Ethics and the Vision of Value Chapter 7: Goods and Persons Inhumanity, it is clear, is a trait specifically restricted to the human race. (Robert Spaemann, Persons: the difference between ‘someone’ and ‘something’, p.8) 7.1. Which are the goods? Chapter 6 has developed a rationally coherent model of response to a plurality of irreducibly different values which does not depend (as consequentialist models do) on the possibility of weighting these different values against each other—thus denying, in the end, that they form a genuine plurality of values or goods. Chapter 6 has not determined which these goods are, nor whether every one of the plurality of goods is of equal importance. These are the next questions that we need to consider. I do already have some commitments in place about which goods there are in the plurality. It became clear in chapter 6 that on the present view, there are at least two sorts of goods: loyalty goods and interest goods. Saying that a good is an interest good means that it is some type of thing such that interest in it, enjoyment of it, is made rational by its goodness. (Pleasure was my primary example of an interest good.) Saying that something is a loyalty good means that it is some type of thing such that not only is interest in it made rational by its goodness; that goodness also gives reason for loyalty to the good in question, indeed makes disloyalty to it irrational. (I gave friendship, patriotism, political ideals, and truth as instances of loyalty goods.) In short, interest goods only rationalise response to them, whereas loyalty goods demand response. The last paragraph already gives us some examples of goods (pleasure, friendship, patriotism, political ideals, truth). It also gives us at least a partial answer to the question about equal importance: loyalty goods will be more important than interest goods, in the sense that they make stronger demands on us. How might we advance beyond these modest and piecemeal results, to a more systematic account of which the goods are and how they relate to each other? Many philosophers have not, apparently, seen much need for system at this point. They have simply suggested lists of the goods and importance-relations holding between them. And certainly the sense that system is unnecessary here is encouraged by the striking similarities between their lists of goods: “Accomplishment, the components of human existence [i.e. life and health, liberty, autonomy, “what makes life human”], understanding, enjoyment, deep personal relations…” (Griffin 1986: 67) “Life, knowledge, play, aesthetic experience, sociability, practical reasonableness, religion.” (Finnis 1980: 80-89) 1 “Happiness, knowledge, purposeful activity, autonomy, solidarity, respect, and beauty…” (Railton 1984: 109-110). “Personal affections and aesthetic enjoyments include all the greatest, and by far the greatest, goods we can imagine…” (Moore 1903, VI, para.113) “…the best life for human beings is one of significant engagement in activities through which we come into appreciative rapport with agent-neutral values, such as aesthetic beauty, knowledge, and understanding, and the worth of living beings.” (Darwall 2002: 75) “[L]ife; bodily health; bodily integrity; use of sense, imagination, and thought; healthy emotional development; practical reason; ability to show concern toward others and to be respected by others; ability to have concern for other species; play; and political and material control over one’s environment…” (Nussbaum 2001: 416-418) “…life, consciousness, and activity; health and strength; pleasures and satisfactions of all or certain kinds; happiness, beatitude, contentment, etc.; truth; knowledge and true opinions of various kinds, understanding, wisdom; beauty, harmony, proportion in objects contemplated; aesthetic experience; morally good dispositions or virtues; mutual affection, love, friendship, cooperation; just distribution of goods and evils; harmony and proportion in one's own life; power and experiences of achievement; self-expression; freedom; peace, security; adventure and novelty; and good reputation, honour, esteem, etc.” (Frankena 1973: 87-88) “Friendship; aesthetic value; pleasure and the avoidance of pain; the contemplation of God; life; physical and mental health and harmony; reason, rationality, and reasonableness; truth, and the knowledge of the truth; the natural world; people; fairness; and achievements.” (Chappell 1998a: 43) 1 Of course, anybody can play this list-generating game. And within the limits of credibility, they can put what they like on their list. It is just this that makes the overlaps between these lists (even between the rest and Aristotle’s list in Footnote 1, which I quote With these lists contrast Rhetoric 1300b18-20, a passage curiously little quoted by Aristotle’s modern liberal admirers: “Let happiness, then, be doing well (eupraxia) combined with virtue; or independence of life; or the life that, as well as being secure, is pleasantest; or an abundance of possessions and of slaves (ktêmatôn kai sômatôn), with the power to guard them and use them…” Some might object to taking this passage as an expression of Aristotle’s own views about the nature of happiness by saying that Aristotle here is only collecting the common opinions about happiness that an orator needs to keep in mind. No doubt that is what Aristotle is doing here. It doesn’t follow that he does not accept the opinions he collects; on the contrary, Aristotle is famous for arguing that the fact that something is a common opinion is itself an excellent if defeasible reason for believing it (NE 1145b3). So, apparently, Aristotle’s considered list of goods would not include having plenty of money (see NE 1097a26-28), but it would include having plenty of slaves (Politics 1254a17-1255a3). 1 2 partly because of its very perversity) so much more striking than the differences between them. Given their disagreements in so many other areas elsewhere, these philosophers’ convergence about the goods is remarkable. Still, in the interests of having beliefs that are justified as well as true, it would be nice to get further than listing lists of goods. In 7.2 I shall discuss four possible tests for what counts as a good: the self-sufficiency test, the isolation test, the eudaimonistic test, and the reasons-based account of the goods. 7.2. Four tests for the goods The first test, the self-sufficiency test, is suggested by Aristotle when he writes that “We take the self-sufficient to be that which on its own makes life choiceworthy and lacking in nothing; and we think that happiness is like this” (NE 1097b16-19). According to the self-sufficiency test, we will have a complete specification of the goods once we have described a possible human life which is such that it could not be made any better by the addition of any further goods. (For this familiar interpretation of NE 1097b16-19, see e.g. Irwin 1989: 134.) Taken strictly, this test does not work. One reason is the difficulty of describing or even imagining a life that is so good that it is unimprovable. (Cp. 5.5’s problem about identifying “the best option”—and multiply the difficulties a couple of thousand times for “the best life”.) Another is the fact that, since the contributions of goods to lives is usually holistic, it can’t be guaranteed that the addition of just any good to just any life will always result in an improvement. Loosely understood, however, we can make something of the basic idea of the self-sufficiency test. Very simply, this is that the goods are things that are nonaccidentally apt to make life better: apt, that is, to make life better in the direct way that more and deeper friendships or understanding characteristically make it better, and not in the indirect way that, say, a house-move from Bognor Regis to Littlehampton might make it better. At the very least, this gives us a way of checking candidates for the list. The main objection to this understanding of the self-sufficiency test, as coming down to the question “What factors are apt to make life better?”, seems simply to be that we could not just as well put this question as “What are the goods?”. A second test for what counts as a good is the isolation test suggested by G.E.Moore’s “method of absolute isolation” (Moore 1903). According to this test, anything will be a good (a good in itself, intrinsically valuable) if and only if, when we consider it completely on its own, and ask whether it is better that it should exist than not exist, we decide that it is better that it should exist. Similar comments apply to this test and to the self-sufficiency test. Taken strictly, it is not clear that it even makes sense to posit “absolute isolation” as a condition, even an idealised or imaginary condition, under which we might consider anything. Nor is it 3 plausible to suppose that there is bound to be anything specially revealing about this condition, even if we can conceive it clearly. Loosely understood, however, we can see what Moore is getting at: his idea is that nothing will be intrinsically valuable unless its goodness is essential to it, not just accidental or circumstantial. This test for a good, then, can be loosely understood as equivalent to the question “What things are valuable in themselves?”. As before, the main thing wrong with that question is only that it is not so very far from the question that we started with, “Which are the goods?”. A third test is a little more substantial. This is the eudaimonistic approach to the goods, according to which the goods are defined as such by their role as ingredients of flourishing: “the human good turns out to be the activation of the soul’s capacities according to virtue” (Nicomachean Ethics 1098a17). The goods, on this sort of view, are “aspects of human flourishing”, or “fulfilments of human capacities”: The basic goods are… aspects of the fulfilment of persons. (Grisez, Finnis, and Boyle 1987: 114) Human good is sui generis. Nevertheless… there is a ‘natural-history story’ about how humans achieve this good as there is about how plants and animals achieve theirs. There are truths such as ‘Humans make clothes and build houses’ that are to be compared with ‘Birds have feathers and build nests’; but also propositions such as ‘Humans establish rules of conduct and recognise rights’. To determine what is goodness and what defect of character, disposition, and choice, we must consider what human good is and how humans live: in other words, what kind of a living thing a human being is. (Foot 2001: 51) There has to be something right about the eudaimonist’s insistence that the place of ethics is in human life (or at least, that it is for humans), and that the place of human life is in “nature” (4.3). For all that, the view faces problems too. One problem is that, although the eudaimonistic rationale fits many goods extremely well, it apparently doesn’t fit all the things that we intuitively want to count as goods. Plausibly, the well-being of the Amazonian rain-forest, or of a school of dolphins, is a good, and even an aspect of flourishing; but neither is an aspect of human flourishing. (Nor is either only a good thing because of its possible contribution to human flourishing. It is not incoherent to think it good that a school of dolphins should flourish even though no one takes pleasure in its flourishing.) Equally plausibly, the truth is a good; but the truth does not seem to be an aspect of any sort of flourishing. (Recall the Trinity Mathematicians’ Toast: “Here’s to our new theorem; may it never be any use!”) Indeed the truth would, apparently, still be a good even if it was directly opposed to human (or any other sort of) flourishing: as Nietzsche remarks in Human, All Too Human (para.517), “There is no pre-established harmony between the furtherance of truth and the well-being of mankind”. The reason why we get these apparent counter-examples to the eudaimonistic thesis that the goods are to be defined as the ingredients of human flourishing takes us to 4 a deeper question about the thesis itself: its anthropocentrism. Some, such as the new natural-law theorists Grisez, Finnis, and Boyle, take human nature as “the standard for morality” (Grisez 1983: 183), or say that “the basic goods are2 basic reasons for acting because they are aspects of the fulfilment of persons” (Grisez, Finnis, and Boyle 1987: 114). But the fact that it is easy to find examples of goods that are not, on the face of it, contributions to the fulfilment of persons in any way except accidentally should alert us to natural doubts about this sort of anthropocentrism. Even Aristotle, who in recent times3 has often been read as one of the chief advocates of the anthropocentric account of the good, can see its cosmic parochialism: If political knowledge or practical wisdom is taken to be the highest sort, that is absurd: after all, humanity is not the best thing in the universe. (NE 1141a21-23) The familiar question for the anthropocentric account of what it is to be a good is: “Why is human (or persons’) flourishing, particularly, such a good thing? Why not a moral focus on cockroach flourishing, or on Martian eudaimonia?” Eudaimonism seems to get the order of explanation back to front. If knowledge and friendship (for instance) are goods, as seems plausible, they are not goods because they are aspects of human flourishing. Rather they are aspects of human flourishing because they are goods, and because human flourishing (unlike cockroach flourishing) is a sufficiently sophisticated form of flourishing to include them. To turn the last quotation from Grisez, Finnis and Boyle on its head: aspects of the fulfilment of persons provide basic reasons for acting because they are grounded in the basic goods. Reversing the order of explanation like this has its uses. For example, it enables us to make at least some sorts of comparison between the goodness (if any) of cockroach flourishing (et cetera) and human flourishing. In each pairwise comparison, the better form of flourishing will be the form that includes participation in more and “higher” goods. It is better to take this route—and face up to the resultant problem about what exactly “higher” means in this context—than to take human flourishing to be such an ultimate terminus of argument that you are left with no defence against the dialectical opponent who takes cockroach flourishing as equally ultimate, and challenges you to dislodge him from this position. Nor, if we put things this way round, do we necessarily lose our grip on one of the chief benefits of the eudaimonistic account—namely that a flourishing human life does seem to be the place where most of the goods we want for our list are likely to be found. Thus we can keep the reference to human flourishing as, at least, a diagnostic aid to locating (many of) the goods, even if we reject the eudaimonistic account as definitive of the goods. More importantly, perhaps, we can also recognise that the nature of some goods—pleasure, for instance—is indeed best understood by considering their place within human eudaimonia, without enforcing the eudaimonistic pattern for all goods. So, for instance, Stephen Darwall is right to say that “a person’s good is constituted… by Sic. Presumably “are sources of” would be more accurate. I am thinking, e.g., of Nussbaum 1978’s reading. An older and far more convincing reading makes Aristotle a non-anthropocentric ethicist: for this see Aquinas. 