Article Text

Download PDFPDF

The moral significance of being born
  1. Neil Levy
  1. Correspondence to Dr Neil Levy, Florey Institute of Neuroscience and Mental Health, University of Melbourne, Parkville, Victoria 3010, Australia; neil.levy{at}


This paper is a response to Giubilini and Minerva's defence of infanticide. I argue that any account of moral worth or moral rights that depends on the intrinsic properties of individuals alone is committed to agreeing with Giubilini and Minerva that birth cannot by itself make a moral difference to the moral worth of the infant. However, I argue that moral worth need not depend on intrinsic properties alone. It might also depend on relational and social properties. I claim that the in principle availability of neonates to participate in scaffolded interactions with carers might plausibly be seen as contributing to their moral worth.

  • Abortion

Statistics from

The howls of outrage provoked by Giubilini and Minerva's examination of ‘after-birth abortion’1 indicate the extent to which the deliberate killing of newborn infants is at odds with common sense. Most people find it obvious that there is a categorical difference between such killing and abortion. Even the most fervent opponents of abortion don't think that abortion is as significant a wrong as homicide, but very many people think that the killing of newborns is every bit as wrong as killing older children or adults. Of course, we cannot simply appeal to common sense to decide difficult ethical issues. Common sense is often confused, at odds with itself and sometimes driven by psychological processes that are not truth tracking. Nevertheless, there are good reasons to take intuitions seriously. Most philosophers believe that the justification of moral claims depends crucially on our intuitions, since justification depends on passing the test of wide reflective equilibrium.2 ,3 Applying the method of wide reflective equilibrium involves the search for consistency between our intuitions, and our moral and non-moral theories, adjusting each as necessary to achieve coherence. For that reason, we cannot simply dismiss the commonsensical conviction that birth makes a big moral difference.

If we cannot simply dismiss the conviction that birth is extremely morally significant, however, it seems just as difficult to accept it. Though some intuitions may be too foundational to morality to require justification, the intuition that there is a categorical difference between neonates and the unborn does not seem to be of this kind. Though we may not be able to offer any positive reason why we ought to accept a foundational moral principle (consider, say, the principle that inflicting pain on an innocent person requires justification), we never have good reason to reject such principles. But there does seem to be good reason to reject the intuition that birth is morally significant to the value of the infant all by itself: the intuition seems to conflict with the principle that moral value supervenes on intrinsic properties alone. There does not seem to be a relevant property on which the difference between infanticide and abortion might rest. Birth is simply a change in the location of the infant, not a change in its intrinsic (non-relational) properties.

Though many philosophers believe that relational facts matter for the morality of abortion, for most it is not because they matter, or can matter, for the moral value of the infant that they matter for the morality of abortion. Instead, it is because they hold that the morality of abortion concerns the mother, and therefore her relationship to the infant, as well as the moral worth of the infant. So far as its moral worth is concerned, most philosophers think only the intrinsic properties of the infant matter. It is these kinds of facts to which Giubilini and Minerva themselves advert, when they argue that the moral status of the newborn and the fetus are equivalent. Both ‘lack those properties that justify the attribution of a right to life to an individual’, they argue. Giubilini and Minerva follow philosophers such as Michael Tooley4 and Peter Singer5 in thinking that the relevant properties are psychological properties; the difference between a person and a non-person is a difference in the ‘level of her mental development’. Now it should be obvious that the location of the child does not correlate perfectly with her level of mental development. Though there is surely a rough correlation between mental development and location, simply because location correlates with age, the correlation between age and location is not perfect: babies are sometimes born premature. Because babies are sometimes born premature, some neonates are many weeks younger than some unborn babies. On the plausible assumption that mental development depends on brain development, and brain development takes time, it seems that in such cases the neonates will be less mentally developed than some unborn children. If moral significance supervenes on mental development, in such cases it seems likely that the unborn child will have a greater moral worth than the neonate—contrary to the commonsensical intuition.

In this paper, I will argue that despite the difficulties alluded to above with accepting the commonsensical intuition that birth is morally significant, there are reasons to think that the intuition is not entirely off-base. That is, I will claim that there are good reasons to think that the fact of being born adds to the moral worth of the neonate. I will not attempt to assess how much moral worth the fact of being born confers; for this reason, I will not attempt to assess whether a premature infant has a greater moral worth than an unborn and more developed child (assuming that psychological development also contributes to moral worth), nor will I attempt to adjudicate the question whether infanticide is impermissible (given that abortion is permissible). I will leave it entirely open how much extra moral worth is conferred by the fact of being born. The strategy I will develop is akin to that suggested by Mary Anne Warren in a now classic article (and also briefly advocated by Susan Sherwin).6 ,7 Like Warren and Sherwin, I will argue that neonates have a moral worth that depends on relational facts about them, and not merely on their intrinsic properties. Prior to developing this argument, though, let me first examine a failed strategy for vindicating the common sense intuition, one that is compatible with the usual view that moral worth depends on intrinsic properties alone.

