The Doctrine of the Plurality of Rationalities and the Polite Bystander’s Perspective

The claim that the answers we give to many of the central questions in genethics will depend crucially upon the particular rationality we adopt in addressing them is central to Matti Häyry’s thorough and admirably fair-minded book, Rationality and the Genetic Challenge. That claim implies, of course, that there exists a plurality of rationalities, or discrete styles of reasoning, that can be deployed when considering concrete moral problems. This, indeed, is Häyry’s position. Although he believes that there are certain features definitive of any type of thinking that can accurately be labeled rational,Footnote ¹ he maintains that nothing about that set of features compels us to conclude that there is a single rationality. What is more, and significantly for the way in which Häyry’s book develops, there is no Archimedean point from which we are licensed to pronounce one flavor of rational deliberation to be intrinsically superior to any other or to be justified to the exclusion of all others. To this belief that “there are many divergent rationalities, all of which can be simultaneously valid,”Footnote ² we can perhaps give the name “the Doctrine of the Plurality of Rationalities” or, for short, “DPR.”

Whether DPR ought to be accepted is, of course, a big question and one that is far too weighty adequately to be tackled in a short article. Accordingly, in this paper, I leave it to one side and instead address a much more parochial issue, albeit one that is intimately connected with the claim made in DPR. To see what that issue is, let us first suppose for the sake of argument that DPR is uncontroversially true. It would appear to follow from that doctrine, as Häyry concedes, that there is a somewhat reduced hope of fruitful dialogue between positions on concrete bioethical problems wherever those positions are grounded in disparate rationalities. And where each party in a bioethical disagreement continues to insist on the universal applicability of her own style of rationality, “the result is a heated doctrinal shouting match camouflaged as a dispute over what makes sense and what is reasonable.”Footnote ³ A largely quietistic, or as Häyry puts it “non-confrontational,” response to the existence of nonconvergent, variegated rationalities and their differing conclusions might seem to be the only sensible one.

This may be unfortunate to the extent that it stymies attempts to reach definitive answers about what we ought to do in relation to a particular genetic technology. Nonetheless, it has the merit of being realistic: where no agreement can, in principle, be reached, it is as well for us to acknowledge this. And Häyry is, at any rate, not entirely pessimistic when it comes to the possibility of common ground and productive dialogue between rationalities. Such dialogue might, for the most part, be beyond reach, but he would appear to think there will be occasions on which at least some meaningful communication and progress can be achieved. So, in characterizing what he calls his polite bystander’s perspective on the particular genethical issues surveyed in the book (where the polite bystander’s perspective is to be understood as a stance of “reflective equipoise” between rationalities), Häyry tells his readers, “I will not intentionally take sides in the issues that I analyse, except in cases in which I think that a solution could be accepted by all parties.”Footnote ⁴ In other words, it seems that we ought not to stand politely by with respect to a bioethical debate until two conditions have been met: (1) the participants in the debate are operating with different rationalities and (2) there is no chance of their reaching, from within their differing rationalities, a mutually acceptable solution.

Again assuming the truth of DPR, this seems to me to be eminently sensible. It is methodologically sound, in the interests of achieving practical guidance on genethical issues, to delay for as long as we can the moment when we admit the in-principle futility of further debate. If DPR is correct, then much of the time we will only be delaying the inevitable. But this may be preferable to our precipitately adopting the polite bystander’s perspective and wrongly declaring that two or more entirely separate rationalities are, or need be, at work in a debate and that no common ground is possible between them. This brings me onto the central topic of my paper.

In Chapter 3 of his book, Häyry examines two views, one advanced by Julian Savulescu and the other by me, on the issue of whether a prospective parent has a moral responsibility to select the child, from a range of her possible future children, who is likely to lead the best life. More precisely, he takes my view (that we have no such obligation) and Savulescu’s view (that we do have such an obligation) to proceed from distinct rationalities, though he at least tentatively suggests that the two views may be more compatible than they appear at first sight. In the sections that follow, I argue first of all that the potential for a rapprochement that Häyry seems to detect in the two views is, in fact, not present. I then go on to urge, significantly, that once this possible source of agreement is gone, there is still no need to adopt a polite bystander’s view with respect to the debate. If Savulescu and I were to abandon our discussion as soon as Häyry’s suggested rapprochement is seen to fail, we would be deserting it far too soon, even if our positions really are rooted in separate rationalities.

