry’s wide appeal, philosophers have worried that Creating Miserable People and Creating Happy People commit us not merely to a moral asymmetry, but to some sort of inconsistency. Although this worry has been presented in a variety of ways, I think it is most clearly explained by way of a view called moral actualism. Moral actualism holds that among all actual and possible people, we have moral reasons regarding the rights and welfare only of actual people – those who have existed, do exist, or will exist. In other words, merely possible people do not matter morally: their welfare and their rights give us no moral reasons to act one way or another.Footnote 8 It is plausible that people whose existence depends on our future choices are merely possible people, at least until we have caused those people to exist.Footnote 9 If that is the case, then moral actualism entails that people whose existence depends on our future choices do not matter morally.
Setting aside for the moment the question of whether moral actualism is true, the view provides a basis for an argument that the Procreation Asymmetry commits us to some sort of inconsistency in our moral beliefs.
The Moral Actualism Dilemma
-
(1) We have moral reasons to promote the welfare and respect the rights of all and only those persons who matter morally.
-
(2) Either moral actualism is true or moral actualism is false.
-
(3) If moral actualism is true, then Creating Happy People is true and Creating Miserable People is false.
-
(4) If moral actualism is false, then Creating Miserable People is possibly true and Creating Happy People is false.
-
(5) Therefore, either Creating Happy People is true and Creating Miserable People is false, or Creating Miserable People is possibly true and Creating Happy People is false.
-
(6) Therefore, either Creating Miserable People is false or Creating Happy People is false.
-
(7) Therefore, the Procreation Asymmetry is false.
Premise 1 is, if not an analysis of what it is for a person to matter morally, a highly plausible assumption about the extent of our moral reasons. Premise 2 is trivially true. Premise 5 follows by way of constructive dilemma from premises 2–4. Premise 6 follows by subtraction from premise 5, and the conclusion (7) follows straightforwardly from premise 6. All of the action happens in premises 3 and 4.
Premise 3 says that if moral actualism is true, then Creating Happy People must be true and Creating Miserable People must be false. Creating Happy People denies that the fact that someone would be happy gives us any moral reason to create that person. This appears to follow straightforwardly from moral actualism: if the welfare and rights of merely possible people never give us moral reasons, and people are merely possible before we choose to create them, then the fact that a person’s life would be overall good for him gives us no moral reason to create that person. The trouble is that moral actualism appears to be inconsistent with Creating Miserable People. If the welfare and rights of merely possible people never give us moral reasons, and people are merely possible before we choose to create them, then the fact that a person’s life would be overall bad for her gives us no moral reason not to create that person.
Premise 4 says that if moral actualism is false, then Creating Miserable People might be true and Creating Happy People must be false. Creating Miserable People claims that the fact that a person would be miserable gives us a strong moral reason not to create her. Although not entailed by premise 1, it is plausible that this claim means that we have strong moral reasons to prevent people from having miserable lives; if moral actualism is false, then such a reason would also extend to merely possible people, including those we might yet create. The trouble is that if moral actualism is false and merely possible people do matter morally, then Creating Happy People would be false in virtue of premise 1. If we have moral reasons to promote the welfare and respect the rights of all people who matter morally and it is true that merely possible people matter morally, then we have a moral reason to give those people we might create the overall good of a life worth living.
Thus the Moral Actualism Dilemma seems to show that whether moral actualism is true or false, the Procreation Asymmetry leads to an inconsistency among our moral beliefs. In the next section, I will consider and reject some strategies that have been proposed for dealing with this argument.
3. Attempts to resolve the inconsistency
Philosophers have taken two broad approaches to resolving the sort of inconsistency uncovered by the Moral Actualism Dilemma. One approach accepts the force of the dilemma, and in turn counsels us to modify the moral intuitions behind the Procreation Asymmetry. The other approach aims to resolve the apparent inconsistency by showing one of the dilemma’s premises to be false. The first (and more common) approach is revisionist, in that it maintains we should let go of one of the moral intuitions behind the Procreation Asymmetry. The second approach is preservationist, in that it maintains we can ultimately hold onto those moral intuitions. In this section, I will show that extant versions of revisionism and preservationism are unsatisfactory.
3.1. Revisionism
One revisionist approach is to reject Creating Miserable People, the proposition that the fact that a person would be miserable gives us a strong moral reason not to create that person. Philosophers who favor this strategy for resolving the Procreation Asymmetry’s latent inconsistency typically have independent reasons to endorse moral actualism. Some believe that denying moral actualism commits us to a counterintuitive or theoretically unworkable picture of morality.Footnote 10 Others argue we should accept moral actualism because it follows from modal actualism, the view according to which there are no non-actual entities (e.g. merely possible people).Footnote 11 Premise 3 of the Moral Actualism Dilemma shows that Creating Miserable People is inconsistent with these commitments, which leads philosophers with these commitments to reject Creating Miserable People. The obvious problem with this revisionist approach is that Creating Miserable People is deeply intuitive; it seems to many that rejecting it would be an unacceptably high ransom for moral actualism.
