1. Introduction
In 1977, John Taurek published an infamous article, “Should the Numbers Count?” In it Taurek argued for the radical claim that, when faced with the choice between saving two groups of people from death, the number of people saved simply didn’t count. Few philosophers agreed with Taurek, and his article spawned a literature of criticisms. At this point, the debate over Taurek’s article seems to be played out, with the available ways of disagreeing with Taurek fully charted and entered into the record.Footnote 1
This is a paper on John Taurek and his infamous article. Why another paper on Taurek? This paper takes a step back and approaches Taurek’s arguments from a higher level of abstraction, in the context of a topic that has received much less sustained attention: the idea of a duty of love. My claim is that when Taurek’s arguments are read as an extended articulation of the logic of the duty of beneficence, understood as a duty of love, many of the more puzzling features of these arguments cease to be puzzling, the intuitive force of some standard criticisms disappears, and new questions are raised. Further, the relationship between different strands of Taurek’s argument is clarified, and some core claims appear as logically separable in a way that has not been appreciated.
The title of this paper poses a question: what is the point of a duty of beneficence? This question might seem strange. Understood as a duty, roughly, to meet human need and provide help – to benefit others – it is an obvious part of commonsense morality.Footnote 2 Its point is that the needs of others have moral weight. If we think of beneficence as a duty of love, however, things become a bit more complex. We need to ask a related question: what point does love have in ethics? For many philosophers, the answer is: none. Morality is about the public sphere and love is about the private sphere. Kieran Setiya has voiced his dissatisfaction with this: there is a tradition of thinking of ethics as a kind of universalized love of the other (2014: 251–52). If love is to have a place in ethics, then, we need to understand how ethics could in some way be grounded in such love.
I begin the paper in §2, then, with two important dissenting voices – two papers that attempt to put love back at the heart of ethics: J. David Velleman’s “Love as a Moral Emotion” and Setiya’s own “Love and the Value of a Life.” The papers move in opposite directions – Setiya arguing that the partiality characteristic of love has a place in ethics proper, and Velleman arguing that love is in fact importantly impartial in ways characteristic of the more obviously moral attitude of respect. Both philosophers’ results are limited, however: Velleman shies away from connecting love to reasons for action at all, and Setiya establishes the narrow conclusion that certain dispositions characteristic of love are rational to have towards any human. However, I think that the papers ultimately represent complementary attempts to think through what it would mean to ground love in moral value and they contain many useful resources for pushing this idea further. Their limits highlight the possibility of conceiving of the duty of beneficence as grounded in love, insofar as the value that grounds the duty of beneficence is one that is grasped in loving another person. The duty, I think, can be understood as explained by the universal importance of human need as it is understood in empathetically engaging the perspectives of particular people.
I think that Taurek’s paper goes a long way towards articulating what a duty of beneficence so conceived would look like, by drawing on an ordinary conception of how empathy works and spelling out some of its implications. I attempt to show this in §3. Section §4 draws some lessons. Taurek’s critics have by and large not recognized what I argue is his implicit strategy, and because of this many of their criticisms fall flat. But Taurek himself fails to reckon with important questions about the point of a duty of beneficence, so conceived. He fails, I think, to grapple with the question of beneficence’s scope. And he fails to fully consider the ways in which others have value that transcends what can be grasped through empathy. Part of understanding the point of love in ethics – and the point of a duty of beneficence – is working out how the value grasped in love could be a moral value at all. But part of understanding the point of love in ethics is articulating its limits. Reevaluating Taurek will, I hope, allow us to see how to move beyond Velleman and Setiya, but it will also show us what questions remain to be answered.
2. Love and morality: two contemporary views
This section is both critical and constructive. I draw out some key ideas from the work of David Velleman and Kieran Setiya that will enable us to appreciate the role that love might play in morality. In the course of doing this, I will be critical of some aspects of both philosophers’ views, but my ultimate goal is to show how their pictures point us in a certain direction that allows us to view Taurek’s arguments with fresh eyes.
In “Love as a Moral Emotion,” Velleman aims to show that love is not antithetical to the impartial spirit of morality. Love should be understood as a distinctively “moral” emotion insofar as it is a response to the very same value – the value of the person – to which respect, the paradigmatic moral attitude, is also a response (1999: 344). Velleman draws on Iris Murdoch’s notion of “valuation-as-vision” in articulating both respect and love (341–44). Both respect and love are attitudes or emotions that involve, in some way, really looking at the other. The two attitudes are distinguished by their contrasting positive and negative characters. Respect is an awareness of the value of personhood that arrests the kind of self-love that leads to selfishness. Love is an awareness of the value of personhood that disarms the kind of emotional defenses that block us from getting involved with and opening up to others (360–62).