Briefly, I maintain that this is so because there is implicit in my view a holism about reasons for action that will arguably be a definitive feature of any system of rationality and so also of the “mathematical rationality” distinctive of Savulescu’s work. Reason-holism is the doctrine that any consideration C that is able to function as a reason for action will invariably be sensitive to the influence of other considerations and background conditions that make up the context in which C occurs. The important consequence of this is that, from the fact that a consideration functions as a reason when it occurs in one context, it does not follow that it will function as such a reason in every other context in which it occurs. What operates as a reason to perform an α-type act in situation S₁, then, might not operate as a reason to perform an α-type act in situation S₂, even in the absence of any countervailing reasons being operative in S₂. This is the picture of reason-holism that Jonathan Dancy paints in his book Moral Reasons,Footnote ⁵ and it is plausibly a universal feature of reasons, even if DPR is true. If that is the case, then Savulescu ought to accept it, and the debate remains open (though unresolved) about whether what functions as a reason for choosing people who will have the best lives in many contexts still operates as such a reason when the people doing the choosing are prospective parents. I maintain that it does not. Savulescu will doubtless persist in holding that it does. But so long as we both suppose that reason-holism is true—which, again, even if Savulescu does not, he perhaps should—meaningful dialogue can continue.

In the next section, in order to provide some necessary background, I briefly sketch both Savulescu’s position and my opposing view. I am not concerned here to argue for my own stance or against Savulescu’s.Footnote ⁶ My sole aim in the current paper is not to say which of the two views is right, but simply to show that, if we embrace holism about reasons, we need not worry about whether the two positions belong to separate rationalities. The debate between Savulescu’s view and mine can remain live and even hold out the (admittedly distant) prospect of a resolution. In short, I show that Häyry would be being prematurely polite if he were to suppose that the failure of his attempted rapprochement signals the end of fruitful discussion between the two views.

The Principle of Procreative Beneficence and the Role of the Prospective Parent

In a landmark paper published in 2001, Julian Savulescu outlined and argued for a moral principle that he labeled the principle of procreative beneficence. His explanation of that principle has since become surely, and justifiably, one of the most frequently quoted passages in the literature on reproductive ethics:

The thinking behind the principle is reasonably straightforward and intuitively appealing. Suppose that a prospective parent is faced with a choice of two embryos, A and B, and that A is likely to develop into a child with a better life than the child who would develop from embryo B. Assuming, controversially, that it could ever be possible reliably to make such judgments,Footnote ⁸ it might seem clear that we have a moral responsibility to choose A over B. If we have the choice to bring into existence a better life or a worse life, then we ought to produce the better life. To do so is to make the world a better place than it would otherwise be, and this is something we should do wherever we can. If we flout the principle of procreative beneficence and choose B over A, our choice will be morally wrong.

We might, admittedly, disagree with this conclusion if we are persuaded by the claim that if any act is to be wrong, it must be person affecting, or such as negatively to affect some definite individual or group of individuals. After all, choosing B over A, or the child who will have the worse life over the child who will have the better life, does not negatively affect any particular individual. It does not give B a worse life than she would otherwise have had, because if the choice had not been made, she would not have existed. Nor does it negatively affect the child who would have developed from embryo A. That child has not been denied a life: she does not exist to have her life denied her. To think otherwise is to commit the mistake of supposing, as Jonathan Glover puts it, that possible people not brought into existence are like would-be immigrants denied entry to a country. But, as Glover urges, “the immigration queue model is inappropriate here. People who are kept out by restricting immigration exist. Those kept out of the world . . . are non-existent.”Footnote ⁹

However, it seems that there are good reasons not to be persuaded by the view that all wrongs must be person affecting.Footnote ¹⁰ Not least among those reasons is the existence in the literature of a number of compelling thought experiments detailing scenarios in which it seems counterintuitive to withhold the judgment that morally wrong choices can be made, even though those choices will not make any individual or set of individuals worse off than they would otherwise have been.Footnote ¹¹

Perhaps the most direct way of displaying the central thought behind Savulescu’s principle is by appeal to a point that he makes concerning reasons for choice or action. Suppose that the child who would develop from A would be free of all genetic abnormalities, whereas the child who would develop from B would be predisposed to asthma. If this information is revealed to us through genetic testing, we now have a reason to choose A over B, but no reason to choose B over A.Footnote ¹² Because, according to Savulescu, morality requires us to do what we have most reason to do,Footnote ¹³ we are morally required to choose A.