Most philosophers who take a revisionist approach reject Creating Happy People, the proposition that the fact that a person would be happy gives us no moral reason to create that person. These philosophers are not particularly concerned to defend moral actualism, and indeed often have independent reasons for denying it. Jeff McMahan (Reference McMahan, Roberts and Wasserman2009), for instance, argues that what is good or bad for merely possible people, or even what is impersonally good or bad, can give us moral reasons. Rejecting the intuition that we have no moral reasons to create happy people can be made even more palatable if one holds, like Elizabeth Harman (Reference Harman2004, 97–98) does, that our reasons to benefit people are generally weaker than our reasons not to harm them. Harman and others hold that, other factors being equal, failing to advance someone’s welfare by some amount is less bad than diminishing someone’s welfare by that same amount.Footnote 12 If McMahan’s or Harman’s views are correct, this could undermine the intuitive appeal of Creating Happy People: perhaps the fact that a person we might create would have a happy life does give us a moral reason to create that person, but it is a relatively weak reason, easily overridden by considerations that apply in most cases (e.g. the risks of pregnancy, the harms of overpopulation, etc.).Footnote 13
This second revisionist approach might seem attractive, at least at first. If Creating Miserable People is deeply intuitive, the truth of moral actualism is uncertain, and it isn’t all that counterintuitive to deny Creating Happy People, why not bring our moral intuitions in line with premise 4 of the Moral Actualism Dilemma? The problem with this approach is that rejecting Creating Happy People has some significantly counterintuitive consequences. First, on McMahan’s view, there would often be cases in which agents violate their moral obligations by failing to procreate, even if such failure would have no negative effects on the agent or third parties.Footnote 14 Such cases might be rarer if one is prepared to accept Harman’s controversial thesis about the relative weights of harms and benefits, but they would occur nonetheless. Second, as Melinda Roberts has argued (Reference Roberts2011b, 773–774), rejecting Creating Happy People entails that agents would sometimes be morally obligated to create a new person rather than prevent the suffering of an existing person.Footnote 15 Again, such cases would occur less frequently on Harman’s view, but they are almost sure to be realized: provided that the life of the person to be created is sufficiently happy, the benefit of creating that person could outweigh the harms that existing people would otherwise suffer.
Third, rejecting Creating Happy People entails that those who refrain from procreating without considering the happiness of the person they would have created are in some way unvirtuous. It is plausible that conscientious moral agents weigh and consider all of the moral reasons that are both epistemically accessible to them and relevant to the moral status of their decisions.Footnote 16 If that is true and Creating Happy People is false, then most readers of this article have failed to live up to this standard of conscientiousness on multiple occasions. It seems patently obvious, however, that this is no failure of virtue on our part, and that it would be obtuse to criticize our refraining to procreate because we failed to fully consider the happiness of those we might have caused to exist. Indeed, it seems that agents who are conflicted about whether to refrain from procreating in light of such considerations would exhibit a different sort of failure in virtue or rationality, akin to having ‘one thought too many.’Footnote 17
Resolving the Procreation Asymmetry’s latent inconsistency by revising the moral intuitions behind one of its conjuncts has steep costs. The fact that some philosophers’ theoretical commitments lead them to reject the intuitions behind Creating Miserable People is a mark against those theoretical commitments. Rejecting the intuitions behind Creating Happy People might seem more palatable, but this revisionist approach has its own counterintuitive consequences. These problems have led some philosophers to advance preservationist arguments, which resist the Moral Actualism Dilemma’s contention that endorsing the Procreation Asymmetry leads to inconsistency.
3.2. Preservationism
As with revisionism, there are two basic strategies for a preservationist approach. One can reject premise 3 of the Moral Actualism Dilemma by arguing that the truth of moral actualism is consistent with the intuitions behind Creating Miserable People. Alternatively, one can reject premise 4 of the dilemma by arguing that the falsity of moral actualism is consistent with the intuitions behind Creating Happy People. Extant preservationist accounts have opted for the latter option, either by arguing directly against premise 4 or by arguing against premise 1, which would indirectly undermine premise 4. I will show in this section that these attempts to reject premise 4 have significant costs, and so we should consider the possibility of rejecting premise 3 instead.