The strategy I have in mind was developed by Jose Luis Bermúdez.8 He argues that birth is morally significant because being born is a necessary condition for exercising the capacity for self-consciousness. Bermúdez cites evidence that the capacities of neonates are very much greater than philosophers tend to think. In particular, neonates as young as 42 minutes old give evidence of a capacity for imitation that seems to indicate an unexpectedly high degree of self-consciousness. Imitating the facial expression of an adult seems to require that one understand that the adult is distinct from, but similar to, oneself, and that seems to indicate ‘a primitive form of self-awareness’. Bermúdez does not believe that the degree of self-awareness for which imitation provides evidence is the degree of self-awareness that is plausibly required for personhood. Nevertheless, this more primitive self-awareness has a derivative moral significance for him. He justifies this claim by reference to what he calls ‘The Principle of Derived Moral Significance’, according to which if a property confers moral significance on a life, then a primitive form of that property also confers moral significance.

The principle of derived moral significance is somewhat plausible, though it is not obviously compelling. It may be that some properties confer moral significance only when a certain threshold is reached; below that threshold a primitive form of that property confers none (the property of being ‘rational’ might be a threshold property: it may be that an organism has to respond in a way that looks reasons-responsive to a sufficient number of stimuli before it counts as rational). However, self-awareness does seem to be the kind of property that makes a moral difference in a continuous fashion. A neonate may not be a person by virtue of a primitive degree of self-awareness, and (therefore) may not qualify as having a right to life, but surely its moral worth is somewhat greater by virtue of having such a primitive degree of self-awareness.

If Bermúdez is right, birth marks a morally important stage in the development of human beings, because only once the infant is born can it engage in imitation, and by imitating come to understand its body as its own. Prior to birth, it might develop an awareness of its body, but in the absence of appropriate perceptual input, it may not understand the distinction between its body and the bodies of other agents.9 However, even if this is correct, it does not vindicate the common sense intuition that there is a categorical difference between infanticide and abortion. A neonate will, in the normal course of events, come to perceive the bodies of other agents and thereby acquire a primitive form of self-awareness very rapidly after birth, but because acquisition of this psychological property takes time, appeal to this fact cannot vindicate the intuition. The intuition is that birth alone, independently of further developments for which it might be a necessary condition, confers additional moral worth on the child. Bermúdez's suggestion does not vindicate that claim.

Perhaps Bermúdez's suggestion nevertheless explains why people have the intuition. Perhaps it is the fact that birth is a necessary condition for the acquisition of important psychological properties (together, perhaps, with the fact that birth correlates reasonably well with age) that explains why people think it is morally significant. I want to suggest an alternative explanation, however, which if successful vindicates the intuition that birth confers moral worth all by itself.

As Mary Anne Warren suggests, any account of rights or moral worth that turns on the intrinsic properties of individuals alone cannot vindicate the common sense intuition, because birth does not alter any intrinsic property of the infant that could confer moral worth on it. But intrinsic properties are not the only properties that infants have. Warren proposes that the relational, in particular the social, properties of individuals may be ‘part of the foundation of moral rights’. For her, the move to relational properties seems to be partially strategic. It is not because she thinks that relational properties really confer moral worth that she defends them, it is because she thinks more rationally defensible grounds of worth have insurmountable problems. She maintains that on narrow rational grounds, ‘the capacity for sentience […] may be the only pragmatically defensible criterion for the ascription of full and equal basic rights’, but resists this criterion on the basis that it requires us to extend full and equal basic rights to fetuses. She resists such an extension, in turn, because ‘it is impossible to treat fetuses in utero as if they were persons without treating women as if they were something less than persons’. Because the only genuinely defensible criterion of rights possession entails a non-contingent conflict between the rights of individuals, Warren suggests that we endorse a partially relational account of the grounds of rights.

Warren therefore suggests that we confer a right to life to infants when they emerge into the social world. Just what constitutes emergence into the social world for her remains unclear. In what follows, I aim to develop her suggestion. I do not do so in quite the same spirit as hers, however: I don't believe that a relational concept of moral worth is rationally inferior to a conception that has it depend on intrinsic properties alone. Rather, I suggest that view has nothing to apologise for. The idea that intrinsic properties must be what matters rests on internalist assumptions; on the assumption that psychological states and capacities supervene on internal states of agents alone. Locational internalism of this sort is today under attack on a number of fronts.10–12 If we embrace locational externalism, we may come to recognise that many psychological properties may be partially world involving.