In a 2006 paper responding to Savulescu’s article,Footnote ¹⁴ I concede that there is, for the most part, a responsibility to make choices that will result in the existence of a person or set of persons with the best life or lives possible. I do not, therefore, hold that all wrongs must be person affecting. Nonetheless, I argue (with qualifications) that no such obligation exists when the decisionmaker is a prospective parent deliberating about which out of a range of possible children to bring into existence. This is because I believe that a prospective parent, just by dint of her filling that role, has, for example, absolutely no reason to choose embryo A over embryo B, provided that both A and B would develop into children with lives worth living. There is something about the perspective appropriate to a prospective parent that strips of their reason-giving force considerations about which possible child will lead the best life. This perspective I label “internal,” concerned as it is only with the future child’s estimation of his or her own individual life; I contrast it with an external perspective that is concerned to compare two or more possible lives from, so to speak, the outside and to pronounce on which is best. When we adopt the internal perspective, the sorts of reasons for choosing one child over another that show up from the external perspective are nowhere to be seen. From the perspective most fitting for the prospective parent, they simply do not exist.

There is no need for me to restate my arguments for this view here. The important claim so far as the current paper is concerned is one that is implicit in those arguments: it is possible for considerations that function as reasons for action in one context (say, the making of public policies that will determine whether people with better or worse lives will be born in the future) not to function as such reasons in another context (when, for example, a prospective parent is selecting her future child).

Room for Agreement between Savulescu and Herissone-Kelly? Häyry’s Suggested Rapprochement

In the spirit of striving to find points of agreement between the proponents of what he sees as distinct rationalities, Häyry makes the suggestion that Savulescu’s view, once its finer points have been considered, is not as far from my own as it might initially seem. He writes:

It should be plain that Häyry is right in claiming that Savulescu’s principle of procreative beneficence embodies a judgment made from the external perspective. To assess Häyry’s suggestion that there may be further agreement between me and Savulescu, then, we need to get clear about the nature of the concessions Savulescu makes to parental decisionmaking. That done, we can determine whether they truly do converge with my “internal considerations.” I aim to show that they do not. My doing so will set the stage for a brief explanation, in the next section, of reason-holism and the manner in which my opposition to the principle of procreative beneficence might be seen as relying upon it.

The first thing to note is that it is not unequivocally the case that Savulescu’s concessions to parental decisionmaking entail that it would be morally acceptable on occasion to select less than the best. So, for example, Savulescu maintains that it is required of society that it allow parental autonomy. The upshot of this is apparently that parents should not be legally obliged to choose the child likely to lead the best life, even though they remain morally obliged to do so. In other words, Savulescu’s claim here is not that a principle of reproductive autonomy ever makes it morally permissible for parents to ignore the principle of procreative beneficence. Rather, his view seems to be, to put it pithily, that society has a moral obligation to permit prospective parents not to fulfill their moral obligation to be guided by considerations of procreative beneficence.

On the other hand, Savulescu seems to suggest that a couple with dwarfism who select a child with the same condition, on the grounds that their house is set up for dwarfs, may be acting in a morally permissible way.Footnote ¹⁶ If this is what Savulescu is saying, then the thought is probably that the couple have most reason to select a child with dwarfism (recall that for Savulescu, morality requires us to do what we have most reason to do). Here, the reason to select the child likely to have the best life is outweighed by a stronger reason. As Savulescu writes, “Procreative Beneficence is a valid principle, albeit one which must be balanced against others.”Footnote ¹⁷

But there is something noteworthy about this last claim that makes it significantly different from the judgments that parents are likely to make from the internal perspective. Savulescu’s talk about exceptions to the principle of procreative beneficence is couched in terms of reasons being balanced against each other and of a reason to select the child with the best life being outweighed by other, more pressing reasons. However, in what we might call a pure case—where there are no other principles in operation that carry greater weight than that of procreative beneficence—he maintains that we have a reason to choose a better-life embryo A over a worse-life embryo B and no reason to choose B over A. That being the case, we are morally required to choose A over B. Note that the case of the couple with dwarfism is not a pure case: even assuming that a child with dwarfism will be likely to lead a worse life than a child without dwarfism (an assumption that, for the record, I do not share), the prospective parents still have good reason to choose a child with the condition. That reason is able to act as a counterbalance to the reason the parents have to select the child likely to possess the best life.