Melinda Roberts (Reference Roberts2011a, Reference Roberts2011b) has developed the most well-known direct argument against premise 4 of the Moral Actualism Dilemma. On her view, which she calls ‘variabilism,’ the welfare and rights of all persons matter equally regardless of their modal status (i.e. moral actualism is false). However, even though all persons matter morally, not all of their missed opportunities for greater welfare (what Roberts calls ‘losses’) matter in the same way. Rather, a person’s losses matter variably, depending on whether the person incurs the loss in the actual world or in a merely possible world. If a person would incur a loss in the actual world, either by suffering a typical harm or by being caused to exist with a life that would be not worth living, then we have a reason to keep the person from incurring that loss. However, if a person would incur a loss in some merely possible world by never being caused to exist with a life that would be worth living, then we have no reason to keep the person from incurring that loss (Reference Roberts2011a, 355–356; Reference Roberts2011b, 773–774). Roberts’s variabilism purports to show that even if moral actualism is false, we will have a strong moral reason not to cause someone to exist just because doing so would impose the actual-world loss of suffering a miserable life, but no moral reason to cause someone to exist just because failing to do so would impose the possible-world loss of never enjoying a happy life.
If Roberts is correct, then premise 4 of the Moral Actualism Dilemma is false and the Procreation Asymmetry is thereby preserved. One problem with variabilism, however, is that it is ad hoc. As Johann Frick (Reference Frick2014, 56–60) has persuasively argued, Roberts offers no reason to accept the claim that people’s losses matter variably apart from an appeal to the intuitions behind Creating Miserable People and Creating Happy People. Why should we think that only some significant losses incurred by merely possible people matter morally (i.e. those that would be incurred in the actual world), while all significant losses incurred by actual people matter morally? To respond to this challenge, Roberts must either assume the truth of the Procreation Asymmetry, thereby begging the question against the Moral Actualism Dilemma, or provide independent reasons for believing that there is an asymmetry between our moral reasons to prevent actual-world losses and our moral reasons to prevent possible-world losses. Even if Roberts can assuage these worries, variabilism would commit us to a baroque and potentially mysterious asymmetry about the relative strengths of our moral reasons. Assuming that we do not want to commit ourselves to more asymmetries than moral intuition and sound moral theorizing can support, this seems to be yet another problem with variabilism.Footnote 18
Rather than attacking premise 4 directly, a preservationist might undermine it by arguing that we should reject premise 1. To see how this strategy could work, recall Harman’s view that there is an intuitive asymmetry between the strength of our moral reasons against causing states of affairs that are intrinsically bad for individuals (i.e. noncomparative harms) and the strength of our moral reasons in favor of causing states of affairs that are intrinsically good for individuals (i.e. noncomparative benefits).Footnote 19 Negative utilitarians and libertarians have argued for a more stringent version of this asymmetry, claiming what while we have strong moral reasons not to bring about intrinsically bad states of affairs for others, we do not have any moral reasons to bring about intrinsically good states of affairs for others. Although this position has been subjected to extensive criticism,Footnote 20 perhaps a stringent asymmetry is more plausible with respect to noncomparative harms and benefits, which are the welfare concepts at issue in the Procreation Asymmetry.Footnote 21 If such a view is correct, then premise 1 of the Moral Actualism Dilemma is false: we do not have moral reasons to promote the welfare of all persons who matter morally, where ‘promoting welfare’ includes both conferring noncomparative benefits and not imposing noncomparative harms. If we have moral reasons only to refrain from imposing noncomparative harms, then the falsity of moral actualism is perfectly consistent with Creating Happy People, and hence premise 4 of the Moral Actualism Dilemma is false.
Jeff McMahan (Reference McMahan2012, 15–16) has observed the following peculiarity with the claim that there is a stringent asymmetry between our moral reasons not to impose noncomparative harms and our moral reasons to confer noncomparative benefits: If we have strong moral reasons against imposing noncomparative harms on people, but no moral reasons in favor of conferring noncomparative benefits, then it seems that procreation would often be morally impermissible, even if the person we might create would have a very happy life. Since every life includes states that are noncomparatively bad for the person – e.g. injury, illness, heartbreak, failure, etc. – we have moral reasons against creating any person; but if the noncomparatively good states in a person’s life do not give us reasons that outweigh the reasons against procreation, then it appears creating any person is morally impermissible (except in the rare case that procreating is necessary to avoid harming others).Footnote 22 This result is highly counterintuitive, as we generally think that it is morally permissible to create very happy people, at least when doing so does not harm third parties.
One might try to solve this problem by arguing that while we do not have moral reasons to noncomparatively benefit persons by creating them, the positive value of noncomparative benefits can nonetheless weigh against or cancel our moral reasons not to impose noncomparative harms. As McMahan (Reference McMahan2012, 20–21) suggests, although noncomparative benefits might not have ‘reason-giving weight,’ in the sense that they can ground moral obligations, they might have ‘canceling weight,’ in the sense that they can discount corresponding noncomparative harms.Footnote 23 One problem with this response to the above objection is that it seems ad hoc. Apart from the fact that this response allows us to salvage the stringent asymmetry argument against premise 4 without running into counterintuitive consequences, there appears to be little reason to expect a divergence in the reason-giving weights and canceling weights of our moral reasons to bring about noncomparative benefits.