2 3 5 what one should want insofar as one cares about her” (Darwall 2002: 4); the only thing to add is that not every good is part of any “person’s good”. A further thought about the eudaimonistic account of the goods, to which I shall turn in the next section, is that it is in a way not anthropocentric (or prosopocentric?) enough. In developing the possibility that goods are aspects of the fulfilment of persons, it overlooks the possibility that persons themselves might be goods—an idea that we can deploy to help us answer the question raised above as to why “human fulfilment” might be thought such a good thing, and so much better a thing than, say, cockroach flourishing. But more about that in its place. The fourth, and in my view the most successful, of these ways of deriving or justifying a list of the goods is the reason-based account of the goods. The basic idea of this account is found—alongside much else, including intimations of the eudaimonistic and self-sufficiency tests—in Aristotle’s ethics: “The ends of action (telê) seem to be many; but we choose some ends for the sake of others—wealth, for example, and flutes, and instruments in general. So it is clear that not all ends (telê) are final ends (teleia)…” (NE 1097a26-28) The reason-based account identifies goods (basic goods as they are sometimes called, e.g. by Finnis and Grisez) by identifying what Aristotle calls teleia telê—final ends. Final ends, or basic goods, are those objectives of action which can rationally be chosen, not for the sake of anything else, but for their own sake. Thus, to re-use one of Aristotle’s examples from the Nicomachean Ethics (an example which contradicts the Rhetoric’s account of the goods as given above—see Footnote 1), wealth is not a basic good, because it could not rationally be chosen as an end in itself. Anyone who did choose or pursue wealth, not for the sake of what wealth could get her, but just for its own sake as wealth, would be open to rational criticism on that score. (“But what’s the point of acquiring all this wealth, if you don’t want to do anything with it? Haven’t you turned into a bit of a miser?”) Likewise catching the bus is not a basic good. There is nothing rational, or even intelligible, about aiming simply and solely at the catching of a bus: there needs to be some further end in the background that is achieved, or in some way advanced, by catching the bus before we can so much as understand the agent who makes bus-catching his aim. Even to say that someone has bus-catching as his hobby is to begin to draw his bus-catching within the reach of intelligible final ends such as those of play and recreation. To add that we are likely to think bus-catching an eccentric hobby is precisely to concede that bus-catching, even understood as a way of pursuing those kinds of ends, remains a peculiar way of pursuing these ends. By contrast—according to the reason-based strategy—the choice or pursuit of typical basic goods, just for their own sakes, is perfectly intelligible. Bus-catching to go to the museum to see some art, or to visit a friend, or to see the doctor about my illness: these activities are intelligible in a way that bus-catching for its own sake is not. The explanation is that art, friendship, and health are all, unlike bus-catching, goods that give us reasons that are independently intelligible: they are goods that it can always make 6 sense—other things being equal—to pursue in their own right, and without any reference to any other or more final end. According to the reason-based strategy, this is what makes them basic or intrinsic goods. The same will hold for all other goods that we can rationally pursue in their own right. Notice that the reason-based strategy is not the same as the strategy that John Stuart Mill is sometimes thought to adopt—simply to take statistical-behavioural evidence about what people do in fact pursue or desire as grounds for conclusions about what people ought to pursue or desire. (The locus classicus for this accusation is G.E.Moore, Principia Ethica Chapter 1.) Pace some interpreters, I doubt his approach was really as statistical-behavioural as some more recent theories in the revealedpreference style have been. But whatever Mill may have been up to, the reason-based strategy is explicitly a normative one, not a behavioural one. For this strategy, what counts is not what people do choose or pursue as an end in itself, but what they can rationally choose or pursue as an end in itself. The point is not about motivating reasons, but about justifying reasons (1.7). The reason-based strategy defines a basic good as any good that can rationally be pursued for its own sake. It is therefore no objection to the reason-based strategy that such goods are not always pursued for their own sakes, even rationally. Aristotle thinks this is an objection—“We say that something is ‘more final’ if it is never pursued for the sake of something else” (NE 1097a31-33); but Aristotle is surely wrong. For one thing, Aristotle should surely have put “is never rationally pursued” rather than just “is never pursued”: obviously someone irrational might pursue the more final for the sake of the less final. Even leaving that aside, it is easy to think up cases where a perfectly rational agent finds himself, say, reading what he can see is a great novel solely for the sake of acquiring a paper qualification, or (sincerely) admiring a fine view simply to overcome a social awkwardness. Again, someone might be a utilitarian social psychologist, and combine a belief that happiness is the only final good with the promotion of someone else’s happiness, not as an end in itself, but so that she will be a suitable subject for his next study in social psychology. Such examples show that it is entirely unproblematic to think both that something is a basic good, and also that the good in question can be, and sometimes is, rationally pursued as a means to something beyond itself. The reason-based strategy raises an issue about distinctions that is worth discussing briefly here. (It appears at first sight to be an issue about terminology, but it is more than that.) I consider “instrumental”, “non-basic”, “extrinsic”, and “derivative” as terms for one side of a distinction among goods, and “intrinsic”, “basic”, “noninstrumental”, and “non-derivative” as terms for the other side of the very same distinction. I am thus ignoring what Korsgaard 1983 has argued is the crucial difference between the intrinsic/ extrinsic and final/ instrumental distinctions. I do this deliberately, because I don’t share Korsgaard’s view that it is possible to value anything intrinsically but instrumentally, or extrinsically but finally. To value something intrinsically but instrumentally is not to value the thing at all, but the fact of its instrumentality; to value something extrinsically but finally is not to value the thing either, but its extrinsic properties. There is no harm, then, in treating Korsgaard’s two distinctions as one. More 7 radically, we could give up both distinctions, and refuse to call anything a good that is not valued for its own sake. For there is a clear sense in which instrumental goods are not goods at all, but merely means to goods; and an equally clear sense in which extrinsic goods are not goods either—it is their extrinsic properties that are the goods in question.4 One has to admit that there is a certain messiness and inconclusiveness about the reasons-based strategy. In one way, the strategy is a strongly a posteriori approach: it commits us to the project of looking to see what agents take to be sources of reasons, and deciding whether they are justified in their views about this. This is a rather messily open-ended project. The task of socio-ethological investigation that it sets us has no obvious end-point; for all we can tell, all sorts of surprises might turn up along the way. In another sense the reasons-based approach might also seem objectionably a priori. What we are looking for when we use the reasons-based strategy is not what kinds of things people do take as reasons, but what kinds of things people can rationally take as reasons. It may seem that this reference to rationality simply provides us with an easy way to beg the question, or reduces the question “What things are sources of reasons?” to the question “Which are the goods?”. If so, the reasons-based strategy for locating the goods is, like the first two tests for the goods, little more than a variant of that question. So all four of these ways of justifying a list of the goods, rather than simply helping ourselves to such a list, turn out to be imperfect tools. Either they get us only a little further forward than the direct question “Which are the goods?” (the first two, and the fourth); or they seem philosophically questionable (the third); or they give us messy and disorganised results (all four). These methods are perhaps none the worse for these flaws; maybe the problems they face are just the kind of problems that we would expect in what Aristotle calls “arguments to, not from, first principles” (Nicomachean Ethics 1095a31). In any case, the imperfections of our four methods of locating the goods are unlikely to make much difference to what goes into our list of goods, given the wide consensus that I have noted about what the goods are. On any plausible account, the list of goods will include interest goods such as pleasure, play, amusement, connoisseurship, achievement, and humour; and it will include loyalty goods such as friendship, love, truth, and (worthwhile) political, philosophical, or religious ideals. We can, then, develop a large part of our account of the goods along such familiar lines as these, without worrying too much about exactly how to justify that account. 4 Michael Zimmerman tries to make a different distinction, arguing that non-derivative and intrinsic goodness are not the same (http://plato.stanford.edu/entries/value-intrinsic-extrinsic/): “As ‘intrinsic value’ is traditionally understood, it refers to a particular way of being non-derivatively good; there are other ways in which something might be non-derivatively good. For example [as well as] thinking of health as intrinsically good [we can also think of it as] ‘good for’ the person who is healthy… If John were a villain, you might well… insist that, in light of his villainy, his being healthy is intrinsically bad, even though you recognise that his being healthy is good for him.” This argument does not show that John’s health is nonderivatively good; it shows only that it is non-derivatively good for John. So it does not show that nonderivative goodness can be anything but intrinsic goodness. 8 7.3. Persons as goods “A large part” only, however, and not the whole. The central part of my account of the goods will not be along the most familiar lines. The most important claim that I want to make about the goods, and the heart of my theory of value, is a claim which is surprisingly rare in contemporary ethical theory: the claim that persons are goods. Perhaps because of the way the four tests outlined in 7.3 are usually phrased, this claim does not usually appear among the results of those tests. But in fact, all four tests point us in the direction of treating persons as goods. The self-sufficiency test does this in a very obvious way, as does the eudaimonistic test: it is, after all, extremely plausible to suggest that no person can live well without other persons. (The idea that a person might live well with no more than the illusion of company is refuted in a familiar way by Nozick’s “experience machine” thought experiment (Nozick 1977, Ch.1). The thesis that a good human life can be a solitary one is unsupported in surprising quarters: the usual conception of religious solitude, for example, sees solitude not as a human norm or even ideal, but as an exceptional state for a chosen few, whose solitude is essentially dependent on its background community, and has a certain sort of contribution to that community as part of its intended purpose. (For more about the individual’s relation to the community, see 7.6.) What about the isolation test? In its pure form, this test asks us to consider the thing in itself, and apart from its relations to anything else, and decide whether the world is a better place with it or without it. If this is difficult to do with anything else, it is particularly difficult to do with persons; for one of the most striking features of persons is (as we shall see) precisely their special relation to all the other goods. So let us take a less pure form of the isolation test, and simply point out that the choices and the lives of persons provide the only possible framework within which all the other goods can be realised; personhood has a creative role in its responses to the (other) goods. Compare what I said in 5.5 about the way in which practical choice can partly determine the values of goods. For many choices between goods, I argued, no antecedently determinate fact of the matter can be found as to which option is the better one to take. Instead, what is true is that, if I take option A, then option A can (and will, if all goes well) become the right one to have taken because of the way that my life will develop as a result of having taken it. Likewise, if I take option B, that too can (and with any luck will) become the right one to have taken. My decision, together with a hundred subsequent decisions that flow from it, creates a narrative, a structure of significance, within which my choice of option A comes to be the right one because of its place in that continuing story. No illusions need be involved in such a process—though perspectives certainly are; indeed the process itself can be called the creation of a perspective. It’s not that previous truths about the relations of value are falsified, but rather that such truths are precisified. Choices made