By now I have, hopefully, provided a good idea of how the traditional debate about personhood’s conditions proceeds, albeit without offering a solution. We have also seen just how difficult it is to propound a criterion of moral status which evades all of the obvious reductio ad absurdum problems, as well as other kinds of counter-arguments.
By the end of any such discussion, supporters of the species membership criterion will be tempted to think that they have come out of it the clear victors. After all, only their favoured criterion of personhood clearly brings all human beings— including infants, the comatose, and the radically cognitively defective—within the parameters of full personhood status. And if the pure species membership criterion for personhood is correct, and if embryos and fetuses are human beings, it follows that they too are included in the category of persons.
But this is a hollow victory for proponents of personhood-from-conception, if it is one at all. It is, to begin with, a victory grounded on purely negative virtues. The strongest recommendation for the species membership view comes from the fact that, unlike its capacities-based contenders, it is consistent with the equal moral status of human beings we are not comfortable leaving out of the personhood category. This is a far cry from offering a compelling rationale in defence of the species membership criterion, or answering the ‘speciesism’ charge. Particularly in light of the obvious conceptual difference between biological humanity and moral personhood (a difference which, I have stressed throughout, cannot be denied), we still require an explanation for why personhood supervenes on human genetics. An explanation, that is, which offers more than just satisfying entailments. Moreover, the negative virtues of the species membership criterion do nothing to neutralize its own reductios. Could it really be true that the death of a single-celled zygote just moments after conception is tantamount to the death of a five-year-old child? Many will have great difficulty believing this to be true.
It is crucial here not to employ double standards with the epistemic value we place on reductio arguments. For example, defenders of the species membership criterion cannot argue that post-conception criteria of personhood must be rejected as soon as they are seen to countenance infanticide, but that the species membership criterion does not have to be rejected for its equally implausible suggestion that zygotes are morally equivalent to five-year-old children. Fair play is paramount here. One cannot say only about one’s own theory that whatever conclusions fall out of it is simply where the logic leads us, whilst maintaining that patently unacceptable implications of rival theories show them to be false. Those who would countenance the moral permissibility of infanticide in the same circumstances we permit abortion also need to be mindful of this when impugning personhood- from-conception accounts for having obviously false implications.
Of course, one’s own judgements about the relative plausibility of the various reductios will always count for much in discussions like these. It is always open to someone, when confronted with a putative reductio of her criterion of personhood, to simply deny that it is indeed a reductio and claim that it is instead a correct entail- ment telling us something we deep down already knew to be true. One woman’s reductio is another woman’s thesis-affirming conclusion. Hence, Kaczor, George, and Tollefsen, and other supporters of the species membership criterion do not reject it once it leads them to admit that zygotes are every bit as morally valuable as adult human beings. This is, after all, what they always imagined to be true. Likewise for supporters of theories which lead inescapably to the view that infanticide is no different from late abortion. For some people, this is an entirely intuitive implication, and therefore not a reductio at all.
All of this places some limitations on reductio reasoning in argument about per- sonhood. Interlocutors may devote endless time to constructing reductios out of one another’s theories, but as soon as there is disagreement on whether something counts as a reductio, it is not obvious where to go from there. It may be feared that, at bottom, reductio reasoning just replays deep-seated disagreements in our starting convictions about what can and cannot plausibly count as a person with full moral rights. This limitation does not make reductio reasoning pointless in moral argument about abortion. At the very least, summoning counter-examples such as radically cognitively defective persons or thought-experiments such as the Embryo Rescue Case compels discussants to reconsider their theory of moral status in response to its apparent implications, or to deepen or elaborate that theory so as to explain them. Still, at the point at which discussants find themselves conflicted about which entailments really are too counter-intuitive to accept, we may well ask ourselves if there is any way of thinking about moral status which moves the constant exchange of counter-examples.
One suggested way of breaking out of this cycle of counter-examples has been to adopt a multi-criterial account of the conditions of personhood. My discussion so far has focussed on the various counter-examples that can be levelled against any one putative criterion of personhood taken to be both necessary and sufficient for full moral status. Some philosophers, however, have eschewed the idea that there is only one core criterion for personhood, favouring instead a ‘cluster concept’ account of personhood’s conditions. Accounts like this propose that personhood does not admit of a strict set of necessary and sufficient conditions, but supervenes on a cluster of properties, only an adequate selection of which are required for personhood.