A relational account of moral worth

An approach that centres on relational properties might instead focus on the value conferred on the child's life. That is, we might say that the life of the neonate is morally more significant just because others see it as more significant. Such an approach has obvious problems, however. For one thing, it seems to be unacceptably subjectivist. It seems to entail that there is no fact of the matter whether a neonate has greater moral worth than an unborn child; rather, it makes moral worth depend on whether the child is valued. Perhaps worse, it is at best vague about whose valuing counts. Is it sufficient that someone value a neonate for its moral worth to rise? Might not the criterion entail that abortion is impermissible, or at very least no more permissible than infanticide, because anti-abortion activists value fetuses? Can moral worth hang from such a slender thread as whether someone happens to value the child?

But the relational approach does not have to take this subjectivist form. Rather than holding that worth depends on the attitudes of other agents, the relational approach might hold that worth depends on the child's actual capacities, but that these capacities depend (in part, of course) on its relationship to other people and to its environment. I shall argue that this second approach is defensible. However, the second approach has problems of its own: like Bermúdez's account, it explains how birth might make a difference by being a necessary condition for certain capacities, but not how birth makes an immediate moral difference independent of the development of such capacities (which takes time). I will suggest that what is needed is a combination of the relational account of worth and the relational account of capacities. The combination avoids the subjectivism of the first, but entails that birth makes an immediate moral difference.

With this in mind, I turn to the question of how the child's actual capacities might depend, in part, on her relationship to other people and to her environment. The ‘locus classicus’ here is the work of Lev Vygotsky.13 Vygotsky argues that many actions are ‘scaffolded’, where an action is scaffolded if it is possible only because some kind of external support is available to the agent. Much cognition is scaffolded: for instance, there is evidence that complex arithmetic requires the use of natural language representations of numbers.14 A scaffold such as this one remains indispensable: some kind of representational system is always required for this kind of cognition. Other scaffolds are required for the development of some kind of onboard capacity that, once in place, can dispense with the scaffold (think of training wheels on a bicycle). Vygotsky's main concern was with this second type of scaffolding. He argued that young children often have scaffolded capacities that greatly outrun their onboard capacities, because the first are acquired prior to, and often as a necessary condition for, the latter.

Vygotsky focused on speech as scaffolding, as when an adult talks a child through a task. But non-linguistic scaffolding of cognitive capacities is ubiquitous; so ubiquitous, in fact, that we typically miss it. We may not encode the layout of a familiar building, because we rely on the corridors and doors to serve as cues and guides for navigation. Children are especially dependent on scaffolds: even getting dressed—from selecting appropriate clothes through to the actions involved in putting them on—may require guidance from an adult or older child, or guidance from the environment (say the layout of the closet). In short, children's capacities may often be far greater than we often think, because we confuse their unscaffolded capacities with their capacities per se.

It is very probable that the capacity for self-awareness is scaffolded. As Kim Sterelny has argued, human beings engineer the child's environment to scaffold the acquisition of the skills involved in interpreting other agents, for instance by feeding the child simplified narratives that make intentions and their attribution salient.15 If, as many philosophers believe, self-awareness requires (or is even constituted by) theory of mind,16 scaffolding these skills would indirectly scaffold self-awareness. More directly (but somewhat speculatively), the use of the child's name by its parents and others probably plays an important role in making her aware of her distinctness and uniqueness. As we have already seen, however, we cannot appeal to facts such as this to account for the categorical difference between infanticide and abortion, because scaffolded or not, self-awareness takes time to develop. However quickly it might develop as a scaffolded capacity, self-awareness is not a capacity that the neonate has. Nevertheless, I shall argue the scaffolding of capacities may play a role in vindicating the common sense intuition that there is such a categorical difference.

Let me turn now to the relational account of moral worth suggested above, according to which the moral worth of an agent may depend (in part) on their social status. As I mentioned, this account faces the objection that it is unacceptably subjectivist: it seems to leave too few constraints on the conditions that must be met for an agent to be recognised as a member of our social group, and entails that the moral worth of identical agents may shift from context to context. These problems are serious in their own right, but they also may entail that the account cannot vindicate the common sense intuition. One might think of the impact of ultrasound images in the abortion debate. Prior to the availability of such images, the fetus was less easily imagined and thought of as one of us; now that such images are common, it easier to recognise—or imagine—their status as a member of our community. As a consequence, a line that might once have divided neonate from fetus has become blurred.