A second problem is that, even if this odd divergence could be shown to be independently motivated, it leaves us with a cumbersome set of views about metaethics or the metaphysics of reasons. In order to reject premise 4 of the Moral Actualism Dilemma without implausible entailments, this preservationist approach requires that we believe (1) there is some asymmetry between the strengths of our reasons not to harm and our reasons to benefit, (2) this asymmetry is stringent in the case of noncomparative harms and benefits, and (3) our moral reasons to cause noncomparative benefits have some significant canceling weight but no reason-giving weight. One might reasonably be skeptical that it is either necessary or wise to adopt all of this theoretical machinery in order to preserve the moral intuitions behind the Procreation Asymmetry.
Although I have not considered every preservationist attempt to resolve the Procreation Asymmetry’s latent inconsistency, I have shown that there is a common set of problems for both Roberts’s variabilism and for the view that our reasons not to harm are stringently asymmetrical with our reasons to benefit. First, these views might fail to plausibly defend the Procreation Asymmetry: variabilism appears to beg the question, while the stringent asymmetry view generates counterintuitive consequences (unless it includes some seemingly ad hoc assumptions). Second, these preservationist arguments require one to accept controversial metaethical or metaphysical claims, which might be objectionable on grounds of parsimony, precision, or preexisting commitments in moral theory. The available preservationist options either fail to save the Procreation Asymmetry from the Moral Actualism Dilemma or do so only by loading us down with heavy theoretical baggage.
I conjecture that part of what pushes some philosophers to revisionism is the judgment that the intuitive costs of rejecting Creating Miserable People or Creating Happy People are lower than the theoretical costs of preservationism. Perhaps the intuitions motivating the Procreation Asymmetry do not seem either so compelling or so widespread as to justify such drastic alterations to our metaethical or metaphysical outlook. On the one hand, I find this sort of motivation for revisionism about the moral intuitions behind the Procreation Asymmetry to be deeply compelling. On the other hand, I find those moral intuitions to be deeply compelling. What we need is a compelling defense of the intuitions behind Creating Miserable People and Creating Happy People that has the virtue of portability, such that philosophers with diverse and conflicting theoretical commitments in metaethics and metaphysics will be able to accept it. I articulate just such a defense in the rest of this article.
4. The Revised Procreation Asymmetry
Let’s pause for a moment to take stock of where we are. We began by identifying two commonsense intuitions about the morality of procreation, which we formulated as Creating Miserable People and Creating Happy People. The trouble with these propositions is that the Moral Actualism Dilemma showed us that accepting both of them leads to inconsistency. Revisionists have tried to resolve this inconsistency by abandoning the core moral intuitions behind either Creating Miserable People or Creating Happy People, but this approach generates counterintuitive consequences. Preservationists have tried to show that the Moral Actualism Dilemma is unsound because premise 4 is false, but their arguments either rest on ad hoc assumptions or commit us to multiple controversial metaethical and metaphysical claims. In this section, I will argue that the Moral Actualism Dilemma succeeds only because Creating Miserable People and Creating Happy People misrepresent our moral intuitions by utilizing an overly broad sense of ‘creation.’ With some careful thinking about what we mean by ‘creation’ in different contexts, we can give a portable preservationist defense of the commonsense moral intuitions behind the Procreation Asymmetry.
Consider the following case:
Anjali: Anjali’s doctor has informed her that if she becomes pregnant at any time in the next six months, any individual she might conceive would have a genetic abnormality such that, were it carried to term, its life would be truly miserable. Acting against her doctor’s recommendations, Anjali intentionally conceives just such an individual the following month. Anjali does this because she wants to have the experience of discovering that she is pregnant, and has no interest in giving birth or becoming a mother. Having satisfied this desire, she terminates the pregnancy in the first trimester, just as she originally intended.Footnote 24
Anjali’s actions might be irrational in some way or might demonstrate a lack of virtue. Her actions might even be morally wrong, perhaps because terminating the pregnancy violates the fetus’s right to life, or because conceiving a fetus with the intention of destroying it is disrespectful of the value of human life. Is it also true, though that Anjali acts against a strong moral reason not to create a person whose life would be not worth living? That is, does Anjali act against the kind of moral reason picked out by Creating Miserable People?Footnote 25
Creating Miserable People says that the fact that a person’s life would be miserable gives us a strong moral reason not to create that person. Setting aside the question of which discrete biological event constitutes conception,Footnote 26 there is room for disagreement about whether intentionally conceiving counts as ‘creating a person,’ as that expression appears in Creating Miserable People. Although something comes into existence at conception, one might think that the thing created is not a uniquely individuated organism (e.g. Bedate and Cefalo Reference Bedate and Cefalo1989; Smith and Brogaard Reference Smith and Brogaard2003). Even if one does believe that conception brings into existence a new individual organism, one might think that this organism is not a ‘person,’ an individual with significant moral status (e.g. Tooley Reference Tooley1972; Warren Reference Warren1973; McMahan Reference McMahan2002, chap. 4). On either sort of view, Anjali does not act against the moral reason picked out by Creating Miserable People, because although Anjali intentionally conceives, she does not thereby ‘create a person.’