within a narrative context of a life can and often do give the values involved particular and definite values for the chooser which they did not have outside that context or perspective. In this sort of way, as I put it in 5.5, “choice can be anterior to value”: more 9 exactly, “choice can be anterior to fully-determinate value even though it is posterior to rough value.” Thus the relationship of persons to the goods that their living instantiates is not just a matter of their passively acting as receptacles for instances of those goods, as a beer-glass holds beer. (More about receptacles in 7.7 below.) It is a creative relationship, in the precise sense that it involves the person’s creating, in the narrative of his life, a particular structure to which those goods stand as ingredients or raw materials. Only in his life do those goods have the precise interrelations that his agency gives them. Those goods are possibilities for him to create with, just as musical themes are possibilities for a composer, and like the musical themes, will have no exact realisations until his choices give them exact realisations. Persons are unique in nature because persons alone can act upon the goods in this way, to organise and co-ordinate them into coherent narrative structures in their lives. The very familiarity of personhood or agency5 so understood should not blind us to its uniqueness—nor to its unique value. There is something more than an argument of the isolation-test form at stake here. There is a whole vision of humanity, and what humanity can be. We may call it humanism, and think it worth defending. Despite its foundational role in almost everything in history that has deserved the name of civilisation, it has never been short of detractors.6 What about the fourth test, the reasons-based strategy for identifying the goods? Does this too point us towards the idea that persons are goods? It might seem not to, since it is hardly natural to suggest that my basic reason for doing anything might be a person; as it stands, this suggestion is barely even grammatical. However, another familiar form of words shows us how to bypass this verbal difficulty. We speak, following Kant, of persons as ends in themselves. By this we mean that persons are not to be seen, or not solely seen, as instrumental to other ends. Instead (it is natural to add) they are to be valued in their own right and for their own sake. The attitude of love, directed towards persons, is a good example—perhaps the paradigm example—of an objectual emotion, not a propositional emotion, i.e. of an emotion with a thing as its object, not a proposition (1.7, 3.3). The phenomenology of loving people—more about it in the next section—is that we love them, not (just) their properties or accidental attributes;7 and the received 5 Strictly speaking I should distinguish agency from personhood. Agency is the practical aspect of personhood: the agent is the person insofar as the person aims to weave eudaimonia out of the moral or practical goods. There is also a theoretical side to personhood, the understanding or cognising side: the cogniser is the person insofar as the person aims to weave understanding out of the epistemic goods. I will not obtrude this distinction into the present discussion, which is complicated enough without it. 6 One recent detractor is Gray 2002, who relies upon two main arguments, or rather rhetorical tactics, against humanism and humanistic morality. The first is the presentation of anecdotes tending to show that historically speaking humans have often behaved badly. The second is an appeal to the theory of evolution, rather loosely conceived, as evidence that humans often can’t stop themselves behaving badly. A third favourite hobby-horse is that humanism is the love-child of Greek philosophy and Judaeo-Christian faith. These darling notions of Professor Gray’s are neither new nor épatant, even to bien-pensants like me. Still less are they reasons to abandon humanism. (Or Greek philosophy; or Judaeo-Christian faith.) 7 It might be suggested that there is an intermediate possibility between loving a person for herself, and loving her for some property of hers: namely loving her for the place she has in my life, and for the place I 10 view about love is that it is the best and most valuable emotion we can feel (or attitude we can take).8 It is, therefore, also part of the received view that the best attitude we can take to others, other things being equal, is to love them for their own sake. The received view, so understood, is getting very close to acknowledging what I want to say: that persons are intrinsic goods. The four tests thus point us towards the conclusion that persons are goods. Given the unique role of persons relative to the rest of the goods, they point us towards the conclusion that persons have a crucial place in the theory of the goods. We can also see now how the claim that individual persons are intrinsic goods can be related to the key claim of chapter 6 that, for any good, we have freedom to pursue it within the limits set by our obligation to respect it. Talk of respecting and pursuing persons seems just as ungrammatical and nonsensical as asking “whether adding a person to a life makes it go better”: but sense can now be made of this sort of talk too. No doubt the locution “pursuing persons” is simply a joke; but we may at least speak of “pursuing the good of persons”, and thereby mean “pursuing the reasons that the existence of some person gives you to promote or help develop that person’s agency”. Similarly we may speak of “respecting persons”, and thereby mean “acting on the reasons that the existence of other persons gives you not to undermine, threaten, or destroy their agency”. No doubt there is a decidedly Kantian ring to these claims. Like Kant, I take as central the unique value9 of persons, and their status as “limiting conditions on agency”— the status, that is, of each agent as a limit on what any other agent can reasonably do. Like Kant’s, my picture is designed to show the independence of each such value from every other such value in a “republic of ends”. My claim is not merely that some abstract property called personhood is an intrinsic value, but that each individual person separately is a free and equal10 intrinsic value. Also like Kant, I see a difference between the minimum requirements of ethics, and its higher aspirations above those minimum have in hers. Solomon 2006: 58: “’Because we fit so well together’ can be an excellent reason for love”, a reason of the sort that Solomon nicely calls “Aristophanic”. It is helpful and revealing to see many cases of love (including the best cases) as based largely on Aristophanic reasons. Seeing love-relationships as typically involving shared projects—whether the project is as narrow as climbing together in the Alps for one fortnight, or as wide as living and raising children together for as long as we both shall live—gets us away from the static view of love that can come from talking only of loving persons in themselves. However, appeal to Aristophanic reasons does not undercut the distinction between loving someone for her properties and loving her for herself; it cross-cuts it. There are, after all, many different contexts in which we can “fit well together”. In a merely carnal relationship, we might literally do that. (This joke is as old as the Symposium.) And then, probably, our relationship would be both Aristophanic and also merely instrumental. 8 Love is naturally understood both as an emotion and as an attitude; and also as an emotion that generates an attitude, and an attitude that generates an emotion. 9 This remark should be read with the caveat about Kant’s view of value noted in 5.4, that for Kant, given his metaphysical scepticism, the point cannot be that persons are intrinsically valuable, but only that we are rationally bound to treat them as intrinsically valuable. 10 “Free” here means that no person has controlling rights over any other person; “equal” here means that each person has exactly the same claim of practical reason on every other person. The quasi-political language is, of course, deliberate. 11 levels: where he talks of “perfect duties” I talk of obligations to respect, where he talks of “imperfect duties” I talk of prerogatives to pursue. 7.4. Three objections to the thesis that persons are goods My thesis that individual persons are intrinsic goods of a special sort should at once be tested as any philosophical thesis must be tested: by considering some objections. One natural objection lies in the counter-thesis that it is only morally good persons who are intrinsic goods; morally bad persons, it might be said, are nothing of the sort. A second possible counter-thesis (call this the receptacle view of the self: 7.7) says that persons’ lives are good if and only if they have good things in them; they are bad if they have bad things in them, but in neither case intrinsically good. These two objections show a common strategy, that of urging reasons why we should take personhood, per se, to be not good, but morally indifferent. The two objections also fall to a single pattern of response: a charge of irrelevance. We should distinguish the claim that a person is an intrinsic good, which is the claim that I am making, from the claim that a life is an intrinsic good, which I reject for exactly the reason given by the present objections—that some lives are morally or prudentially good, and others morally or prudentially bad. My thesis was that persons are intrinsically good for a number of reasons; for one thing, because of their unique creative power over the other intrinsic goods. Remarks about the possible goodness or badness of lives are remarks about how that power can be used or misused, fortunate or unfortunate. They have nothing to do with the assessment of the goodness of the power itself. “But if a power is neutral between good and bad results, then surely that power is itself neutral in value.” This simply doesn’t follow; a capacity can be apt to produce either good or bad results, and still be good in itself. Health and strength can be used to save lives or to murder, the imagination can be used to create great art or dreary pornography. No one doubts for that reason that health, strength, and imagination are good things. As these examples show, it is not controversial that capacities can be both of instrumental value, and also of intrinsic value: good things to have in their own right, quite aside from the further goods that we can achieve by having them. Likewise with the power over the other goods that persons characteristically display. A third objection to the thesis that persons are intrinsic goods arises from the following puzzle. It can seem that it is not even possible to value some particular person X for herself, or for himself, alone. We can get into a puzzle (one I consider further in Chappell 2004) by asking why we value X: it can seem that the only possible alternative answers are “Because X has property F” or “For no further reason”. But if we value X because X has property F, then it seems that what we value is not X, but the property F, or at most the instance of F that happens to be X. Whereas if we say that we value X “for no further reason”, it is easy to hear that as meaning that we value X for no reason at all. Obviously that is the wrong way to hear it. What is meant, or what can be meant, by the 12 claim that we value the person X “for no further reason” is simply that we value the person X in herself. This possibility is easily overlooked, perhaps because of a tendency, more common than commented upon, for philosophers to assume that a justificatory argument about values is bound to terminate in references to properties as basic or final values. But it is easy to show that this is a quite groundless assumption; there is no reason at all why only properties can be basic values. Let us suppose we agree that something x is good, and we ask “In virtue of what is x good?”. We then pick a feature F of x—a property of x, a relation in which x stands, or whatever—as the answer to this question. Now we may ask the same question again: “In virtue of what is feature F good?”. To answer it, we have two alternatives. The first is to pick a feature of this feature—call it G—and say “F is good in virtue of G”. The second is to invite other parties to the debate to agree that F is not good in virtue of any further feature; instead, F is good in itself. I’ll call taking this second alternative making the Intrinsic Appeal. Obviously, unless we make the Intrinsic Appeal at some stage, we will find ourselves in an infinite regress, or else working in a circle: either x is good in virtue of F, and F is good in virtue of G, and G in virtue of… (ad infinitum), or else x is good in virtue of F, and F is good in virtue of G, and G is good in virtue of H, and H— closing the circle—in virtue of F. To prevent regress or circularity, we have to make the Intrinsic Appeal at some point. Nothing in the structure of the argument limits where we may make the Intrinsic Appeal, nor what sorts of things we may appeal to in making it.11 There is no reason, then, to assume that anything that is intrinsically good must be a property, and cannot be any other kind of thing—unless one has an argument in metaphysics, as some philosophers (among them Hume, perhaps12) have thought they had, for thinking that there is no other kind of thing except properties (or perceived property-instances). Nor therefore is there any serious puzzle about how intrinsic goods could be items in some other ontological category: facts, or tropes, or capacities, or things: persons, for instance.13 Hence it is no less reasonable and intelligible to claim that persons are intrinsically good than to claim that this or that property of persons is 11 Zimmerman 2002: “Suppose that someone were to ask you whether it is good to help others in time of need... you would answer, ‘Yes, of course.’ If this person were to go on to ask you why acting in this way is good, you might say that it is good to help others in time of need simply because it is good that their needs be satisfied. If you were then asked why it is good that people's needs be satisfied, you might… say, ‘It just is.’ But then… your interlocutor could ask once again, ‘What's good about that?’... you would be forced to recognise that… there must come a point at which you reach something whose goodness is not derivative in this way, something that ‘just is’ good in its own right, something whose goodness is the source of, and thus explains, the goodness to be found in all the other things that precede it on the list. It is at this point that you will have arrived at intrinsic goodness.” 