My position is in stark and irreconcilable contrast to Savulescu’s view of pure cases. I do not hold that the prospective parent has a reason to select A but an equally strong reason to select B, thus rendering it a matter of indifference which she chooses. Instead, I maintain that even in a pure case she has no reason at all to select A over B and, for that matter, no reason at all to select B over A. It is the absence of reasons on both sides, and not the presence of equally weighted reasons on either side, that makes it the case that it does not matter morally which embryo she chooses. So, even though Savulescu apparently agrees that there can be occasions when the right thing to do is to choose a worse-life embryo over a better-life embryo, he (a) does not think that this would ever be a permissible thing to do in a pure case and (b) thinks that there is always a reason for a prospective parent to choose the better-life embryo, even though that reason might in impure cases be overridden by stronger reasons in favor of the worse-life embryo. I disagree with him on both counts. So, I am able to say:

I suppose it would be possible, but mistaken nonetheless, to interpret me as saying something subtly, but importantly, different to this and something that would more nearly approximate the concessions that Savulescu makes. I might be thought to be holding that, while there is a reason for choosing A over B in the shape of A’s better prospects in life, this reason is overridden by a stronger reason for not choosing either embryo in preference to the other. That stronger reason might be characterized as the fact that the decision maker is a prospective parent, and it is appropriate for prospective parents not to favor either potential child. Although this position clearly has something in common with my view, it does not accurately capture what I think about prospective parents’ reasons for action or the lack thereof. According to my model, that someone is a prospective parent is not a consideration that features in the deliberation that she ideally undergoes when faced with a choice about which child to bring into existence. It is not, therefore, a reason for action. Rather, it is an explanation of her inhabiting a certain perspective, from which there is seen to be no reason to prefer A to B or B to A. It is a position from which she deliberates, rather than a consideration that enters into her deliberation. In addition, it is, I think, wrong to portray the prospective parent as having a reason for not choosing one embryo over the other. Instead, I want to say that she has no reason for choosing one over the other. This distinction is a delicate one, but is of the highest significance.

Because Häyry’s attempt to find an overlap between my position and Savulescu’s has, if I am right, turned out to be something of a cul-de-sac, are we forced to conclude that the possibility of dialogue between the two viewpoints is similarly doomed? Must we stand politely by and acknowledge that each approach is sensible, but only by the standards of the particular rationality from which it proceeds, and that it would be misguided for either one to try to engage meaningfully with the other? I do not think so. I think instead that even if we do consign each to a separate rationality—and, incidentally, I remain for the purposes of this paper agnostic on the question of whether we should—there is still the opportunity for fruitful debate between them. This is because, as I mentioned earlier, I think that anything that can qualify as a rationality will be subject to a holism about reasons, because I take it that all reasons for action are subject to such holism. Any consideration that can function as a reason for action is sensitive to the influence of other considerations and background conditions, and those considerations and conditions determine whether it functions as a reason here, in this particular context. And holism about reasons entails the existence of a phenomenon that Jonathan Dancy and others call “silencing.” That is, it is possible for a consideration that stands as a reason to perform α-type actions in one context to be no reason at all to perform such actions in another context. In that other context, the consideration is silenced by the presence of other conditions—it is rendered entirely inoperative as a reason, rather than outweighed by a more powerful reason. Indeed, according to Dancy, the consideration in question might in certain contexts function as a reason not to perform α-type actions or as a reason to perform β-type actions instead, though I am not, in this paper, directly concerned with these last two possibilities.

A Brief Account of Reason-Holism

My own view that a prospective parent has no reason to select a better-life embryo over a worse-life embryo relies on the existence of a holism about reasons and the attendant phenomenon of silencing. That is, I think that for the most part—in most contexts—there exists good reason to take that course of action that will result in the existence of better lives, even where those lives are not better for any set of particular individuals. However, I think that the consideration that usually functions as a reason in such cases is silenced when the person selecting fills the role of prospective parent. The consideration that operates as a reason elsewhere does not function in the same way here: it is silenced by facts about the identity of the decisionmaker.

Savulescu is centrally concerned with rationality. If reason-holism, being universally applicable to reasons for action, is a definitive feature of anything that can qualify as a rationality, then it must be a feature of Savulescu’s rationality too. In this section, I briefly set out the reasons for believing that reason-holism is generally true.

In Moral Reasons, one of Dancy’s primary aims is to deny what he calls the universalizability of reasons, understood as the view that what operates as a reason here must likewise operate as a reason in all circumstances or even in all relevantly similar circumstances. But the claim, as we have seen, is not simply that reasons that are decisive in one context can be outweighed by more powerful reasons in another context. Rather, it is that reasons that are decisive in one context can be silenced—have their status as reasons removed—by the considerations and background conditions present in another context.