Assume for the moment that neither of these views is correct, and that by conceiving one brings into existence a new individual organism with significant moral status (e.g. Brody Reference Brody1973; George and Tollefsen Reference George and Tollefsen2008). Given this assumption, Anjali’s act of conceiving does count as ‘creating a person,’ but then Creating Miserable People appears to commit us to making an implausible moral judgment about her case. Anjali’s plan to terminate her pregnancy shortly after conception means that there is no significant risk that the person she creates will ever go on to have a life not worth living. Although the fetal person would come to have a miserable life on the outside chance that Anjali’s plan were to fail, by the time Anjali terminates her pregnancy, that person neither suffers nor is put at significant risk of suffering such a life. If Anjali acts against the moral reason picked out by Creating Miserable People, then this would mean that an action can be contrary to that reason even if it brings about no significant risk that any person will suffer the sort of diminished well-being or rights violations that come with a miserable life. But then Creating Miserable People would be implausible as a general claim about our moral reasons; at best, it would seem to describe a prima facie moral guideline, with exceptions for cases like Anjali’s.
One might raise the following objection to the claim that if ‘creating a person’ refers to conception, then we are committed to saying that Anjali acts against the reason picked out by Creating Miserable People: If Anjali’s plan to terminate the pregnancy early on is almost certain to succeed, then it is false that the fetal person would have a miserable life. Being created and then aborted within a few weeks seems to be neither an overall bad life nor an overall good life for the person to live, if it can properly be called a life at all. This objection is right to scrutinize the underspecified counterfactual element in Creating Miserable People, but it seems that how this counterfactual should be interpreted cannot hinge on what procreators plan to do following conception. As McMahan (Reference McMahan, Roberts and Wasserman2009, 52) suggests, including Anjali’s intentions in the counterfactual would have the strange consequence that the existence of her moral reason would vary inversely according to whether she intends to do what (de re) the reason requires. Holding constant any external impediments to carrying out her plans, as Anjali’s intention to abort weakens, it becomes more likely that the moral reason exists, but as her intention strengthens, it becomes less likely that the moral reason exists, presumably vanishing entirely at the moment before the pregnancy is terminated. Maybe a reason with properties like these is possible, but it seems more likely that whatever reasons Anjali has, she has them in virtue of facts that do not depend in this way on her own agency. The counterfactual in Creating Miserable People, then, is most plausibly understood to include only those facts that do not depend on a procreator’s future actions or intentions.
If conception does not count as creating a person, then Anjali does not act against the moral reason picked out by Creating Miserable People. If conception does count as creating a person, then Anjali does act against the ‘reason’ picked out by Creating Miserable People, but then it is implausible that this is a genuine moral reason, rather than some sort of prima facie moral guideline. Creating Miserable People specifies some conditions under which we have a moral reason not to ‘create a person,’ but it seems that ‘creating a person’ cannot refer to intentionally causing conception. Parallel arguments could be made that ‘creating a person’ cannot refer to causing a number of events that occur in the early stages of the biological process of ontogenesis: formation of the primitive streak, differentiation of various essential organs, beginning of the heartbeat, formation of sexual organs, etc. Like conception, no person is put at significant risk of suffering a life not worth living merely by undergoing these early developmental milestones, and so it seems that bringing them about cannot be against our moral reasons not to harm or wrong people by causing them to suffer a miserable life.
Creating Miserable People gives an overly broad statement of our moral intuitions. I propose that we narrow it as follows:
Finishing Miserable People: The fact that a person would be miserable – i.e. his or her life would be not worth living – gives us a strong moral reason not to finish creating that person.
The expression ‘to finish creating a person’ as it appears in Finishing Miserable People excludes conception and other events that occur in the early stages of ontogenesis, thus excluding the possibility that Anjali acts against this moral reason. At the same time, Finishing Miserable People can be endorsed regardless of one’s views about when and how persons come into existence. If one believes that a person is brought into existence at conception or some other early-stage event, then Finishing Miserable People can be plausibly construed as telling us not to allow a person to develop the properties necessary for suffering a life not worth living, such as sentience, consciousness, or self-awareness, all of which arise only in the later stages of ontogenesis. If one believes that no distinct biological individual or no individual with moral status comes into existence before one of these late-stage developmental events, then Finishing Miserable People can be plausibly construed as telling us, at a minimum, not to allow an existing fetal person to actually suffer the kinds of experiences that make one’s life not worth living. Although when an agent has a moral reason to prevent the developing person from being harmed or wronged by being put at risk of suffering a miserable life will vary according to when persons come into existence, Finishing Miserable People identifies a strong moral reason to prevent the process of ontogenesis from reaching its completion, most likely by inducing abortion.Footnote 27
Given that we have revised Creating Miserable People in order to narrow the meaning of ‘creating a person,’ we should similarly revise Creating Happy People. I propose the following:
Beginning Happy People: The fact that a person would be happy – i.e. his or her life would be worth living – gives us no moral reason to begin creating that person.