12 Hume, Treatise 1.4.6: “For my part, when I enter most intimately into what I call myself, I always stumble upon some particular perception or other… I never can catch myself at any time without a perception, and never can observe any thing but the perception.” 13 Various candidates have been advanced as possible bearers of intrinsic value: properties (Butchvarov 1989: 14-15), tropes (Rabinowicz and Rønnow-Rasmussen 2003), facts (Ross 1930: 112-13; cf. Lemos 1994, ch. 2); states of affairs (Chisholm 1968-69, 1972, 1975), “states” (Zimmerman 2001, ch. 3), ecosystems (Naess et alii 1988), and indeed persons (Gaita 2000). 13 intrinsically good. (Even if it was, the following simple manoeuvre would always be available: we could just define a property of being a person, and ascribe intrinsic value to that property.) Nor, therefore, should we be much impressed by the complaint which one sometimes hears, that anyone advancing the claim that persons are intrinsic goods must be engaging in mere assertion: that the thesis that persons are intrinsic goods, even if true, cannot be argued for. For one thing, I spent the whole of 7.3 arguing for it. For another, if there was this problem about arguing that persons were intrinsic goods, the same problem would arise for anyone who attempted to argue that anything whatever was an intrinsic good. If there can be intrinsic goods at all, then there can be items from some or other ontological category (call them Fs) which are valuable in themselves, and not because of some further feature that they possess. That does not mean that all we can say to explain why Fs are good is merely “Because they’re Fs”. We might, for instance, fill out the remark about by running 7.2’s four tests backwards so to speak, and describe how Fs are naturally apt to add value to lives; how they have value in themselves and apart from other things; what they distinctively contribute to eudaimonia; and how they stand at the termini of rational practical reasoning. But even if there was no more to say about why Fs are good than “Because they’re Fs”, it would make no difference which ontological category Fs belonged in. The problem would be the same whether F was a thing or a capacity or a property or a trope or a fact or a relationship or anything else. What, however, are persons? That is the question I shall answer in 7.5. 7.5. Persons as human animals My argument under this heading is very simple. The argument itself can be very briefly stated, though developing it in response to objections may take longer. First, persons are a kind of things, not of properties or relations or capacities, or items from any other ontological category. Second, the only kind of thing (that we all know of) that persons could be is things of the kind: human animals. So, third, for each and every person that we (at any rate) all know about, to be a person is to be a human animal. It is worthwhile to spell this argument out step by step. The first claim, however, is so obviously true that it is hard to know what to say to spell it out. Any plausible set of ontological categories will include a category at least resembling the category I have simply called things—a category of substances, individual existents, subjects, or something of that sort—and will oppose this category to all of a group of categories at least resembling the Aristotelian categories of properties, relations, capacities, and so on (though it may well also include un-Aristotelian categories such as facts, tropes, or states of affairs): we might call this the thing-quality or noun-adjective opposition Since, platitudinously, “person” is a noun not an adjective, it is not much more than a platitude that, given such an opposition, persons will fall on the things side of it. Persons are things, not qualities. 14 This first point seems obvious and commonsensical. If a further argument is needed, we may think further about what is involved in loving a person for herself.14 As I said in 7.3, to love a person for herself is not just to love her properties or capacities or relations, or facts about her, or states of affairs involving her, etc.: it is to love her, the thing that she is. Love of the person is love of the thing, not merely of the qualities; so the person is the thing, not merely the qualities. Of course, the distinction between a thing and its qualities (or properties) is apt to seem puzzling on a Lockean view of substance (Locke, Essay 2.23.2): If any one should be asked, what is the subject wherein Colour or Weight inheres, he would have nothing to say, but the solid extended parts; and if he were demanded, what is it, that that Solidity and Extension inhere in, he would not be in much better case, than the Indian… who, saying that the World was supported by a great Elephant, was asked, what the Elephant rested on; to which his answer was, a great Tortoise; but being pressed again to know what gave support to the broad-back’d Tortoise, replied something, he knew not what…. The Idea then that we have, to which we give the general name Substance, being nothing but the supposed, but unknown support of those Qualities we find existing, which we imagine cannot subsist, sine re substante, without something to support them, we call that support Substantia. —But that is not a reason to worry about whether we can love persons in themselves. Rather, as Elizabeth Anscombe points out, it is a reason to give up on the Lockean view of substance (Geach and Anscombe 1960: 10-11): Because Aristotle distinguishes between substance and quality, those who take a predicate like [“human”] to signify a complex of properties readily suppose him to be distinguishing between the being of a thing and the being of any attributes that it has. They then take the thing itself to have no attributes. It would be almost incredible, if it had not happened, to suppose that anyone could think it an argument to say: the ultimate subject of predication must be something without predicates; or that anyone who supposed this was Aristotle’s view could do anything but reject it with contempt. If this is where Locke’s approach to substance gets us, then so much the worse for Locke’s approach. In metaphysics, we should prefer a view of substance (of things as opposed to qualities) that, like Aristotle’s, allows us to be aware of the substance itself, directly, in our immediate experience—and not just the properties of the substance. What I am (representationally15) aware of when I perceive a cat, say, is not just blackness, warmth, and furriness, but the cat that is black, warm and furry. Likewise (or indeed as one application of this), we should prefer a view of the person that makes persons themselves directly available to be loved, and not just their properties. It isn’t only the properties of the particular existing thing—not even only its essential properties—that are 14 15 Some of what follows restates or refines the argument of Chappell 2004. As opposed to “sensorily”; cp. 1.3. 15 available to be known. We can and should hold to the commonsense belief that the particular thing that has the properties can itself be known. And then it can meaningfully be said that there is a difference between loving the person’s properties, and loving the person. My argument’s second claim is that the only kind of thing known to us that persons could be is things of the kind human animals; and its third claim is that, therefore, being a person is being a human animal. What other kind of thing could persons possibly be? Anyone who wants to claim—as many philosophers do—that persons are some other kind of thing needs to give us reasons to believe that this other kind of thing exists as more than a philosophers’ stipulation. For instance, it will not do, common though this proposal is (e.g. MacMahan 2002), to suggest that persons might be psychologies. Psychologies are not a genuine kind of thing; psychologies are not things at all. Psychologies are a kind of properties of things—namely, of persons. In Strawson’s words (1959: 102), “States of consciousness could not be ascribed at all, unless they were ascribed to persons.” A second and even more widespread proposal says that we should stipulate that persons are human (or similar) animals that satisfy certain further tests, or with certain further properties added, or considered under phase sortals. The best known example of this sort of proposal is Singer’s: I propose to use “person”, in the sense of a rational and self-conscious being, to capture those elements of the popular sense of “human being” that are not covered by “member of the species Homo sapiens”. (Singer 1993: 87)16 The question for Singer’s proposal is whether his creative definition of “person” corresponds to any non-stipulative kind of thing. Pretty plainly, the answer is “No”. To talk about “human beings who are rational and self-conscious” is not to talk about a different kind of things from human beings, no more than to talk about “kites which are blue” is to talk about a different kind of things from kites. Rather, it is to talk about a subclass of that kind, the members of which subclass are picked out by their possessing the further properties of rationality and self-consciousness. The same conclusion applies to John Harris’s proposal (1985: 16-17) that “Persons are beings capable of valuing their own lives”; this makes persons, not a genuine kind, but that subclass of the pseudo-kind beings, or better of the genuine kind human animals, which consists of humans that have the further capacity to value themselves.17 The same applies also to Tooley’s proposal (1972: 82) that “An organism possesses a serious right to life only if it possesses the concept of a self as a continuing subject of experiences and other mental states, and believes that it is itself such an entity”; this makes persons, not a genuine kind, but that Rather strikingly, this stipulation is made on a page of Singer’s book (87) that begins with a denunciation of stipulative procedures: “The morality of abortion is a substantive issue, the answer to which cannot depend on a stipulation about how we shall use words.” 17 Harris’s idea that x’s capacity to sense value is the criterion of x’s having value just seems a confusion. (An originally Kant-inspired confusion, but still a confusion.) Geiger counters sense radioactivity; that does not make Geiger counters radioactive. 16 16 subclass of the kind human animals that has the further property of possessing a particular concept and a particular belief. The same goes too for Mary Anne Warren’s proposal (1997: 83-84) that the six key markers of personhood are “(1) sentience... (2) emotionality... (3) reason... (4) the capacity to communicate... (5) self-awareness... (6) moral agency”; this makes persons, not a genuine kind, but that subclass of the real kind human animals that has these further six properties or capacities. The same also goes, finally, for Olson 2007’s suggestion that “person” is a phase sortal (“Being a person may be only a temporary property of you, like being a philosopher”): this makes persons, not a genuine kind, but that subclass of the kind human animals that has, at any given time, whatever further capacities Olson takes to be needed.18 But why does it matter whether persons constitute a genuine kind, or merely a subclass of some genuine kind? Haven’t I just said myself (7.4) that it is no more or less rational, other things being equal, to value items from one ontological category than from any other? Suppose we grant the assumption I am apparently making (more about it in 7.7) that human animals constitute a genuine kind of things: that the ontological classification human animals is one that “cuts nature at the joints”. It doesn’t follow that our moral classification must cut along the same line. For all I have shown, the line of moral classification could cut anywhere. If we can (other things being equal) with equal rationality value properties, or capacities, or things, or items that fall under any other heading in the list of categories, presumably it is no more and no less rational to value a subclass of a kind of things than to value a kind of things. This objection is perfectly correct.19 The fact that ontology makes a given distinction does not entail that ethics must make that distinction too. This conclusion does not follow, for instance, from the thesis of the supervenience of the normative on the descriptive: that thesis says that for every normative distinction there is a descriptive distinction, not that for every descriptive distinction there is a normative distinction. The real problem with Singer’s and the others’ creative definitions of “persons” as a subclass of the kind human animals is not that their proposals about what to count as a person are metaphysically or logically inadmissible. As my own argument in 7.4 implies, these proposals are no less metaphysically or logically admissible than any other proposal we might make about that. –Though certainly it is salutary to note that they are no more metaphysically or logically admissible, either: such proposals are often presented as if they were irresistibly obvious, but in fact their logical standing is at most no better than the standing of any other possible stipulation about what to count as a person. The real problem is not about the metaphysics of these stipulated meanings for “person”. The real problem with these proposals is a moral one. In a nutshell, it is that they constitute an attack upon the community of humanity. I shall explain and expand this remark in the next section. Some examples of the restrictive approach to personhood go even further than these: “If being a person is to be [sic] a responsible agent, a bearer of rights and duties, children are not persons in a strict sense” (Engelhardt 1989: 120). 19 In conceding the objection I am, in effect, recanting Chappell 1998a, ch.4’s “symmetry thesis”. 