It is clear that Warren, in her important paper, does not suppose moral worth to hinge on such subjectivist factors. For her, birth marks ‘the beginnings of the infant's existence as a socially responsive member of a human community’. Warren urges the recognition, rather than the conferral, of the neonate's status. However, as she appears to acknowledge, the criterion of social responsiveness might not do the work of vindicating the common sense intuition, because responding to its environment and carers may take time (an infant may be born premature and unresponsive). Better, I think, to say that birth marks the entry of the neonate into full sociality. Even prior to its birth, the child may begin to enter into social relations, To the extent to which societies are, liked nations, imagined communities’,17 depending for their existence of the attitudes of members, the child may be accorded some social status well before it is born (indeed, even prior to conception its parents and other may already attribute to it social properties). The discovery of the child's sex, which is now common well before birth, and its naming, accelerate this process. But the ability to interact with the child independently of the mother marks a qualitative leap in our attitudes to it and consequently to its social being. It is, as Sherwin notes, now for the first time capable of the broad range of interactions with an indefinite number and variety of others that mark it as an individual.7 It becomes a member of our community by virtue of birth. The extra value that accrues to it as a consequence is not conferred by us; rather, it is incumbent on us to recognise this value. Because it is now fully interpellated into social relations, it is, in this respect, one of us.

However, the interpellation account of moral worth seems to face an analogous objection to a more projectivist account. Doesn't interpellation shift with technological and social changes? Isn't the fetus interpellated earlier, thanks to ultrasound, and the social changes it brings? Perhaps in part this is true: if parents begin to think of a child in more concrete ways thanks to ultrasound images, this may actually change the child's worth (and not only its worth to them). However, I want to claim that birth marks a significant enough change in the child's status to make it reasonable to identify it as the entry into full sociality, independent of such imaginings. It is here that I appeal to scaffolding. Once the child is born, and only once it is born, its capacities are scaffolded by the parent. As Vygotsky emphasised, the neonate is far from an independent, autonomous, being; in many ways, it exists as a part of carer-child dyad. Just as it is misleading (for many purposes) to consider the child in abstraction from this dyad, I suggest, it is misleading to assess its worth independently of its capacity to form a part of such a carer-child relationship. Its worth is partially dependent on its availability to form part of that dyad. That is, its moral worth comes, in part, not from actually being a part of such a dyad (of course many children are abandoned or orphaned) but from the fact that it is capable of being part of such a dyad. We recognise that status in recognising its entry into full sociality (note that this claim is entirely compatible with also thinking that actually being valued by parents or others, and actually being part of a carer-child dyad also confer additional worth on the child. It is also compatible with some prenatal entry into the social. Again, we aim to explain how birth increases moral status, not to argue that it does so in a unique or uniquely significant manner).

Notice that being available to enter into a carer-child dyad is very significantly dependent on the child's psychological capacities: the child is equipped to take advantage of scaffolding in virtue of its internal states, as a consequence of the coevolution of onboard capacities and social environment.15 However, the child alone is available to enter into the relationship in only an attenuated sense. Full availability depends on the equal availability of appropriate partners (we can attribute availability to the isolated neonate only by virtue of our prior knowledge of how it would relate to potential carers; further, actually interacting with an infant activates processes—imitation, gaze tracking, mirroring of facial expressions—that together constitute the carer-child dyad). Though its availability is dependent on its intrinsic properties, what it is available for depends on an appropriate environment, not as mere triggers but as ongoing supports, guides and inputs. The capacities manifested in the carer-child dyad are therefore categorically different from capacities that can be exercised unaided (it is also worth remarking that since availability, in this sense, depends on intrinsic properties of the infant as well as its environment, absence of these capacities may be relevant to its moral worth and therefore to the permissibility of infanticide).

These remarks are sketchy and somewhat speculative. They do not constitute a full defence of the claim that moral worth might be increased by birth. They leave it entirely open just how much moral worth is increased by birth. They are, and are intended to be, consistent with the claim that other factors (such as the development of psychological capacities dependent on intrinsic properties, or the conferral of value on the child) are far more significant. They leave it open whether the increase in value is significant enough to rule out the permissibility of infanticide. I have attempted only to argue that birth increases moral value somewhat. This entails only that killing the neonate is a more serious matter than killing the fetus; how much more serious I cannot yet say. The aim of this paper has not been to rule on questions such as the permissibility of infanticide, but merely to show that the commonsensical intuition that birth makes a moral difference might rest on a truth, however obscurely grasped.


The author is grateful to the Australian Research Council for funding that supported this work, and to two anonymous reviewers for this journal.



  • Contributors NL is sole author of this paper.

  • Funding Australian Research Council, FT100100261.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Other content recommended for you