The notion of ‘to begin creating a person’ hews more closely to the intuitive sense of ‘to create a person’ as it appears in Creating Miserable People and Creating Happy People. It refers to whatever events are necessary to bring a person into existence, whatever one’s view of how people come into existence. If one believes that a person is brought into existence at conception or some other early-stage gestational event, then Beginning Happy People tells us that the fact that such a person would be happy gives us no moral reason to bring about that early-stage event. If one believes that an individual organism or a person is brought into existence at some late-stage gestational event such as the development of sentience, consciousness, or viability, then Beginning Happy People tells us that the fact that such a person would be happy gives us no moral reason to bring about that late-stage event.
One advantage of Beginning Happy People over Creating Happy People is that it can be consistently endorsed by those who believe both (1) the fact that a person one might create would be happy gives one no moral reason to become pregnant and (2) the fact that a person one has already begun to create would be happy gives one a moral reason to finish creating that person. That is, we cannot immediately infer a lack of reasons to finish the process of ontogenesis from a lack of reasons to begin the process of ontogenesis. Since ‘creation’ as it appears in Creating Happy People is ambiguous across the early and late stages of gestation, it is not clear that one could consistently endorse that proposition while also believing, for instance, that the fact that a 15-week-old fetus would be happy if brought to term gives one a moral reason (of some strength) to continue gestating that fetus. Beginning Happy People, however, implies no inconsistency between those two positions.
Call the conjunction of Finishing Miserable People and Beginning Happy People the Revised Procreation Asymmetry. I propose adopting the Revised Procreation Asymmetry in order to escape the force of the Moral Actualism Dilemma. Consider a version of the dilemma aimed at this asymmetry:
The Revised Moral Actualism Dilemma
-
(1) We have moral reasons to promote the welfare and respect the rights of all and only those persons who matter morally.
-
(2) Either moral actualism is true or moral actualism is false.
-
(3) If moral actualism is true, then Beginning Happy People is true and Finishing Miserable People is false.
-
(4) If moral actualism is false, then Finishing Miserable People is possibly true and Beginning Happy People is false.
-
(5) Therefore, either Beginning Happy People is true and Finishing Miserable People is false, or Finishing Miserable People is possibly true and Beginning Happy People is false.
-
(6) Therefore, either Finishing Miserable People is false or Beginning Happy People is false.
-
(7) Therefore, the Revised Procreation Asymmetry is false.
The Revised Moral Actualism Dilemma is unsound because premise 3 is false. The truth of moral actualism would not entail the falsity of Finishing Miserable People: Finishing Miserable People claims that we have a strong moral reason to prevent an actual person from suffering (or being put at significant risk of suffering) a life not worth living. This claim is perfectly consistent with the view that we have moral reasons regarding the rights and welfare only of actual people. Beginning Happy People does refer to merely possible people, but only to deny that the fact that they would have happy lives gives us any moral reason to ensure that they in fact enjoy such lives, and this is also consistent with moral actualism.
My preservationist proposal, then, is that we should accept the Revised Procreation Asymmetry. The Revised Procreation Asymmetry preserves the core moral intuitions that motivated the original Procreation Asymmetry without leading to the sort of inconsistency identified by the Moral Actualism Dilemma. The Revised Procreation Asymmetry has three additional advantages over the original. First, Finishing Miserable People is consistent with our moral intuitions about cases such as Anjali’s. Even on various conflicting views about when and how persons come into existence, it need not entail that Anjali’s conception counts as acting against the strong moral reason not to create a person just because that person’s life would be miserable. Second, Beginning Happy People is consistent with the common (though by no means universal) belief that the fact that a person that one has already begun creating would be happy can give one some moral reason to finish gestating that person. Third, accepting the Revised Procreation Asymmetry avoids the problems faced by other preservationist approaches: it is not ad hoc, nor does it require us to accept highly controversial metaphyisical or metaethical views. In these respects, the Revised Procreation Asymmetry offers a plausible, portable justification of the core moral intuitions that were initially formulated in the Procreation Asymmetry.