Velleman’s paper involves an extended reading and reconstruction of Kant’s ethics. I’m not interested here in whether Velleman gets Kant right. I am interested, however, in a connection between Kantian respect, as described by Velleman, and moral duties – a connection that Velleman doesn’t emphasize. As noted, Velleman understands respect as a kind of awareness of the value of the other that plays a certain motivational role; in truly recognizing the value of the other I am moved in certain ways. Certain selfish motives fall away, or I am motivated to combat such motives, or I am put on alert to be wary of the possibility of selfish motivation, etc. Respect is also, as Velleman is aware, at the heart of a system of duties for Kant. We have duties of respect – not to, for example, lie and murder. These two aspects of respect – as a kind of motivational awareness of value and as at the heart of a system of duties – are naturally understood as connected. Velleman’s understanding of respect as a motivational awareness of value invites a particular interpretation of how respect lies at the heart of a system of duties. The value of the other that I grasp in the awareness that constitutes respect is, ex hypothesi, not a value that should simply arrest my selfishness on this particular occasion. What I am supposed to have become aware of is a value that arrests my selfishness precisely because it is a value that all persons have; recognizing another through respect is supposed make my self-interest seem petty because it is an awareness of some particular person’s value as that of the person as such. This awareness, then, is of a value that can – with a bit of imagination and abstraction – underwrite a system of negative duties. Duties against lying and murder, for example, are prohibitions on actions that, in recognizing this value as limiting, one recognizes as transgressing the limits. They are duties whose discharge, ideally, would proceed from the kind of feeling of recognition of the other that puts a halt to self-interest – that is: respect.
I think that this is a compelling picture of how recognition of value, emotion, motivation, and moral theorizing are interconnected. Further, all of the moves just sketched suggest similar moves for love. After all, morality is not exhausted by the negative. We have, for example, duties of beneficence – duties to offer help, to meet human need. (Indeed, Kant complements his duties of respect with duties of love.) Love is, on Velleman’s account, an awareness of value that emotionally disarms us and draws us closer to the other. Why not tell the following story? In being emotionally disarmed by the other and recognizing that what we are responding to is a feature possessed by all, we come to the recognition of grounds for duties to be involved in the lives of others more generally. Duties of beneficence and mutual aid are duties that, ideally, would proceed from the kind of feeling of recognition of the other that disarms our emotional defenses – that is: love
This is not the route that Velleman takes. He thinks of respect as a mandatory minimum and love as an optional maximum response to the same value: that of a rational agent (366). As optional he doesn’t think of love as connected to something like a system of duties. This leads Velleman to puzzling places: he denies, for instance, that Bernard Williams’s man who saves his wife from drowning is best understood as motivated and justified by the reasons of love as theorized throughout most of the paper, but rather by reasons stemming from his relationship (372–73). Divorcing the reasons of love from reasons for action in this way certainly blocks the grounding of moral duties in love, but it is not clear that this divorce is required. Kieran Setiya, while sharing much with Velleman, also finds this move puzzling and rejects this divorce between reasons for the emotion love and reasons for action; it is to his “Love and the Value of a Life” that I now turn.
Setiya’s paper plays a complex dialectical role here. We might simply bypass it, as the position that I find in Taurek is essentially a concrete development of Velleman’s path-not-taken just sketched. Setiya’s paper is significant because it develops a picture that has many affinities with Velleman’s – but which crucially does connect love with reasons for action in a way that makes it suitable, at least in principle, to ground moral duties. Setiya, however, does not make use of apprehension of value – doesn’t follow Velleman’s path-not-taken as I’ve called it – and stops short of grounding duties in love. In order to see the prospects and difficulties of the view I find in Taurek, I think it is useful to sketch Setiya’s alternative, raise some worries about it, and flag some of its limitations.
To start, Setiya’s basic position is that “another’s humanity is sufficient reason for love” (2014: 276). With Velleman, Setiya thinks that there are reasons for love – that love can be rationally assessed – and that reasons stemming from qualities or relationships are not necessary. Further, in making humanity sufficient reason for love, Setiya also follows Velleman in making love suitable to play a role in ethics in two interconnected ways. Although Setiya writes of “humanity” where Velleman writes of “rational personhood” – and takes this debate to be perhaps more than terminological – he flags that he thinks of love, like Velleman, as justified by whatever property gives human beings moral status (262). An upshot of this fact – for both Setiya and Velleman – is that love is not meritocratic (as on views that justify it on the basis of good qualities) or private (as on views that justify it on the basis of personal relationships).
Setiya’s major divergence from Velleman comes at this point: regarding reasons for action. Setiya appreciates the force of some of Velleman’s worries about the explication of love solely in terms of the desire to benefit another (253). Nevertheless, Setiya is skeptical that criticisms of views that assimilate love to the desire to benefit show that the reasons for love as an emotion and the reasons for action motivated by love diverge sharply (265, fn25). Setiya concludes – plausibly in my view – that a common element of love is “vulnerability to the needs and interests of another human being” (253). This sort of vulnerability, further, involves, in part, the disposition to accord the beloved’s needs a special weight (267–68). The disposition in question is the one that is paradigmatically manifested in cases like Williams’s husband saving his wife from drowning.
Setiya argues that such a disposition can be had rationally towards any human being, justified simply by the target’s humanity. He gives two main arguments. The first is simply an application of the conclusion of earlier arguments against alternative views about the reasons of love – arguments which lead him to the view that the humanity of its object is sufficient justification for love. Given that love involves the sort of disposition at issue, such a disposition is justified in the same way that love is generally (264–65). The second is an independent argument, considering the kinds of features of the world a disposition needs to be sensitive to in cases like Williams’s, for the conclusion that such a disposition is rational if it moves from the fact that a human is in need and that the situation is not extraordinary to a saving action, without needing any of the other aspects of love to justify it (265–70).