18 17 7.6. How we make each other into persons I said in 4.2 that “Persons are unique in nature because persons alone can act upon the goods… to organise and co-ordinate them into coherent narrative structures in their lives”. What I did not say there, but will now add, is that no person does this completely; that no person does it purely; and that no person does it alone. No person does it completely, in the sense that every actual exercise of this capacity to shape the goods is always less than it could be. All of us are bearers of unrealised potential for agency. No person does it purely, in the sense that no actual exercise of this capacity is ever free from the possible sceptical imputation that it was not, in fact, an exercise of good-shaping capacity at all, but some manifestation of ego, or brute instinct, or sheer mechanical accident. All of us can be interpreted, by a sufficiently determined sceptic, from a non-intentional stance, as nothing more than currents of primitive deterministic urges that bubble along, with agency as no more than a froth on the surface. And no person does it alone, in the sense that our agency always has, and arises out of, a history in our community. Every, or almost every,20 actual agent came to be such because he is surrounded by other, similarly placed persons, who teach him how to actualise his agency. Normally, of course, these others are his parents, older siblings, and other friends, and ideally they teach him not by accident but by taking the trouble to instruct him directly; but this pattern is not the only one possible, even if it is, so far as we can tell, the one that works best. In the words of Schapiro 1999: 723: …there is a sense in which no one, regardless of age or maturity, is able to achieve autonomy on Kant’s view. This is because the notion of autonomy in Kant is an ideal concept which outstrips all possible realisations in experience. Strictly speaking, every instance of human willing is necessarily an imperfect realisation of transcendental freedom, and every virtuous character necessarily falls short of perfect virtue. And yet the applicability of the moral law depends on our mapping these ideal concepts onto ourselves and one another for the purposes of guiding action. So we are to regard the social world as a community of autonomous agents despite the fact that perfect realisations of autonomy are nowhere to be found… I am not ignoring wolf-boy cases, solitary Robinson Crusoes, etc. Such cases don’t affect my argument here, which is about how agency arises in the non-exceptional cases (where these are the huge majority), and about how agency essentially arises (where “essentially arises” does not mean “always arises”, any more than “humans essentially have two legs” means “all humans have two legs”). 20 18 There is at least this much truth in Dennett 1987’s well-known idea of the “intentional stance”: persons constitute each other as persons, and as agents, by treating each other as persons and as agents. Something like the Davidsonian “principle of charity” (Davidson 1980) is at work in our mutual interpretations; and the pun is, perhaps, a felicitous one. (“My attitude towards him is an attitude towards a soul. I am not of the opinion that he has a soul”: Wittgenstein, Philosophical Investigations, II, p.152.) There is a kind of constitutive prolepsis21, as we might call it, in the most basic attitude that one person extends to another: by interpreting the other as a person, I make it true that he is a person. This is most obviously true in the case of children. Think of how parents interact with toddlers. The child staggers across and plonks a book down on his sibling’s lap; the parent’s reaction is “How kind of you to let her share your book!” Is that, in fact, what the toddler was doing? The right answer to this question can be: “Yes it is—once the parent has given this reaction.” The parent’s reaction is an interpretation of the child’s deed. Before the reaction, perhaps, there was no fact of the matter about what the child was doing. The child did not know himself what he was doing; for all he knows about it, his action might as well have been simply random, an indeterminate piece of behaviour. But now that his mother offers her interpretation, the child accepts that interpretation as true; on his mother’s authority, he learns to see himself a certain way. Because his mother sees his act as having certain a meaning (as the giving of a gift), and because the child sees his mother as seeing his act this way, and because his mother sees the child as seeing her as seeing his act this way… (Grice 1957), the act comes to have that meaning. Talking of Grice, the same pattern is even clearer in the case of languagelearning. A baby begins by babbling—that is, by producing all the sounds the human mouthparts can make. Some of these get a response because they are sounds that occur in the parents’ language, while others don’t because they are not. The baby “homes in” on the sounds, and then on the patterns or combinations of sounds, that get a response; then it learns to correlate particular patterns of sounds with particular contexts by the same homing-in process (cp. 1.3). What the parents say is “She’s learned to say ‘tiger’” (my own first child’s first word). Seen another way, what has happened is that the parents, by treating their daughter as if she has the word for, and then the concept of, a tiger, have made it true that she has that word (and, a little later perhaps, that concept too). This “constitutive prolepsis”, whereby treating someone as if she was a person makes it true that she is a person, is not only a feature of parent-child relations. Throughout life it remains true, as I put it above, that no actual person’s agency is either complete or pure. Every person at every time is incomplete, because each is a bearer of Darwall too talks of prolepsis in this connection (2006: 87-88): “there seem to be many cases where we swish to hold others accountable though we seem to have very good evidence that they are not free to act on moral reasons in the way our practices of holding someone fully responsible seem to presuppose… In some instances, for example with children, we seem simultaneously to move on two tracks in the process of inducting them into full second-person responsibility, sometimes treating them proleptically as though 6they were apt for second-personal address as a way of developing moral competence while nonetheless realising (“objectively”) [in the sense meant by Peter Strawson in ‘Freedom and Resentment’] that this is an illusion that must also be recognised.” 21 19 unrealised potential for agency: if for no other reason, because no one can pursue all the goods that are available at any time (5.5), but most likely also because no actual agent does all they could to be an agent. It is familiar to the point of banality that we all lapse sometimes into torpor or sloth or an inability to take the opportunities in front of us to “become what we really are”; even more banally, we all go to sleep and relax sometimes. And every person at every time is impure, because each can be interpreted, by a sufficiently determined sceptic, as not really an agent at all, but a victim of the instinctive or other determining external forces that constantly threaten anyone’s agency. Given these forms of essential human imperfection, the constitutive prolepsis always involves a kind of idealisation. No real agent is entirely pure or entirely complete, but the constitutive prolepsis enables us to understand other beings as persons by understanding them in the light of the ideals of the purity and the completeness of agency. These two forms of imperfection also give persons a crucial sort of vulnerability. It is remarkably easy to attack persons by withdrawing the constitutive prolepsis from them. We can do this, to one degree or another, by explaining away what they do, or by refusing to listen to them or to take them seriously or as persons, or (in Kant’s phrase) by refusing to treat them as “ends in themselves”. At the limit, we can do it by declaring them “non-persons” (as the Nazis did to the Jews), by pretending that they simply aren’t there (as was done to the aboriginal Australians under the notorious doctrine of terra nullius), or by treating them merely as objects (as often happens in pornography, violence, and murder—and sometimes in applied ethics, too). The reason why persons are so vulnerable to these forms of attack is because it is so easy to stop seeing them as persons. The constitutive prolepsis is, as I say, an idealisation, and that means that there is some truth, or at least plausibility, in views that reject that idealisation: for example, the person’s body is a physical object. This is why all of the agents we all actually know about—the human ones—are, in Alasdair MacIntyre’s splendidly apt phrase, “dependent rational animals”. Humans are dependent rational animals because they cannot live well except against the background of a society of other dependent rational animals who will extend to them the constitutive prolepsis. This society is what I mean by “the human community”. Perhaps it is now beginning to be clear why I say that proposals like Singer’s about how to use the word “person” are an attack on this community. The human community is made possible only by the willingness of agents to extend not only to each other, but also to those who are not clearly actual agents at all (not yet, or not at the moment, or not any more), a certain kind of interpretive charity. This is needed because—very obviously to start with, less obviously, but still genuinely, later on— everyone is a marginal case of agency. Hence, it is only by way of a society of agents’ willingness to treat marginal cases of agency with some idealisation, as at least potentially more than marginal, that any agents at all (including the socialised agents themselves) can ever become actually more than marginal. The continuing process whereby this “de-marginalisation” happens and keeps on happening is the heartbeat of the human community. Singer’s proposal stops this 20 heartbeat, because it disallows the attitudes of interpretive charity. On the story that he and his allies tell us, it is (in effect) only those who are currently displaying the personal qualities who are to count, at present, as persons (and so, to count morally at present). The objection to this restriction is not just that it is very difficult to assess the “factual evidence” for what Singer is prepared to count as agency (though it certainly is difficult). Nor is the objection just that, as compared with the normal extent of ordinary good people’s natural interpersonal sympathy, concern, and interest, Singer’s proposal is a disastrous narrowing of the circle22 (though that is also true). More profoundly, the objection is that Singer’s proposal commits him to rejecting the very attitudes of openness towards other human agents, actual or potential, central or marginal, without which there can be no human agents (still less a community of human agents). The human community’s openness to, and interpretive charity about, this sort of variety of degrees of agency is mirrored in some linguistic evidence: the looseness and ambiguity of the term “person” itself. In ordinary language, “person” can certainly, at one end of its semantic range, mean something like what Singer wants it to mean, a “rational and self-conscious being”. Yet at the other end of its range, the word can also be a synonym for “human animal”. If what I have just argued is right, then it is part of the point of our person-concept to allow those who are persons only in the “human animal” sense to enter into a continuum of meaning towards the far end of which they can become persons in much stronger senses. (Even if it is also true that, in a way, no one ever becomes a person in the fullest sense: for no one actually achieves the ideals of purity or completeness that I described above. They are ideals.) Singer’s refusal to countenance this inclusive continuum of meaning exactly parallels his restrictive refusal to let the human community’s interpretive charity range as widely as ordinary good people naturally allow it to: which is at least widely enough to include all human animals. Such proposals as Singer’s about how to define or redefine “person” are therefore an attack on the human community, because they are a denial of its constitutive interpersonal attitudes. As such, these proposals should be firmly rejected. Instead we should hold fast to the claim that I made at the start of this section: that for each and every person that we (at any rate) all know about, to be a person is to be a human animal. There may yet appear to be further difficulties facing that claim. I address three in the next section. 7.7. Three objections to the thesis that persons are human animals 1. When does the human animal begin and end? In 7.5-7.6 I have argued against Singer-style revisionism about the meaning of “person”, and in favour of retaining its normal continuum of meanings from “human animal” to “rational self-conscious being”, I am of course alluding to Singer’s book title The Expanding Circle. There is a fine Orwellian irony about more than one of Professor Singer’s book titles. Just to cite the most egregious example, that the author of Unsanctifying Human Life should be director of a “Centre for Human Values” seems to be a bad case of putting the fox in charge of the chickens. 22 21 with a default presumption in favour of the former, wider sense. My case has been that Singer-style revisionism is wrong, and bad, not for a metaphysical reason but for a moral one: because it undermines our participation in the fundamental interpersonal attitudes without which none of us can hope to be understood and accepted as an agent. If this argument against revisionism is correct, then the normal meaning of “person” still applies; so in the broad sense of the word, “persons” means “human animals”. But what, it might be asked, are human animals? When do they begin, and when do they end? Despite endless attempts to muddy the waters, the answer to this question remains very clear—indeed, given recent advances in medical technology, it is now clearer than at any previous point in human history. It does not take metaphysics to teach us that a human life begins at conception and ends at death; medical science does that, if common sense has not already done it. True, the process whereby sperm and ovum fuse and produce a zygote is not instantaneous and has trackable stages. And—depending of course on how death happens—there are often a number of different points in the dissolution of a human physiology any of which might, with some plausibility, be labelled as the point of death. So there is some unclarity about both these concepts. But not enough to matter here. I need not get into contemporary debates about the definition of death to make the only point I want to make, namely that, whatever exactly we count as death, death has, by definition, to be what constitutes the natural23 end of the human animal. As for the unclarities in the production of a zygote, these typically all happen within approximately a one-day span, so that it is enough for my purposes simply to bracket that span, and say that anything before the brackets is definitely not an individual human being in utero, and anything after the brackets definitely is. Do we then still feel the urge to insist that the early foetus is in some sense not really a human being—because it is not “viable”, perhaps, or because it does not yet have a well-formed brain, or because at some early stages it (= the one human being) might yet split and become two human beings, or because it is, in the usual phrase, “only a cluster of cells”?24 But these are pitiful evasions. As excuses for refusing to bring human animals who happen to be unborn just as much within the scope of the human community’s constitutive interpersonal attitudes as those who happen to be born, they are interesting mainly because they are such clear illustrations of how swiftly and abruptly Singer-style revisionism can narrow the circle of our natural sympathies. Here the truncation is effected mainly, perhaps, by a blank refusal to think about early foetuses as existing in time. It serves Singer’s purposes to look on the foetus, not as an astonishingly dynamic physiological system that is changing and developing towards a 23 A natural end is one thing, a supernatural end another. But I leave that aside here. Both these last two ideas are in Singer 1993: 156-7, which presents the startlingly bad argument that since an early embryo can split into two embryos, it is “not even an individual”. So does my unity as an individual now depend upon a fact about the future—that ingenious aliens will never abduct and replicate me? Nowadays this startlingly bad argument is startlingly often taken seriously. For a forty-page discussion which its own author admits gets us no further forward, see Guenin 2006: 463-503. 