Mary Anne Warren enumerated five such characteristics: 1. consciousness (especially the capacity to feel pain); 2. reasoning ability; 3. self-motivated activity (or, we might say, agency); 4. the capacity to communicate; and 5. a concept of the self.25 Warren contended that there is no strict formula regarding exactly how many of the five traits an individual must possess, or in which combinations, to warrant an ascription of personhood. The only clear thing, she argued, is that a creature possessing none of the traits is not a person. Warren thus rejected the idea that any single property demarcates those creatures which possess full moral standing from those which do not—what she called, ‘the single-criterion assumption’^6
Reflecting on the previous discussion, it is easy to see that, on its own, each one of Warren’s criteria is subject to a counter-example. Is a human being afflicted with a rare disease rendering her incapable of pain sensation thereby not a person? Is someone who becomes paralyzed and incapable of any self-motivated activity stripped, for that reason, of his personhood status? On Warren’s account, no one single trait need bear all of the weight. Someone incapable of self-motivated activity could still be a person if, say, he possesses consciousness, reasoning ability, and a concept of the self. A human being who lacks reasoning ability may still be a person if she is conscious, has the capacity to communicate, and is capable of some agency. We might wonder, then, whether a multi-criterial approach such as Warren’s, if persuasive, can break us out of the stalemate induced by the countless reductios.
Against the cluster concept approach, Kaczor has argued that such a view cannot rescue developmental criteria for personhood from the counter-example problem. He rejects a solution by which a selection of traits which have all been shown to be individually inessential for personhood (by means of counter-examples) could be simply lashed together in a theory which takes moral status to supervene on all of them, although not, necessarily, on any of them. Kaczor frames the objection as one which cuts against the gradualist (in sense 2) view of fetal moral status, according to which the fetus strengthens in moral value throughout development. However, it is more fundamentally an argument against any multi-criterial concept of personhood, gradualist or not. Kaczor illustrates the multi-criterial approach by analogy with the construction of a rope.27 Each individual stage of development from conception to birth is a single thread which is easily broken (by a counter-example which shows it to be over- or under-i nclusive). When added together, however, the various ‘threads’ of development construct a rope which is difficult to split: a sound basis for the attribution of moral status. Taking the rope analogy as a target, he writes:
The weakness of the rope analogy ... is the comparison of the various arguments in favour of personhood to threads which, though weak taken individually, make an increasingly strong case. For unlike a thread, an argument that is invalid or unsound does not have any ‘strength’ which could then be added to other arguments, the sum of which would amount to a stronger case for one’s view. If, for instance, Tooley were correct that viability, spontaneous  
movement, and human form do not distinguish mere human beings from persons, then putting viability, spontaneous movement, and human form together would not make the gradualist case any stronger ... A thread has some strength: a bad argument has none.
But this is a misleading analogy. In effect, it claims that the complex moral quality of being a person cannot supervene on a cluster of simpler properties if it can be shown that they are not individually necessary for personhood. But any number of complex properties seem to be constituted thusly. Let us take, as just one example, the concept of being a friend. The property of being someone’s friend appears to supervene on a cluster of base properties none of which is individually essential. The core constitutive features of friendship are qualities like loyalty, regularity of social contact, enjoyable companionship, shared interests or ideology, and mutual care and concern. It would be hard to make the case that any one of these properties is individually essential for friendship; counter-examples could, surely, always be summoned to prove otherwise. Friends can be distant, ideologically antagonistic, no fun to be around—even disloyal, up to a point. When it comes to friendship, this goes to show that a property is not conceptually redundant merely for being inessential. Loyalty is still a constitutive feature of friendship, even if one can adduce a real or hypothetical example of a friend who does not typically behave thus. Doubtless, one can rightly rebuke a treacherous friend for behaving in an unfriendly manner—for failing in his capacity as a friend. This only reinforces the rightful place of loyalty in our cluster concept of friendship, whether or not it is a necessary condition.
Consequently, the entire thrust of the reductio-type argument against developmental accounts of personhood might well fail to appreciate that the test for conceptual salience regarding a moral quality such as personhood is not whether or not we can manage to adduce instances of persons that are lacking in the relevant feature, but whether the absence of that feature is a deficiency that is explained by reference to what it means to be a person. Emotional ineptitudes like the inability to empathize, or cognitive ones like the inability to use language or reason, are, as Kaczor rightly says, deficiencies in human beings but not in cats because they amount to the failure to realize characteristics quintessential of persons. This is exactly how we know that they are constitutive elements of personhood.
In fact, the same seems to be true of every putative criterion of personhood on Warren’s list. Human beings that lack rationality, agency, communication ability, and so forth may be persons nevertheless, but the failure to realize any of these capacities is without doubt a deficiency relative to what it means to be a person. The only trait of which this does not appear to be true is human species membership. The fact that this is so can be demonstrated by considering the simple ‘intelligent alien’ hypothetical that we have already encountered elsewhere. Consider a member of an intelligent alien species which resembles typical human beings in all of their emotional, social, and cognitive capacities, with the one difference that in the place of human DNA, there is alien genetic coding. It does not appear that the alien’s different biology is any kind of deficiency by reference to what it is to be a person. We would all agree, I believe, that the alien is every bit as much a full instantiation of a person as the typical adult human—that nothing whatsoever is taken away by the pure fact of his different species membership. This seems to establish a good prima facie case for favouring developmental criteria such as that propounded by Warren over the human species membership criterion supported by defenders of personhood-from-conception. On the very same test that shows psychological capabilities to be salient to the concept of a person, human species membership appears to fall down.