The arguments just sketched require a good deal of unpacking to do them justice, but I am not going to do that here. For my purposes, what is important is to appreciate the extent to which such arguments establish a conclusion which amounts to giving love a role in ethics. The upshot of Setiya’s argument is that the disposition from which, for example, a husband may rationally act in saving his wife from drowning – a disposition that manifests his love for her – might be rationally had and manifested as well by a complete stranger. This upshot is, in a way, a vindication of Taurek’s infamous idea that the numbers don’t count. I need not save the greater number even when all are strangers.
In what sense does this conclusion amount to giving love a role in ethics? To answer this question, I want to make two general points about Setiya’s view – one about its narrowness and one about the downsides of eschewing the sort of appeal to apprehension of value that Velleman relies on. First: the narrowness of his conclusions. Setiya’s arguments do establish a morally interesting conclusion: that morality doesn’t demand we save the greater number, ceteris paribus. And Setiya points out that this has implications for significant moral issues:
Even if your money would save more lives in distant countries, it does not follow that you should give to Oxfam rather than to local charities or individuals in need, not because there is insistent reason to favor those who are close to home, but because it is rational to give priority to anyone. (278)
Finally, Setiya thinks that his account gives insight into the “irreplaceable” value of a human life, such that one might think of his reflection on love as giving insight into the value of the thing that undergirds moral status (275–76).
However, Setiya also stresses the narrowness of his conclusion, especially in contrast to Taurek and contractualists who have drawn some inspiration from him. He points out that although he has argued that a certain disposition characteristic of love is rational to have towards anyone, he doesn’t think the disposition is required. For all he has committed to, it may be rational to save the greater number simply because they are greater (272). He notes that his argument for the rationality of the disposition has not turned on criticisms of aggregation or the impersonal good. As such, there may be circumstances where the aggregate or impersonal good make it impermissible to save the one. All Setiya has got us is the conclusion that there are cases where it is rational to save, for example, one’s wife simply on the basis of the fact that she is human, and in those sorts of cases strangers would be permitted to do so as well – although they would not be obliged to do so (273–74).
Setiya is right to stress the limited basis of his arguments and the limited scope of his conclusions. He has pulled off the enviable feat of drawing out surprising conclusions from premises that are hard to resist. But the flip side of this coin is that a lot more work would be needed to see how love could play a substantive role in ethics – we’d need, for instance, to work through all of the factors that Setiya sets to the side in order to get a better sense of the shape of love in ethics as a whole.Footnote 3 This leads me to my second point. Although Setiya moves beyond Velleman, towards a position in which love could play a role in ethics, by drawing the necessary connections to reasons for action, he drops Velleman’s use of the notion of apprehending value through an emotion or attitude. This has to my mind an unfortunate effect on Setiya’s account. Without focusing on apprehension or awareness of value, Setiya has made the case that what seems to be a partial disposition is justified whenever its object is human. Without some story about how the disposition can be intelligibly understood as responsive to and motivated by an apprehension of a value had by all humans, however, the reader may be left suspicious and skeptical of an abstract argument. How can a disposition that gives one human’s interests special weight be responsive to anything other than the particularities of that human? If I started to share Setiya’s disposition towards his wife, what would have to be going on in my mind for me not to regard myself as having a mental breakdown?
In concluding his paper, Setiya seems to anticipate this worry and gesture towards a response (276–79). Using Taurek’s own case of a choice between saving strangers or David, whom you know and like, Setiya invites the reader to imagine their responses:
Try to picture it. On one side, three anonymous faces; on the other, David, whose face you know. You are not obliged to save him. Should you let him drown in order to save the others?… As I would put it, such claims are incomparable. It is not that there is a reason to favor those you know and like over those you have never met. It is rather that, in the absence of special conditions, there is sufficient reason to give priority to the needs of any human being. This is what you do for David, not irrationally, and what you might do for anyone else. (277)
Here’s a natural way of understanding the force of Setiya’s imagined exercise: if we reflect on what we are responding to in giving David’s interests a special weight, we should recognize that we are not being selfish or partial but rather responding to a value that we apprehend in David, because we know and like him, but which we recognize shows up wherever there is a person. This is how I am inclined to read it, anyway, and I am inclined to read this imagined exercise as returning us to Velleman’s basic ideas. Setiya is correct in thinking that if love should play an interesting role in ethics, this must be understood in terms of making room for recognizably loving actions as a part of morality. But, to understand why such actions count as responses to moral rather than some other private value, we will need to advert to the kind of awareness of value that ideally motivates those actions and understand how it could be anything other than partial – a kind of awareness of value that Setiya officially wants to eschew.