24 22 visible form maximally like “our”25 own all the time we are debating whether to destroy it, but as it were as an alien species consisting of time-slices. The thought to be avoided at all costs, in Singer’s way of looking at things, is the thought that we too are, if you like, no more than “clusters of cells”, that we too are surprisingly bad at “surviving independently”, that we too have depressingly undeveloped mental powers, and that in fact that thing there is just another instance of the very same kind of thing as we are—at an earlier stage of development, to be sure, and yet with a heartbeat from twenty-two days, brain activity from seven weeks, the ability to hiccup from twelve weeks, the ability to sleep from eighteen weeks, to suck its thumb by twenty weeks, to feel pain by perhaps twenty weeks, and with the same brain-wave patterns as a newborn by twenty-six weeks. But this thought, that human foetuses are things of the same kind as we are (in most of the senses of “kind” that could conceivably apply), is a thought that will occur to anyone who thinks clearly about them for thirty seconds. That Singer-style revisionism must, apparently, seek to prevent this thought from occurring is just more evidence of the disturbing impairment of our moral vision that is induced by Singer-style revisionism.26 2. The constitutive prolepsis can be misplaced. My argument has been that the human community depends for its existence upon the willingness of agents to extend to one another an attitude of interpretive charity—in effect, to make it true that others are more fully and truly agents than they previously were, by treating them as if they already had a pure and complete measure of agency to which, in fact, no one ever entirely attains. Singer’s revisionism is an attack upon the human community because it is an attack upon this constitutive prolepsis (as we might call it). However, I am sure to meet the response that the constitutive prolepsis is not always justified, and that where it is not justified, there is nothing damaging or undermining of human community about failing to offer it. No one above the age of six attempts to draw an armchair, or a boulder, or a fire engine, into the community of agents by taking a fundamental attitude of interpretive charity towards it. The question is, then, where the constitutive prolepsis is the right attitude to take. And (it will be said) it is surely not the right attitude, wherever the creature to which it is extended could not become an agent in any sense—which is surely the case with some human animals. The first thing to note about this objection is how much a Singer-style revisionist would have to concede to offer it. To object that we should not extend the constitutive prolepsis to cases where there is little or no realistic chance of the emergence of anything like agency is to concede that we should extend it to cases where there is such a chance. Singer-style revisionists do not typically wish to concede that: they usually want to withdraw the constitutive prolepsis from human animals in all sorts of cases where, quite uncontroversially, those human animals not only can become as fully agents as anyone is, 25 Scare quotes because, of course, a central part of what is at stake here, as so often elsewhere, is who “we” are, and whether “we” includes the unborn. (In general it is a good idea to avoid appeal to the “us” in ethics; the results are nearly always unfortunate.) 26 A striking, and psychologically revealing, example of this (Singer 1993: 182, my italics): “Killing infants cannot be equated with killing normal human beings”. In what possible sense of the word are infant human beings abnormal human beings? 23 but will become such agents unless something goes wrong—healthy neonates and lateterm embryos, to cite the most notorious examples. The second thing to note about the objection is that it misses part of the point of my proposal, as I have stated it. I have argued that constitutive prolepsis is always, in a sense, less than fully justified, because it is always an idealisation. It is only by way of our treating the agent under the aspect of the ideals of purity and completeness that I have described that any agent can realise those ideals even partially. But no actual agent ever fully realises those ideals; as I pointed out above, they remain ideals. That does not make it wrong or unjustified to apply them to persons. Very many things from the world of human life can be seen as exercises in constitutive prolepsis which appeal in this sort of way to an ideal which, like any ideal, is never fully realised. (The basic sort of constitutive prolepsis, the sort that grounds attributions of personhood, is not the only sort.) Consider the theory of rights, for instance. We ascribe the rights to freedom of assembly and freedom of speech to every human being, even though we know that acutely asocial loners may well never (or at least never again) actually exercise either right. The behaviour of parents, besides (as described) involving the basic constitutive prolepsis in a particularly clear form, is also full of other constitutive prolepses that might, on a hard-headed revisionist analysis, be regarded as quite unrealistic. Parents give their children the opportunities to play the cello, to learn Spanish, to use their monthly allowances sensibly, to be polite to Great Aunt Maud (…and so on indefinitely), even though they may be fully aware that every one of these opportunities will be spurned. In minimally decent jurisdictions the treatment even of hardened criminals displays the same pattern. The jurisdiction goes on offering the criminals opportunities to reform, even though it is as good as certain that these opportunities will not be taken up. So in general, the constitutive prolepsis that accords X the status due to a φer does not have to involve any kind of expectation that X will ever actually φ. What it does have to involve is the understanding that X is of a kind that characteristically includes φers. The human kind characteristically includes individuals who freely assemble and speak; that is why it is not ridiculous to give the rights to free assembly and speech to any and every human, including asocial misanthropes and those who choose never to utter a word. Contrast the equine kind, which does not characteristically include any such individuals; that is why it is ridiculous to give these rights to any and every horse. Considered as a kind, children characteristically include individuals who learn the cello or Spanish or wise use of money or good manners to venerable relatives (well… eventually). That is why the parents who try to teach these lessons to their children are not acting absurdly even if they know in advance that their children are not going to respond—whereas parents who tried to teach these lessons to the family goldfish or the rubber plant would be acting absurdly. Similarly, there is nothing absurd about offering the old lag a place on the rehabilitation programme, even if it seems certain that this old lag will always refuse or abuse the offer: and the reason why not has to do with the fact that some old lags do sometimes reform. 24 And so in the basic case of the constitutive prolepsis that accords the moral status of persons, what is needed to justify this constitutive prolepsis to a given individual X is not the expectation that X will ever display actual agency. What is needed is the understanding that X is of a kind that characteristically includes persons. Now the kind human being is such a kind (cp. 7.5); and that is why all human beings have the moral status of persons. That brings me on neatly to the last objection I wish to consider in this section: the charge of “speciesism”. 3. Speciesism. Peter Singer writes: whether a being is or is not a member of our species is, in itself, no more relevant to the wrongness of killing it than whether it is or is not a member of our race… we can now look at the [human] foetus for what it is—the actual characteristics it possesses—and can value its life on the same scale as the beings with similar characteristics who are not members of our species. (Singer 1993: 150-1) A number of curiosities arise in this passage. Perhaps the most obvious is Singer’s move from the explicit claim, before the dots, that the foetus’s being a human is not a morally relevant characteristic, to the implicit claim, after the dots, that the foetus’s being a human is not a characteristic at all—not even a morally irrelevant one. “We can now look at the foetus for what it is”, he tells us, at its “actual characteristics”. Apparently he means that being human (or “fully human”) is not an actual characteristic of the foetus, that being human is not even part of “what it is”. This conclusion is not only a non sequitur from the premiss that being human is a morally irrelevant characteristic. It is also a bizarre conclusion. For any kind of animal, the question “What is it?” evokes the species-classifying answer (“It’s a cat”, “It’s a dog”, “It’s a human”, etc.) more naturally than almost any other answer. Common sense takes any being’s species-membership not only to be one obvious characteristic of that being, but perhaps its most obvious characteristic of all—in a sense, its defining property. So if human foetuses are indeed human, and beings (and despite much well-intentioned coaching I fail to see what else they could be), then “looking at the foetus for what it is—the actual characteristics it possesses” is bound to involve looking at the foetus’s humanity. For this is an actual characteristic that it possesses. (Singer’s blindness to this possibility is strikingly reminiscent of Locke’s blindness about how we can be aware of a thing as the thing it is, as well as of its properties: 7.5.) “But its being a human is not a morally relevant characteristic.” Isn’t it? Why shouldn’t it be? As I argued in 7.5, there are no metaphysical or logical constraints on what kinds we may or may not take to be morally significant or relevant. So there is no reason in logic or metaphysics why we should not regard “mere difference of species” as morally significant. If there is an argument against seeing them that way, it will have to 25 come from somewhere else. But so far as I can find, Singer’s only direct argument for the moral irrelevance of species is the following argument by analogy27: (1) Race- and gender-classifications do not ground morally relevant characteristics; (2) species-classifications are like race- and gender-classifications; so (3) species-classifications do not ground morally relevant characteristics. Any argument by analogy is formally invalid. And this particular argument by analogy would be unsound even if it were valid, because its first premiss is false. In all sorts of ways, race- and gender-classifications do ground morally relevant characteristics. When, for example, three new employees arrive in an overwhelmingly white male workplace, there are two very good reasons for the employer to show special courtesy and warmth towards the one who is black and female—as I rather think Singer would agree (see 1993: 40-41). So Singer does not provide us with a sound argument that species is morally irrelevant. He provides us only with an analogy. Like any analogy, the most that this can legitimately achieve is an invitation to see things a certain way: in this case, an invitation to see species-classification as morally on a par with race- and gender-classification. The upshot of 7.6’s argument about constitutive prolepsis is that we should politely refuse this invitation. In all sorts of ways, species-classification is not morally on a par with raceand gender-classification. In particular, I have argued, “humanity” is the name of a kind such that its members are characteristically persons and agents; so it makes sense to extend to all members of that kind the basic attitude of interpretive charity that involves treating its members as at least potential persons and agents. By contrast, our best information is that other terrestrial species are not such that their members are characteristically agents; so it does not make sense to extend that attitude to all the individuals of any other species. No parallel contrast holds with race or gender. Though both distinctions are real and (in some circumstances) morally significant, neither of them discriminates between a pair of kinds such that members of the first kind are characteristically agents, and members of the second kind are not: that is self-evident. In the one respect that really matters for Singer’s case against speciesism, the analogy on which his argument depends is simply a failure. We have, therefore, been given no reason to accept Singer’s contention that species-classifications are morally insignificant; his argument for that conclusion is simply unsound. I hope to have presented a more successful argument for the opposite view: that species-classifications have, or can have, a fundamental moral significance. If this view is speciesism, then what I have presented is a defence of speciesism. However, a little more precision is still needed about what the view is; so let me register some qualifications and caveats. 