The intuitively compelling nature of this claim is, I think, powerfully brought out by an example supplied by Dancy. He recalls hearing R.M. Hare claim that cruel, pain-inflicting actions from which their perpetrators gain satisfaction have at least this to be said for them: that they give pleasure to those who carry them out.Footnote ¹⁹ Hare’s thought seems to be that the pleasure to be gained from the performance of such actions is some sort of consideration in favor of their being performed, some sort of reason for executing them, albeit one that is massively outweighed by the reasons against those acts (that they are cruel, that they cause pain, and so on). Dancy, I think rightly so, finds this claim to be quite startlingly wrong. He wants to maintain that, in the case of such acts, the fact that they give pleasure to their perpetrators is no reason at all—not even an invariably overridden reason—in their favor. Although the fact that an act will give pleasure to the person who performs it is in many contexts a reason why that act should be performed, in this sort of context, it is silenced by the presence of other considerations that function as reasons against performing the act. Indeed, in the case of cruel actions, the consideration that they give pleasure to their perpetrators plausibly functions as a further positive reason against their performance. So, what frequently functions as a reason for an action here functions as a reason against.

Now, the sort of silencing outlined in this example involves a consideration that can function as a reason for action in some contexts, being silenced by other considerations that function as reasons against action. But this, Dancy thinks, need not be the only way in which silencing occurs. Considerations can also be silenced by background conditions that do not themselves function as reasons, or motivators. This is precisely, I submit, what happens when the person required to choose which of two possible people to bring into existence, where one will be likely to have a better life than the other, is the prospective parent of both. There is not some further reason for action, peculiar to the context, that silences the default obligation to bring into existence the person who will have the better life. Rather, that default obligation is silenced by the background conditions of the case, in the shape of the identity of the person called upon to do the selecting. That embryo A will be likely to develop into a person with a better life than the person who would develop from embryo B provides no reason at all for a prospective parent to choose A over B. Nor is there any reason to select B over A.

Again, I am not concerned here to rehearse the arguments for such silencing occurring in the case of prospective parents faced with a choice between possible children. I have supplied those elsewhere. I simply want to point out that my stance, and my opposition to the principle of procreative beneficence, relies on the existence of the phenomenon of silencing and to urge that the holism that makes silencing possible is a universal feature of reasons for action.

Postponing the Adoption of the Polite Bystander’s Perspective

Once I have supplied a defense of my position that is rooted in reason-holism and the possibility of silencing, Savulescu has open to him two potential responses to it. First, he can deny reason-holism. If holism about reasons is unwarranted, then silencing cannot occur, and if silencing cannot occur, then the default reason in favor of bringing into existence, from a range of possible people, those who are likely to live the best lives will be operative. His second alternative would be to admit that reason-holism is true and that silencing does sometimes take place, but that it happens not to occur, as I argue it does, in the case of prospective parents choosing their future children.

It seems to me that Savulescu would be ill advised to embrace the first alternative. If the style of rationality employed by Savulescu is to be taken seriously, it must be sensitive to the way in which reasons actually function in everyday life. And, though I do not here have the space to argue for this claim in any detail, I think that they do, in fact, display the sort of plasticity that Dancy attributes to them. Indeed, in allowing a role, albeit a rather diminished one, to general moral principles, Dancy makes a claim that might equally be used to defend his own particularism:

I think there would be a fault in any philosophy that tried to deny the sensitivity of reason to context and that tried to suggest that certain features of contexts are not capable of silencing considerations that, elsewhere, operate as reasons. That, it can be argued, is just not how reasons work.

Of course, this leaves open for Savulescu the second alternative: he can argue that I am wrong in thinking the default consideration in favor of selecting people with better lives is silenced when it is a prospective parent doing the selecting. But note that, if he takes this approach, the debate is still live, and the possibility of engagement in meaningful dialogue is still open. Although we may not convince each other, he and I can each put forward our reasons and hope for progress. Of course, if DPR is correct, it might turn out (1) that Savulescu and I are working from within separate rationalities and (2) that those rationalities preclude from the start our agreeing over whether or not silencing occurs where I say it does. But whether both those conditions are met is not something that can be decided in advance of our attempting the debate. Häyry has no need to stand politely by just yet.