I don’t want to push this criticism too hard. I may be wrong in reading this passage and wrong in thinking that Setiya needs to appeal to apprehension of value to make his disposition intelligible. What matters to me here is to use this point to draw a contrast between Taurek and Setiya. On the reading of Taurek that I will develop, he manages to combine ideas that show up in both Setiya and Velleman in a novel way. Whether or not it is necessary to think in terms of the apprehension of value, this is what I think Taurek does in discussing empathy. Like Velleman, I will argue, he thinks of us as grasping a value that all humans have that cannot be aggregated. Unlike Velleman, though, and closer to Setiya, he is friendly to thinking that the value that we grasp in this way is connected to reasons for action; indeed what we grasp is that the needs of others matter in a way that is directly relevant to action – the kind of value that is suitable to ground a duty of beneficence. The idea that such a non-aggregative value is grasped in empathy gives Taurek more resources than Setiya but also commits him to more controversial conclusions. Ironically, however, I will argue in the final section that Taurek is not only more committal than Setiya, but more committal than he needs to be. Seeing this illuminates, I think, future work needed for thinking about love in ethics. But first, Taurek.
3. Taurek revisited
My reading of Taurek stems from the following conviction: the structure of his paper is very misleading.Footnote 4 It begins straightforwardly enough: Taurek introduces the idea of a trade-off situation – a situation in which we can bestow benefits on or prevent harms from befalling one of two different groups of people, but not both. In such situations, we must choose which group to help. Taurek tells us that his paper is about the following question: is the relative number of the people in each group in itself significant in deciding which group to help? And, Taurek tells us his answer, which amounts to the thesis of the paper: no (1977: 293–94).
Taurek then proceeds to approach the question posed in a manner familiar from papers in ethics: he poses hypothetical cases and appears to be pumping the reader’s intuitions in order to support his conclusion. But, the structure of the paper is misleading, I think, because although the form appears unremarkable, the content is strange. An intuition pump only works if your intuitions are shared by most of – or at least some important subset of – the people you are in dialogue with. But the intuitions that Taurek appears to want to pump are not only not widely shared – they have struck many readers as bizarre. Taurek is himself aware that his arguments involve crucial premises which he finds attractive, but which many readers will think false. The essay moves then to argue for the suspect premises, at first by additional hypothetical cases which themselves involve premises readers have found bizarre. The dialectic bottoms out, however, in reflection on human value apprehended thorough empathy. At this point, the ladder has fallen away. Taurek’s hypotheticals get us to the point where we can appreciate this reflection and it starts to look like perhaps the paper has been written backwards; Taurek’s hypotheticals are an illustration of thinking about human value in a certain way. This is how I want to read Taurek – backwards, with an emphasis on empathy.
3.1. The big picture: empathy
As a setup, here is a gloss on the arguments that get Taurek to empathy. (I’ll return to discuss these in more depth when we’ve appreciated the role that empathy plays in his thought.) In his first main argument, Taurek imagines a situation in which he can save the life of David, whom he knows and likes, or five strangers. Taurek thinks that he can save David. But knowing and liking someone surely can’t override an obligation to save the greater number. So, there is no obligation to save the greater number, even when all are strangers (295–99). But why think it really is permissible to save David? In his second main argument, Taurek points out that David could save himself if he had the power to do so. But if David could save himself, surely Taurek could as well. And if that’s right there is no obligation to save the greater number (299–301). But is it really so obvious that Taurek could save David? This brings Taurek to empathy.
Here is the key passage on empathy:
It seems to me that those who, in situations of the kind in question, would have me count the relative numbers of people involved as something in itself of significance, would have me attach importance to human beings and what happens to them in merely the way I would to objects which I valued. If six objects are threatened by fire and I am in a position to retrieve the five in this room or the one in that room, but unable to get out all six, I would decide what to do in just the way I am told I should when it is human beings who are threatened. Each object will have a certain value in my eyes. If it happens that all six are of equal value, I will naturally preserve the many rather than the one. Why? Because the five objects are together five times more valuable in my eyes than the one.
But when I am moved to rescue human beings from harm in situations of the kind described, I cannot bring myself to think of them in just this way. I empathize with them. My concern for what happens to them is grounded chiefly in the realization that each of them is, as I would be in his place, terribly concerned about what happens to him.… The loss of an arm of the Pietà means something to me not because the Pietà will miss it. But the loss of an arm of a creature like me means something to me only because I know he will miss it, just as I would miss mine. It is the loss to this person that I focus on. I lose nothing of value to me should he lose his arm. But if I have a concern for him, I shall wish he might be spared his loss. (306–07)
Taurek’s description here, it seems to me, is of empathy as an apprehension of the value of other people which is motivational – which moves us to respond to their plights. It is, that is to say, consonant with love as understood by Velleman. This is perhaps not surprising, as it is natural to think of empathy as a kind of love. But – as brief as his discussion is – Taurek is more concrete, fills in some more details, relying on our ordinary grip on the specific kind of love that empathy is. In what way am I appreciating the value of another in empathizing with them? I take up their perspective and recognize what matters to them from their perspective as significant in such a way that it should matter to me. And we get a little more traction by contrast with other ways of appreciating value. I may appreciate the value of the Pietà, and such appreciation may be motivational in all sorts of ways. But it can’t be by way of apprehending what matters to the Pietà as something that should matter to me, because nothing matters to the Pietà.
This sketch of empathy, however, raises two interconnected questions: how should we think of this apprehension of value as underwriting a duty of beneficence? And why should we think of the empathy that Taurek describes as an apprehension of value that is universal in a way that moral value is supposed to be? (Taurek is, after all, empathizing with David.)