27 At any rate, when arguing against speciesism Singer only provides us with an analogy. We are supposed to come to that argument with our minds prepared by Singer’s argument for his own positive view about what matters: his sentientism (1993: 57). More about that later in this section. 26 I have argued that there is a human moral community which is created and sustained by the basic constitutive prolepsis: by the adoption of an interpersonal attitude of interpersonal charity in ascriptions of agency. This attitude is naturally, and rightly, extended to every member of the human species. This view does not imply that members of other species could not be drawn into the human moral community. On the contrary, it implies three distinct ways in which this could happen. One is that we could discover a whole species—aliens, or angels, or whatever—which is like the human species in that its members are characteristically agents. A second is that a species already known to us might, en masse, somehow change so as to become characteristically agents. As Peter Carruthers depicts this second possibility: Suppose that the experiments attempting to teach language to chimpanzees had been successful beyond their originators’ wildest dreams. The apes in question gained a complete mastery of English within a few years, were able to attend school and later university, and made many close friendships with human beings. (Carruthers 1992, Ch.3) I agree with Carruthers that “in these imaginary circumstances it would plainly be absurd to claim that the apes lacked moral standing, or had a moral importance that was lower than our own”: if this scenario is realised, we will then be just as obliged to extend the basic constitutive prolepsis to every member of the chimpanzee species as we are to extend it to every member of the human species. Or is it just “to every chimpanzee involved”? The question—to which I think the answer is indeterminate—has point because it brings us to the third way in which members of other species might be drawn within the scope of the constitutive interpersonal attitudes. We could discover good evidence that some particular individual creature—Locke’s parrot, say (Locke Essay 2.27), or Washoe the chimp (Singer 1993: 111)—is in some good sense an actual agent, even though it is a member of a species whose members are not characteristically agents. If this happens, we will then be obliged to extend the basic constitutive prolepsis also to this individual. I say “if this happens”. As I said above, I suspect our best information so far is that the only persons we all know about are humans. But I am simply too ignorant of the evidence to be able to commit myself on the interesting question of whether Washoe, for instance, really is (or was) an agent, or a person. Persons, I said in 7.3, “are unique in nature because persons alone can act upon the goods… to organise and co-ordinate them into coherent narrative structures in their lives”. Does Washoe do this, or can she do it? (De re, of course; to require that she be able to do it de dicto would be to require her to accept my theory of ethics.) That is a difficult evidential question. By which I do not exactly mean an empirical question: the question is a hermeneutic one, about how it is right and reasonable to interpret Washoe, and in particular how much interpretive charity it is reasonable to extend to her. There are other constitutive prolepses besides the one that I have called “basic”, the one that grounds ascriptions of personhood. Some of these are reasonably extended to a cat or a dog—both species to whose typical members it is perfectly reasonable to ascribe desires, beliefs, intentions, decisions, emotions, and 27 moods, even if not moral agency in the sense of decisions about how to pursue the goods—and Washoe surely gets at least as much interpretive charity as a cat or a dog. And maybe more; but however it turns out with Washoe, proponents of the present theory can afford to watch the research develop with open-minded interest. There is nothing in the present theory to say that we could not discover that Washoe is a person. For all the present theory says, we could yet discover that every chimp is a person. Getting straight what we should say about Vulcans and Washoe may also clarify how we can answer an interesting objection to speciesism that is offered by Peter Carruthers: It seems plain that species membership is a morally irrelevant characteristic… suppose it had been discovered that human beings in fact consist of two distinct species, otherwise hardly distinguishable from one another, the members of which cannot inter-breed. In these circumstances it would plainly be objectionable for the members of the majority species to attempt to withhold moral rights from the members of the minority, on the mere ground of difference of species. This, too, would be obvious speciesism. (Carruthers 1992 Ch 3) Carruthers is quite right to claim that it would alter nothing morally speaking if we discovered that “the human species” was in fact two different biological species— perhaps homo neanderthalis and homo sapiens—living together in a single moral community.28 Pace Carruthers, however, his argument does not show that that there is no species-classification that bestows participation in the moral community on its members, and so that species-classifications are morally irrelevant. What his argument shows is only that there may be more than one species-classification (or other kind-classification) that bestows participation in the moral community on its members. That is something that we should grant anyway given the possibility of other species whose members are characteristically persons; and it is not inconsistent with the claim that speciesclassifications are morally significant.29 A different objection to my argument is the claim that species don’t exist: that as a matter of strict biology, there is no clear differentiation between humankind and “other species”, only a differentiation between different parts of the same evolutionary family tree, on which family tree all actual parent-child relations are literal parent-child relations holding, not between species, but between individual organisms; species therefore are populations, not Aristotelian (or Kripkean) natural kinds (Sober 1994). This point does not affect my argument, which can work equally well with the population conception of species (or indeed with any conception at all that will allow “humanity” to count as the 28 Some anthropologists think that this has actually happened: Trinkaus and Shipman 1993. In any case Carruthers’ own view about animal rights is much closer to speciesism than to animal liberationism. He holds that there is an indirect argument, arising from contractualism, for granting all humans and no animals moral standing (for a brief exposition see his online paper “The animals issue” at http://www.philosophy.umd.edu/Faculty/pcarruthers/The%20Animals%20Issue.pdf). Since I think it is obviously true that (non-human) animals do have moral standing in their own right, based roughly on the wrongness of causing pain and suffering, it seems I am much closer to sentientism than Carruthers is. 29 28 name of a kind of some kind). The idea that a population of creatures sharing a common genetic heritage, physiology, and ethology can sensibly be treated as a unitary entity does not seem a controversial one. And that is all I am committed to meaning by “species”. One last clarification brings us, as promised in footnote 27, to Singer’s sentientism. Nothing in the view I have developed here implies that the suffering and pain of other species does not matter (at all, or very much). The view does not even imply that animal pain, considered just as such, is any less morally significant than human pain, considered just as such. “No matter what the nature of the being, the principle of equality requires that the suffering be counted equally with the like suffering of any other being” (Singer 1993: 57): on this principle, I simply agree with Singer, and agree with him too that the principle implies that our current practices towards animals, in particular experimentation and meat-eating, need radical re-examination. I would only add two remarks. First, though pain matters, it is not the only thing that matters, and not the thing that matters most, either: that is part of the point of postulating a variety of goods. Second, however alike human and animal pain may be, human and animal suffering (which is not the same thing as pain) are usually quite unlike; so Singer’s words “the like suffering” will not often apply. A cat can be tortured, but it cannot undergo what King Lear suffers; human happiness is capable of far greater heights and refinements than animal happiness, and the depths and agonies of possible human suffering are correspondingly deeper.31 This is why “the limit of sentience” is not “the only defensible boundary of concern for the interests of others” (Singer 1993: 58). Certainly that limit marks one distinction between the ways we should take an interest in the welfare of other creatures. Hacking an orange to pieces is morally unproblematic in a way that hacking a frog to pieces is not. And this is partly because an orange certainly cannot feel pain, and a frog pretty definitely can. But interestingly, it is surely not only because of that; destroying a living creature for fun is morally objectionable in a way that wantonly destroying a piece of fruit for fun is not, even if the living creature cannot feel pain. (Hacking trees to pieces for fun is morally objectionable too.) Unless we think nothing matters except pain and pleasure—and this is not, officially, even Singer’s own position (1993: xi)—there is no reason to think that the sentientist’s distinction is the only moral distinction we can make among beings; nor that it is the most important. 7.8. The wrongness of killing I’ve argued (7.3-7.7) that each person is an individual good, and that his or her capacity for agency is a central part of what makes him or her an individual good of the Singer himself seems to recognise this possibility when he writes (1993: 107) that “it would not necessarily be speciesist to rank the value of different lives [e.g. a fish’s and a snake’s] in some hierarchical ordering… In general it does seem that the more highly developed the conscious life of the being, the greater the degree of self-awareness and rationality and the broader the range of possible experiences, the more one would prefer that kind of life, if one were choosing”. But, he concludes, “at this stage the question remains open” (1993: 109); so far as I know, he does not return to it at any later stage. 31 29 distinctive kind that persons are; and that we should take the interpersonal attitude at least to all human animals, and possibly to some other animals as well. These commitments show us a little more about how to spell out what counts as pursuing the good represented by any person, and what counts as violating that good. I suggested in 7.3 above that “pursuing the good of persons” means “pursuing the reasons that the existence of some person gives you to promote or help develop that person’s agency”, and that “respecting persons” means “acting on the reasons that the existence of other persons gives you not to undermine, threaten, or destroy their agency”. We can further develop these two characterisations by connecting them with the idea of the interpersonal attitude. If taking the interpersonal attitude towards a person is as morally crucial as I have described it, that helps to explain why it is good to pursue a person’s good by helping that person develop her agency, and what might be involved: again, a loving parent’s attitude towards a child is a paradigm here. It also helps to explain why it is good to respect the value of a person by giving that person, so to speak, the space for agency—not undermining or attacking her capacity freely to act so as to shape and organise the goods that she finds around her. Here the paradigm, or one paradigm, is exactly the sort of non-interfering respect that good adults display towards each other, often (but not only) within a relationship of friendship. Since persons are loyalty-goods rather than interest-goods (6.2), there are, no doubt, many ways to fail to respect the good of another person, and so violate that good. Broadly speaking the importance of these violations will be proportionate to the degree to which they threaten or disable the existence of that person. The most significant violation of the good of another person, therefore, will be the one involved in destroying him: that is, in killing him. Persons are individual goods; killing them means violating those goods, because it means deliberately and directly choosing to destroy them. This is my theory’s explanation of what’s wrong with killing. The explanation is not only simple; it also avoids the “odd indirectness” that I noted in some other moral theories in 5.2-5.3. Compare four other answers to that question, which I would say exemplify this indirectness: The wrongness of killing us is understood in terms of what killing does to us. Killing us imposes on us the misfortune of premature death. That misfortune underlies the wrongness... The misfortune of premature death consists in the loss to us of the future goods of consciousness... What makes my future valuable to me are those aspects of my future that I will (or would) value when I will (or would) experience them... What makes killing us wrong, in general, is that it deprives us of a future of value. Thus, killing someone is wrong, in general, when it deprives her of a future like ours. (Marquis 1997: 95-96) [The wrongness of killing a person is] the wrongness of permanently depriving her of whatever it is that makes it possible for her to value her own life. (Harris 1985: 17) 30 What is fundamentally wrong about killing, when it is wrong, is that it frustrates the victim’s time-relative interest32 in continuing to live. (McMahan 2002: 195) According to preference utilitarianism, an action contrary to the preference of any being is, unless this preference is outweighed by contrary preferences, wrong. Killing a person who prefers to continue living is therefore wrong, other things being equal. (Singer 1993: 94) These answers treat the wrong of killing as a wrong of deprivation. Killing is wrong because it is taking something away from someone—his future, or