5. Objections and replies
One objection to my argument is that the Revised Procreation Asymmetry does not preserve all of the moral intuitions captured by the original Procreation Asymmetry. Consider the following case:
Creation Machine: An eccentric scientist has built a machine that, when switched on, will first combine a sperm cell and an egg cell to form a zygote, and then will gestate the zygote until it exits the machine as a fully formed newborn infant. Not wanting anyone to interfere with his glorious machine, the scientist has cloistered himself in a remote location and hidden his activities from all other agents. After a lifetime of testing and recalibrating the machine, the scientist finds himself on his deathbed. He can use the very last of his strength to activate the machine, but if he does, the machine would combine a sperm and egg that would inevitably result in the existence of a new person whose life would be not worth living.Footnote 28
Creating Miserable People would tell us that the scientist has a strong moral reason not to activate the creation machine. But Finishing Miserable People would not tell us this, since the uncombined sperm and egg are not yet an individual organism, let alone an actual person. Intuitively, the scientist should not activate the machine in this case, so it seems the Revised Procreation Asymmetry fails to account for all of the moral intuitions that the original Procreation Asymmetry accounts for.
One response to this objection is just to admit that the Revised Procreation Asymmetry cannot support all of the same intuitions as the original Procreation Asymmetry, and then to deny that this is much of a problem. The original Procreation Asymmetry was shown by the Moral Actualism Dilemma to lead to inconsistency, so of course it can account for more of our commonsense moral intuitions, since it is likely that the full set of intuitions that constitute commonsense morality is not perfectly self-consistent. All things considered, it is better to have a consistent set of beliefs that cannot support all of our intuitions about extremely improbable cases than an inconsistent set that can. Another response is to note that even if the scientist has not acted against the moral reason picked out by Finishing Miserable People, it is likely that his actions can be morally criticized on other grounds. Perhaps the scientist’s action is wrong because it adds more unnecessary suffering to the universe, because it expresses disrespect for currently existing human beings, or because it instantiates vices such as imprudence, capriciousness, or selfishness. Similarly, even if the scientist had no moral reason not to activate the machine, it might yet be true that the scientist can be appropriately blamed in retrospect for causing a person to suffer a life not worth living. Given how peculiar and unrealistic this case is, it is not clear whether our intuitive response indicates that the scientist acts wrongly because he acts against the sort of moral reason picked out by Creating Miserable People or merely that the scientist acts wrongly in virtue of some other set of considerations.
Another objection to my argument is that Finishing Miserable People assumes that abortion is in some sense feasible for the gestating woman. If Anjali lives in a world that lacks the technologies necessary for terminating a pregnancy, then it cannot be the case that she has a strong moral reason against finishing her pregnancy, since there is no way she could act on this reason. But the original Procreation Asymmetry did not seem in this way to be limited to our current technological or institutional circumstances, so the Revised Procreation Asymmetry does not preserve all of the commonsense moral intuitions which were supported by the original Procreation Asymmetry.
I will grant that the Revised Procreation Asymmetry does assume the technological feasibility of abortion, but deny that this means that it preserves fewer commonsense moral intuitions than the original Procreation Asymmetry. Abortion has been technologically (and often socially) feasible for much of human history. Abortion has not always been safe for women, but recall that Finishing Miserable People does not hold that there is an all-things-considered moral obligation not to finish creating a person whose life would be miserable, just that there is a strong moral reason not to do so. Even if abortion is not a prudent (or even morally permissible) decision for the woman, this strong moral reason not to finish creating such a person would plausibly justify regret over failing to abort in such a case, or could render an otherwise imprudent or impermissible attempt to abort in such circumstances rationally or morally justifiable.
It is important to note that the original Procreation Asymmetry is also set against a certain background of technological development, one in which we can have some epistemic traction on whether a merely possible person’s life would be worth living or not. This kind of epistemic position was consistently realized only in the twentieth century, with the advancement of ultrasound technology, genetic testing, etc. Prior to these developments, it seems more likely that finishing creating a child whose life would be miserable would have been seen as an unavoidable tragedy, rather than a state of affairs that commonsense morality says an agent has a moral reason to prevent. For the limited historical period for which we are entitled to assume our commonsense moral intuitions related to the original Procreation Asymmetry, abortion has not only been technologically feasible, but often socially acceptable and reasonably safe. So the feasibility constraints on the Revised Procreation Asymmetry closely (if only contingently) match those of the original Procreation Asymmetry.
A related line of objection to my argument is that endorsing Finishing Miserable People commits one to counterintuitive or controversial normative claims. According to Finishing Miserable People, Anjali has a strong moral reason not to finish creating the person she previously conceived because that person would have a life not worth living. But if this is the case, then it seems consistency would demand that a gestating woman would have an equally strong moral reason to finish creating any person she previously conceived if that person would have a life worth living. My discussion of Beginning Happy People takes it to be a virtue of that formulation that it leaves such a position open, but how are we to avoid being committed to that position, especially given premise 1 of the (Revised) Moral Actualism Dilemma? It is at least highly controversial, if not highly counterintuitive, to claim that a fetus’s future welfare gives us a strong moral reason to finish gestating it. Many people do not accept that the value of a fetus’s future gives a gestating woman a strong moral reason to refrain from terminating a pregnancy, even in the later stages of pregnancy.