I want to begin in answering these questions by noting that, whether he is justified in doing so or not, Taurek’s essay proceeds on the implicit assumption that the value he ends up articulating in terms of empathy is the value that grounds the duty of beneficence. Taurek does not himself mention the duty of beneficence explicitly. But, in posing the first case glossed above, Taurek assumes as obvious that we are required to save someone when confronted with human need. We have a duty to act on. Taurek is manifestly not worried about whether we are required to help, but rather about whether we are required to save the larger number in order to fulfill our duty.Footnote 5 So, the issue is what must be done to discharge our duty. And, although I will later worry that Taurek is too quick in implicitly lumping all helping actions together, we can postpone those complications and note: his topic is helping actions and so beneficence (broadly construed) rather than, for example, justice. In sum: I think it is an overlooked aspect of Taurek’s paper that his claim that we can save David rather than a larger number of strangers is a claim that we can discharge our duty of beneficence in doing so.
If this interpretation is correct, Taurek’s description of what I appreciate about David in empathizing with him must aim to show how someone could be responding to a moral value in choosing to save him rather than a larger number. This is crucial for two related reasons. First, as a moral duty, the duty of beneficence is supposed in some way to be grounded in moral value. If we can discharge this duty by saving David, it has to be because this would be a way of responding to that moral value. So, Taurek owes us a story of how someone could be saving David out of such an appreciation. Second, many philosophers who balk at the thought that we might save David rather than a larger of number of strangers think this precisely because they take the number of lives saved to be of moral importance – to be the moral value we need to respond to if we are to be genuinely beneficent.Footnote 6
However, I want to flag a crucial fact for readers who are not fully convinced that Taurek takes himself to be discussing the value that undergirds a duty of beneficence: the reading I will pursue gives an interpretation of Taurek on which he does give an account of a value grasped in empathy that could ground a duty of beneficence. If this is the case, it fits well with an overall reading on which he is discussing beneficence. And, if the reader is still not fully convinced as a matter of exegesis, I believe such a view raises philosophically interesting questions which will be discussed in the paper’s final section.
Complications aside: if Taurek is aiming to articulate a value that grounds a duty of beneficence, we should see in his discussion of empathy and related remarks connected answers to both of the questions just raised. However, these questions are not always clearly distinguished in Taurek, as exemplified in the passage quoted. Taurek’s response to his opponents is clear enough: Taurek claims that he can make no sense of the idea that saving a larger number of lives would be better overall. The value apprehended in empathy is of how things matter from another person’s perspective. But perspectives are not the sort of thing that can be coherently combined; I can empathize with many different people suffering by taking up their perspectives, but there is no perspective that does all the suffering with which I can empathize. But Taurek claims that he sees no other value that humans have that would make it particularly important to do anything at all by way of helping them. If we consider humans as objects – as like the Pietà – they are “nearly as common as toadstools” (307), perhaps not worth saving at all.
If Taurek is right in this line of argument, then he will have shown that the only available value to ground a duty of beneficence is the value apprehended in empathy. But we may remain dissatisfied. If persons had an additive value and that value were high, then we’d have weighty reasons and those reasons would look recognizably impartial: every person adds the same value into our calculations. But what am I apprehending in empathizing with David’s plight that can be understood as recognizably impartial? Taurek’s official description of his position here may give us further reason to worry, as he describes his motive for saving David as a “purely personal preference” (297).Footnote 7
To answer this question, I think we need to get a bit clearer on what is essential in Taurek’s position. The idea that we may save David rather than a larger number plays a major role in key arguments. But, once we see the role that empathy plays in Taurek’s overall view, I think that we should recognize that David is inessential. We can, that is, deny Taurek’s claim that we can discharge a duty of beneficence by acting on purely personal preferences while accepting the rest of the picture that he has developed. Seeing how this is so will show us how empathy – how love – can ground a duty of beneficence.
I want to focus on two elements of Taurek’s essay: coin-flips and the impartial character of his description of empathy. First, Taurek famously claims that if he were confronted with the choice of saving one or five strangers, he would, perhaps, flip a coin. Doing so would give every individual an equal 50/50 chance at survival (303). One may read Taurek here as simply exasperated and at sea, given that he can make no sense of human value as additive. He sides with a sort of equality, given that he is not sure what other value is at stake.
But I think that Taurek’s settling on a coin-flip fits naturally with his description of what is apprehended in empathy. Although Taurek describes his motive in saving David as one of purely personal preference based on the fact that he knows and likes David, his description of what he recognizes in reflecting on the plights of others suggests something else. Notice that what Taurek describes is not being moved by idiosyncratic features of those with needs. Rather: “My concern for what happens to them is grounded chiefly in the realization that each of them is, as I would be in his place, terribly concerned about what happens to him” (306–07). What he describes recognizing here is the importance of what happens to others as making a claim on him – something that is appreciated in taking up particular perspectives – but the claim is recognized as one that is not grounded in the particularity of any one perspective. In empathizing, I recognize the plight of a particular person as important in a way that should move me because they are a person, not because they are particular. If this is the value recognized, it would make sense of a duty of beneficence: since the needs of others make a claim on me generally, I am duty bound to respond to those needs. But it would also make sense of flipping a coin when confronted with trade-offs: if I’ve recognized the importance of human need as such through empathy, but also recognized that the sort of importance I’ve appreciated cannot be added, I can best do justice to the equal claims of human need by treating the perspectives equally – by flipping a coin.
But perhaps we may also act on a “purely personal” preference. Perhaps the fact that I know and like David moves me not idiosyncratically, but because the significance of human need as such is made visceral by my connection to David, in a way it is not in the case of strangers. I’m moved, as it were, by the significance of the general in the particular. I’m inclined to read Taurek in this way. But, for the purposes of grounding the duty of beneficence in love, it doesn’t really matter. The idea is that in empathy, we recognize what matters by taking up a particular perspective, but not as important because of any particularities of that perspective. That is what is necessary to ground a duty to meet human need. Whether trade-offs can best be negotiated by our natural inclinations to take up particular perspectives or by a coin-flip is a further issue.Footnote 8
I think we have, now, a reading of Taurek that shows him to be fleshing out the idea of a duty of beneficence as grounded in love. This has involved reading Taurek in a somewhat strange way – backwards as I called it. I’ve placed a lot of emphasis on Taurek’s admittedly terse and gnomic discussion of empathy, and less on the details of his hypothetical cases. I want to justify this strategy now by showing its fruitfulness. First, I want to note the upshots for my reading for Taurek’s arguments in a bit more detail. In the next section, I will sort out the plausible and less plausible in Taurek and show how this sorting points the way forward for granting love a convincing place in ethics.
3.2. Taurek’s hypothetical cases revisited, in detail, in light of empathy
Here’s the structure of Taurek’s first argument (295–99), which I’ve put – as he does not – explicitly in terms of a duty of beneficence:
-
(1) If the duty of beneficence requires us to save the five rather than the one when all are strangers but permits us to save David whom we know and like, then this is either because a personal preference can override a requirement of beneficence or because a personal preference can generate a requirement of beneficence. (premise)
-
(2) A personal preference cannot override a requirement of beneficence. (premise)
-
(3) A personal preference cannot generate a requirement of beneficence. (premise)
-
(4) The duty of beneficence permits us to save David whom we know and like. (premise)
-
(5) So, the duty of beneficence does not require us to save the five rather than the one when all are strangers. (1, 2, 3, 4, modus tollens)Footnote 9
The argument is just a complicated modus tollens. Premise (4) is actually the starting point: Taurek begins the stretch of the paper in which this argument is developed by observing that he finds this premise plausible. He suspects that some – though not all – of his readers will too. What he is interested in at this point is assessing whether (4) can be held by those readers who are also inclined to think that, all else equal, we should save the greater number. This is the genesis of (1): the antecedent of the conditional is (4) combined with the idea about numbers that Taurek takes to be widely held. The consequent is, I take it, supposed to exhaust the relevant options. The only thing different between a case in which all six are strangers and the case under consideration is that we happen to know and like David – that is, as Taurek puts it (misleadingly, I’ve claimed), a “personal preference.” And the cases are supposed to be morally different in that the duty of beneficence requires something in the first that it does not in the second – that the greater number be saved. Taurek takes there to be two relevant ways in which this could go – either because personal preferences themselves can override moral requirements, or because they can generate opposing moral requirements of greater significance. Premises (2) and (3) deny these two possibilities. For (2), Taurek thinks that mere personal preference is simply too feeble to override a moral requirement. For (3), Taurek makes the same point regarding the generation of an opposing requirement, with the added observation that we are not required to save David at all – we are permitted to do so.
The controversial premise here is usually taken to be (4). In light of the reading I’ve given of Taurek we can see more clearly why he thinks that it is true, in a way that also involves (2) and (3) – although in a way that draws in material that the case is ostensibly supposed to motivate. If my knowing and liking of David amounts to a kind of empathy through which I appreciate what is significant about human suffering, acting on a recognition of this value can be an appropriate way of discharging my duty of beneficence, which is grounded in the value I appreciate in empathizing with David. The mere fact that I happen to appreciate this value in David cannot override a duty of beneficence to others, however, as they too are loci of this value. On the other hand, the mere fact that I happen to appreciate this value in David cannot ground a duty of beneficence to David in particular for the same reason: if I’m properly appreciating David’s value in empathizing with him, I should recognize that the strangers are loci of this value too. Thus, (2) and (3).
Here’s a regimented form of Taurek’s second argument (299–301):
-
(6) If it is permissible for David to save himself rather than five strangers if he possesses some medicine that would do so, then I can discharge my duty of beneficence by saving David if I possess the medicine.
-
(7) It is permissible for David to save himself rather than the five if he possesses the medicine.
-
(8) So, I can discharge my duty of beneficence by saving David if I possess the medicine.
Premise (6) expresses a kind of self-other symmetry that has puzzled many philosophers, especially because it is not locally clear in the paper why Taurek thinks it is true. But on the reading of Taurek I’ve suggested the puzzlement should disappear. In empathizing with David, I appreciate the importance of his suffering as an instance of human suffering. If he is allowed to save himself because he cares about his own suffering, then his suffering looks to be an instance of human suffering that is important enough to justify his saving. And if that is true, then I can empathize with him, appreciate his suffering as important enough to justify his saving, and discharge my duty of beneficence in acting on that motive.