I grant that, if we assume that Finishing Miserable People is true, then norms of consistency favor endorsing the proposition that we have strong moral reasons to finish creating people whose lives would be worth living. However, I do not think that this result is all that counterintuitive or controversial. First, the claim that one has a strong moral reason to finish creating a person does not entail that one has a moral obligation to refrain from abortion; strong moral reasons might be outweighed by other moral reasons, or rights to bodily or reproductive autonomy might exempt one from countenancing such reasons (Little Reference Little1999). Second, and perhaps more importantly, this is not merely a problem for my view, but for anyone who holds that we have a strong moral reason to stop creating a person due to the fact that his or her life would be miserable. That is, this problem is perfectly general to anyone who is committed to Finishing Miserable People, which is intuitively plausible in its own right, even if not as a replacement for Creating Miserable People.Footnote 29
A final objection to my argument is that the Revised Procreation Asymmetry is not as portable as it might seem. Finishing Miserable People maintains that one has a strong moral reason not to finish creating (i.e. gestating) a person whose life would be not worth living, presumably because having a life that is not worth living would be bad for that person. But this conflicts with some well-respected views about the grounds of personal identity and rational egoistic concern, due to the fact that even late-stage fetuses (and newborn children) are psychologically isolated from their later selves. If personal identity is grounded in psychological connectedness, then many late-stage fetuses will not be numerically identical with people who suffer lives not worth living (McMahan Reference McMahan2002, 269–270). Similarly, if rational egoistic concern is grounded in psychological connectedness, then a fetus might have only weak time-relative interests in avoiding suffering a life not worth living (79–80). It seems, then, that either (1) Finishing Miserable People tells us that we might have a moral reason not to create a particular person before that person (i.e. the subject who would be afflicted with a life not worth living) comes into existence, in which case Finishing Miserable People makes no progress over Creating Miserable People, or (2) accepting the Revised Procreation Asymmetry commits us to the view that personal identity and rational egoistic concern are not grounded in psychological connectedness. Accepting the Revised Procreation Asymmetry therefore cannot preserve the commonsense moral intuitions expressed in the original Procreation Asymmetry without requiring us to take on controversial metaethical or metaphysical commitments.
I have two brief responses to this objection. First, as we saw with the previous objection, Finishing Miserable People is intuitively compelling in its own right, apart from its role in my preservationist argument. Assuming that Finishing Miserable People is inconsistent with psychological accounts of the grounds of personal identity or rational egoistic concern, this seems to be a pro tanto reason for rejecting those accounts in favor of alternatives that are consistent with this highly intuitive moral proposition. Similarly, the fact that accepting the Revised Procreation Asymmetry seems to provide a plausible justification of some core moral intuitions about creation might in itself give us reason to reject those accounts of the grounds of personal identity or rational egoistic concern that cannot be reconciled with such a justification. One philosopher’s modus tollens is another’s modus ponens. Second, even if this objection does show that accepting the Revised Procreation Asymmetry is not a fully portable defense of the commonsense intuitions behind the original Procreation Asymmetry, it is nonetheless a highly portable defense. The Revised Procreation Asymmetry is compatible with a number of metaethical and metaphysical commitments, and unlike the other preservationist approaches, accepting it does not commit one to any novel or highly speculative metaphysical or metaethical theses. There is a number of well-developed views that do not ground personal identity or rational egoistic concern exclusively in psychological connectedness, and such views are fully compatible with the Revised Procreation Asymmetry.
6. Conclusion
I have argued that, pace the Moral Actualism Dilemma, we can hold onto the commonsense moral intuitions behind the Procreation Asymmetry. Unlike Creating Miserable People and Creating Happy People, Finishing Miserable People and Beginning Happy People are not shown to generate inconsistency by the Revised Moral Actualism Dilemma, as the third premise of that argument is false. I have not argued, however, that the moral intuitions behind the conjuncts of the Revised Procreation Asymmetry are true. Indeed, Beginning Happy People, the claim that the fact that a person would be happy does not give us a requiring moral reason to begin creating that person, is true only if moral actualism is true as well – premise 4 of the Revised Moral Actualism Dilemma establishes this. Although I have not given an argument for moral actualism in this article, I have addressed one of the most significant challenges facing this view. The moral actualist, contrary to first appearances, can consistently endorse the commonsense moral intuition expressed in Finishing Miserable People: the fact that a person would be miserable gives us a strong moral reason not to finish creating that person.Footnote 30