Now, I want to step back and draw some lessons about the point of a duty of beneficence.
4. Taurek read backwards: worries and lessons
4.1. Overview
You might react to my reading so far by concluding that Taurek should have written a clearer paper. But my point in devoting effort to reading Taurek backwards is that I think it helps us to address my paper’s titular question: what is the point of a duty of beneficence? Or, more specifically: what is at stake in thinking of beneficence as a duty of love – as a duty grounded in a value that is best articulated by appeal to what we appreciate when we empathize with someone?
In the second section of this paper, I claimed that Velleman and Setiya gave us resources to think about duties in terms of love. And I’ve offered a reading of Taurek on which he does this for the duty of beneficence. Is his proposal successful? Is it even attractive?
I think that it is attractive. However, the obscurity of Taurek’s presentation has prevented readers from properly assessing its attractiveness, and led to the raising of worries for which Taurek already has implicit answers and the missing of difficulties (for Taurek, and for the idea of a duty of love in general) which need addressing.
Here is what I have in mind: on the reading of Taurek I have given the duty of beneficence is grounded in a value recognized through empathy or love. That value is non-aggregative – Taurek argues as much, and I think he is right about this. (Whatever I appreciate in empathy, it is something that shows up from a perspective, not from the point of view of the universe – just as is the case for stronger bonds of love between friends and romantic partners.) However, Taurek supplements this picture with two other ideas. First, he argues that there is no intelligible notion of human value that is aggregative. Whatever is valuable about us can only be appreciated by empathetically engaging another’s perspective. This is stronger than what Taurek really needs to give an account of beneficence as a duty of love. Consider: it might very well be that a duty of beneficence is best understood as grounded in love, that the value appreciated in love is non-aggregative, but that nevertheless there are aspects or kinds of human value that can be aggregated and which make different sorts of moral claims on us. Second: Taurek assumes that a duty of beneficence has very wide scope – that the duty relevant in mundane contexts of everyday aid is the same one that is relevant in disaster scenarios. Or, in other words, his argument implicitly treats saving lives as on a spectrum with, for example, giving a lost person directions – all under the amorphous heading of “helping actions.”
Both of these ideas are, I think, problematic, but in a way that shows what has gone wrong with Taurek’s critics and how we might proceed in future work aiming to grant love a place in ethics. In the next two subsections, I will conclude by taking these points in turn.
4.2. Critics’ Problems
I think that critics have been motivated primarily to reject the idea that the numbers of human lives saved is intrinsically insignificant. When they respond to Taurek’s arguments, however, they object in ways that have ready replies if my reading of Taurek is correct. I will take Derek Parfit’s “Innumerate Ethics” – an early and classic critique of Taurek – as representative. Parfit systematically fails to register the role that appreciation of human value through empathy plays in Taurek’s argument – no doubt, as should be clear, in part due to the baffling structure of Taurek’s paper. Confronted with Taurek’s first argument, Parfit takes premise (4) to be based on brute intuition, and counters that it would surely be more intuitive to reject the conclusion (1978: 289–92). But, as I argued above, Taurek is best read as having a rationale for (4) that rests not on brute intuition, but on an account of the value recognized in empathizing with another person.
Confronted with Taurek’s second argument, Parfit criticizes Taurek for failing to recognize the possibility of agent-centered permissions – permissions to do less than the best in your own case (287–89). But Taurek is best understood not as failing to recognize this theoretical device, but rather as articulating part of a picture of morality that refuses to rely on such a notion. Taurek wants not a theoretical black box that captures intuitions, but rather an account of the value that grounds our duties. And, as I argued above, if human value in general is appreciated through love, it is natural to think of self-love and other-love as symmetrical.
Finally, when Parfit reaches Taurek’s discussion of empathy, his arguments are directed solely at Taurek’s stronger thesis – not at the idea that there is a non-aggregative value that can be appreciated through love, but at the idea that one cannot appreciate human value from the point of view of the universe (292–301). His discussion of adding up the badness of headaches is thoughtful and puts pressure on Taurek’s argument for his stronger thesis.Footnote 10 But this fails to engage with Taurek’s discussion of empathy or consider its independent upshots.
There is a lot of literature on Taurek after Parfit, but it seems to me that it reproduces, in different forms, Parfit’s failure to see what Taurek is really up to.
4.3. Concluding: real problems
But the failure of the critics is instructive. As I noted, most responses to Taurek are ultimately driven by incredulity at the thought that numbers do not count. Even if their particular arguments fail to engage Taurek, and even if Taurek does attempt to capture our intuitions about numbers by appeal to fair social policies, I think that the critics’ incredulity gets something correct about the radical nature of Taurek’s ultimate position.Footnote 11 If I’m correct, Taurek offers an account of a non-aggregative human value grasped through empathy that justifies a general requirement to help – a duty of beneficence. But, as I noted above, he supplements this with two additional ideas: that there is no intelligible notion of human value that is aggregative and that the kind of “help” covered by a duty of beneficence is to be understood in the most general possible terms – as covering disasters as much as supermarket interactions. These two supplementary ideas entail that numbers don’t count in a really radical way: they couldn’t possibly count in themselves for decisions regarding any kind of helping action. When confronted with this conclusion, we might, like Parfit, simply be inclined to a modus tollens, rejecting the account of human value that leads us to such conclusions.
But this would throw the baby out with the bath water. There is something compelling in Taurek that requires neither of these two additional ideas. I do seem to recognize something of general human importance in empathizing with others. And, if what I recognize is of general human importance, it can stand as a justification of a general requirement to help. But, if what I’ve recognized is not something that can be aggregated because it requires a perspective, then the general requirement is not one that directs me to aggregate, but rather to act as I ideally would from empathy towards those around me. This, it strikes me, is consonant with basic moral experience. I try my best to help others when I can. But my concern is manifestly not with aggregation. I’m not sensitive in my daily life to opportunities to bestow the largest number of aggregate benefits as I can as I walk to the library on campus and do my shopping at Target. Rather – even though I am not deeply virtuous – I’m prone to empathize with neighbors who have lost their keys or students who need a paper extension and this motivates me to help them. Even if I can’t empathize because I’m having a bad day, I think that I should act as if I did empathize. I know that the needs of these people matter in a way that would show up if I could empathize, even if I’m not emotionally capable in the moment.
Recognizing this does not require that there is no intelligible notion of aggregative human value. All it requires recognizing is that there are aspects of human value that are non-aggregative, but which are not purely part of our private lives – the realm of romantic love and friendship – but rather at the root of a basic part of commonsense morality and moral experience. Taurek’s supplementary ideas simply push this element of common moral experience to the extreme.
In this regard, Taurek ironically appears to share deep assumptions with his consequentialist opponent. The consequentialist often starts not with the fabric of ordinary moral life, but with extreme cases like Taurek’s. These cases are taken to motivate the thought that we have a positive duty to do the best we can – to bring about the best consequences. This is taken to be a general admonition, and the fact that it characterizes so little of ordinary moral life is explained by appeal to things like permissions to do less than the best or to accounts of how ordinary behaviors and motivations actually do maximize the best consequences in the end. What is morally important about humans becomes crystal clear in such extreme cases, and the motives of ordinary human interaction are morally justified – if they are – in these terms. We start with disasters and read everything back into ordinary life.
Taurek shares the idea that disaster cases are on a continuum with ordinary moral life, but he moves in the opposite direction. The common moral attitude toward others – one of empathy – is extended to disaster cases. The duty that the value grasped in empathy underwrites is taken to cover everything. And so, moving from ordinary life, we deny what seemed so clear to the consequentialist in disaster cases.
Perhaps these basic assumptions are correct. Perhaps we have no good reason to refuse to view things on a continuum. Perhaps Taurek is right and there is no intelligible notion of human value that is aggregative. Or perhaps the consequentialist is correct and it is just obvious that five headaches are intrinsically worse than four. I am struck, however, by the large variety of kinds of helping actions and as well by the different ways that humans seem important when everything hangs in the balance. It may be that we can give a unified account of why I should grant extensions and help my neighbor search for their lost keys and give a desperate stranger a ride to the hospital, and so on, in terms of a duty grounded in a non-aggregative value grasped through love. But it isn’t obvious that the same value is at stake – or that the notion of helping hasn’t been stretched beyond recognition – when figuring out how to prevent famine, or when I am worried about the threat of nuclear war. And when I start to contemplate these cases, I do start to think – pace Taurek – of humans, perhaps, as a special kind of object, insofar as I am thinking about what is good about there being humans at all.
What is the point of a duty of beneficence? As Setiya and Velleman both note, the idea that love is distinctively partial is what has made it seem to be opposed to morality. Their work calls this into question. And it provides the materials for understanding Taurek as I’ve read him: as articulating love as the grasping of a value that can ground a duty. It is an account that I’ve claimed is attractive, insofar as it plausibly articulates the role that the injunction to help plays in ordinary moral life. But this should push us to ask the titular question in a different way: what role does this duty play in our overall, systematic account of morality? How should we understand the paradigm cases in which it applies to relate to other, more exotic cases? And here something like the worries about partiality which were the initial source of difficulties appears in a different form. The value appreciated in love can be understood as a moral value. But it is not a value that can be aggregated. On these points Velleman, Setiya, and Taurek are in agreement, and I think they are right. Love is not partial, but it refuses to aggregate – cannot aggregate. But is the refusal to aggregate hostile to morality? Or, less strongly, how much of our moral system can get by without aggregating? The point of a duty of beneficence in a moral theory is to articulate the way in which the non-aggregative value grasped in love makes general demands on us. If that is right, love is not essentially hostile to morality. But this shows us the task that needs to be undertaken in order to bring the relationship between love and morality fully into view. The duty of beneficence certainly has a point, but that point is not yet fully clear. The question is not: can we root ethics in love? Rather, it is: how much of ethics can be rooted in love, given that love refuses to aggregate? This is the real question that a full account of the role of love in ethics needs to answer.
Acknowledgements
I owe debts of gratitude to Barbara Herman, Pamela Hieronymi, A.J. Julius, Gavin Lawrence, Sandy Diehl, Christopher Woodard, and two anonymous referees.
Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Open access