whatever he values, or the satisfaction of his interests or preferences. This sort of view of the wrongness of killing has an obvious and fatal flaw: it cannot explain what is wrong with killing someone who is not going to have a “future life of value” anyway, e.g. because if you don’t kill him at 3.58.59, he will be struck by a freak thunderbolt at 3.59.00. (Or does the wrongness of killing him diminish, the closer you get to the time when he will die whatever happens?) That point aside, there are deeper grounds for objecting to such answers. The more we consider their sort of view of the wrongness of killing, the more it is likely to seem unnatural and superficial. It is apt to leave the reader haunted by an inarticulate sense that something crucial has been left out. It is usually true, of course, that to kill someone is to deprive them of good things. Still, that deprivation does not seem the main thing wrong in killing, even when it is part of what is wrong. In killing, the main point is not that something is taken away from someone, but that the someone is taken away. The central wrong in killing is not the wrong of depriving the individual person (not even: the wrong of depriving her of everything). It is the wrong of destroying the individual person.33 Elizabeth Anscombe shares this sense of something crucial left out: You can argue, truly enough, for example, that general respect for the prohibition on murder makes life more commodious. If people really respect the prohibition against murder life is pleasanter for all of us—but this argument is exceedingly comic. Because utility presupposes the life of those who are to be convenienced, and everybody perceives quite clearly that the wrong done in murder is done first and foremost to the victim, whose life is not inconvenienced, it just isn’t there any more. He isn’t there to complain; so the utilitarian argument has to be on behalf of the rest of us. Therefore, though true, it is highly comic and is not the foundation; the objection to murder is supra-utilitarian. (Anscombe 1972: 25)34 32 McMahan 2002: 80 makes the distinction between an interest and a time-relative interest like this: “One’s interests are concerned with what would be better for one considered as a temporally extended being... One’s present time-relative interests are what one has egoistic reason to care about now”. 33 There is then nothing vacuous about saying that what’s wrong with killing is that it violates a good; pace Christopher Coope, 2007: 121, who describes this explanation of the wrongness of killing as a “rigmarole”. Surely it is the convoluted attempts to bring in “future lives of value” etc. that deserve this unkind epithet, not the explanation that says simply that killing is wrong because it means destroying something good, namely the person. 34 Thanks to Cora Diamond for bringing this passage to my attention. 31 Proponents of the deprivation view sometimes seem to recognise the force of this objection. So, for example, Jeff McMahan can sometimes present a view about the wrongness of killing that looks very similar to my own: To kill a person, in contravention of that person’s own will, is an egregious failure of respect for the person and his worth… because killing inflicts the ultimate loss—the obliteration of the person himself—and is both irreversible and incompensable, it is no exaggeration to say that it constitutes the ultimate violation of the requirement of respect. (McMahan 2002:.242) How then does McMahan square this apparent recognition of the intrinsic value of the person with his claim, cited above, that the wrongness of killing lies in the frustration of interests? His answer is to distinguish between the badness of death and the wrongness of killing: While the badness of death is correlative with the value of the victim’s possible future life, the wrongness of killing is correlative with the value or worth of the victim himself… a person, a being of incalculable worth, demands the highest respect. (McMahan 2002: 242) But it is hard to make sense of this distinction in the context of McMahan’s own philosophy. That the wrongness of killing “is correlative with”—i.e., depends on?—“the value or worth of the victim” is not what McMahan says on p.195 of his book, as quoted above. If, despite p.195, this is the source of the wrongness of killing, then it is hard to see why the badness of death does not arise from the same source. But if both the badness of death and the wrongness of killing arise from the victim’s value or worth, then the deprivation view no longer gives the main reason for either.35 The deprivation view of the wrongness of killing shares the key weaknesses of its close philosophical relative, the receptacle view of the self. To see killing as wrong only because of what it prevents persons from instantiating, and to see persons as valuable only because of what they can instantiate, are obviously two parts of the same implausible outlook. Perhaps McMahan moves some small distance away from this outlook in the last quotation—but it is not obvious that the move takes him to a coherent position. It is interesting to note that, at least once, Singer too explicitly claims not to hold a receptacle view about persons: Rational, self-conscious beings are individuals, leading lives of their own, and cannot in any sense be regarded merely as receptacles for containing a certain quantity of happiness… In contrast, beings who are conscious, but not selfconscious, more nearly approximate the picture of receptacles for pleasure and pain. (Singer 1993: 126) However, in assessing McMahan’s claims about the value of the person, we should bear in mind that McMahan’s views about what makes something a person are revisionist in the same sense as Singer’s. 35 32 However, this denial makes little difference to Singer’s overall position. To put it simply, he may not see persons (in his sense of the word) as receptacles for happiness, but he does, apparently, see them as receptacles for preference satisfaction. 7.9. The wrongness of killing: three apparent, but unreal, exceptions So far, my claim is that killing persons is wrong because individual persons are intrinsic goods of the highest level, and killing persons involves choosing to destroy these goods. In this last section of this chapter, I shall discuss three cases that might seem to be counter-examples or exceptions to this rule, namely self-defence, abortion, and euthanasia, and argue that none of them is. a. Self-defence. In 6.5 I argued that the value ethicist holds that we are required to honour (or respect) all goods at all times, and free to promote (or pursue) any good at any time. The value ethicist is not, therefore, vulnerable to the objection raised by “the paradox of deontology”, that if he thinks violations of goods are so bad, then he ought to see it as permissible to perform one violation in order to prevent five violations from being performed. For the value ethicist is not committed to promoting the nonperformance of violations tout court, but to promoting the non-performance of violations within the limits set by the requirement not to perform violations. Interestingly, common moral intuitions, while supporting this line at many points, also seem to suggest one (at least one) exception to it. This is in the case of killing in selfdefence, or more generally, in defence of anyone against unjust aggression. A very widespread moral intuition says that there is nothing wrong with killing an aggressor to prevent him killing you, or others who are presenting no parallel threat of aggression to anyone else. It might seem hard to square this intuition with value ethics’ solution to the paradox of deontology. If I defend five innocents (i.e. people who are not engaging in unjust aggression themselves) by killing an aggressor who is threatening their lives, it looks like I am killing one person to prevent five people from being killed. In which case, apparently, I am promoting the non-performance of violations without limiting my promotion of that value by the requirement not to perform violations myself. My response to this case is to bite the bullet: we should question the intuition that it is permissible to kill one person to prevent other people being killed. More precisely, we should deny that it is permissible to choose, deliberately and directly, to kill one person to prevent other people being killed. What is permissible is to use potentially lethal force against someone to prevent other people being killed, foreseeing that this force may cause death, but not intending it. If someone (or some group of persons, like an army) chooses unjustly to make a lethal weapon of himself, then lethal force may be used against him qua lethal weapon; but even then, not qua person. An individual, or an army, may be targeted with lethal force so long as (we have good reason to think) he/ it remains a lethal threat—which normally36 means, so long as the intention persists to destroy us or 36 There can be innocent threats, of course, such as those discussed at length in Kagan 1991. In general, it is open to us to destroy such innocent threats, especially where the innocent threat only exists because some malicious agent has created it; though of course the moral limits on what we may do to ward off innocent 33 other innocents; but no longer. The fact that someone is an unjust aggressor changes the rules about proportionality, about what we may foresee and allow to happen to persons who are unjust aggressors, up to and including their deaths; it does not change the rule that we may not aim, deliberately and directly, at killing them (as opposed to aiming at destroying lethal weapons). Here I am of course appealing, in advance, to chapter 8’s defences of the principle of double effect and action/ omission distinction. “But if this were how people thought about war, it would be very difficult to fight wars at all!” Is it so easy to find reasons to regret that? Wars cannot be justified unless the analogy that is so often used to justify them, that between the individual person’s selfdefence against an individual aggressor and the state’s self-defence against a state which is an aggressor, is genuinely apt, as in fact it rarely is. Hence, the cases where wars are genuinely justified are far fewer than they are generally taken to be. (Our intuitions about war are a striking case of the harm that can be done by Plato’s long-lived analogy between state and individual; we must learn to see that in reality states are much less like individual persons than they are often supposed to be, and in particular much less real than individuals.) b. Abortion. Abortion as it happens in our society involves the killing of a foetus, an unborn child. This does not happen by accident, but by design; in abortion as our society knows it, the death of the unborn child is intended. This makes abortion a deliberate choice to kill a person. My argument thus far therefore implies that abortion is wrong. It also implies that the most persuasive argument yet produced against the wrongness of abortion, Judith Jarvis Thomson’s (Thomson 1971), is an irrelevance. Suppose we grant that Thomson’s famous “violinist” argument shows that someone (call her “the donor”) who is hooked up, against her will, to become a crucial part of the lifesupport system of some random stranger (call him “the violinist”), is permitted to extricate herself from the set-up, even though she foresees that this will lead to the violinist’s death. Suppose we grant also that the violinist/ donor relation is relevantly similar to the foetus/ mother relation. (I’m sure we shouldn’t grant this—as Finnis 1973 points out, parent-child relations are morally quite unlike relations between any two random strangers; but I don’t need that objection to make my point here.) Even with those assumptions, Thomson’s argument does not show that the donor may extricate herself from the set-up by intentionally killing the violinist. Likewise in the case of pregnancy, then, even if Thomson’s argument establishes the permissibility of the mother’s foreseeing the child’s death, it does not establish the permissibility of her intending the child’s death. But as I pointed out in the last paragraph, in abortion as our society knows it, the death of the unborn child is intended, not merely foreseen. (Here is a second advance appeal to chapter 8’s distinctions.) Thomson’s argument is therefore irrelevant, in this respect, to abortion as our society knows it. (It is irrelevant in other respects too: Thomson builds it in to her story that the mother needs a good reason to threats are more constricting than on what we may do about non-innocent threats. A hostile fighter plane may simply be destroyed; a hostile fighter plane with innocent hostages on board should be brought down without killing anyone on board, if that is possible. 34 extricate herself from the life-support set-up, whereas many real-life abortions are performed merely, as they say, “on demand”.) I said that Thomson’s is the most persuasive argument37 for treating abortion as a counter-example, or an exception, to the rule that killing persons is impermissible; but we’ve just seen that her argument fails. It follows that abortion is no exception to that rule. Indeed there are fewer cases where the moral imperatives that lie behind that rule, which as I said in 7.6 have to do with taking the interpersonal attitude to other human animals so as to draw them proleptically into the human community, are more strongly in force. The rearing of children, and their introduction into the human community, is a paradigm case of good human action: if anyone deserves to be accorded the privileges of treatment as a person, it is those who are not yet, or who are now coming to be, persons. c. Euthanasia. I have argued that there is a general rule against killing persons, and that the cases of self-defence and abortion provide no counter-example to this general rule. Does euthanasia? The commonest arguments in favour of euthanasia are these two: (1) “It is sometimes in a person’s interest to die; therefore euthanasia should sometimes be administered.” (2) “Some patients choose to die, and we have no business interfering with their choice; therefore euthanasia should sometimes be administered.” We may call these the interests and the autonomy arguments for euthanasia. Both fail to establish any reason for making an exception to the general rule against deliberately and directly killing persons. The interests argument is unsound for several reasons. One is that it is a non sequitur; “p is in X’s interest” in no way entails “Y should act to bring about p”. Even if there were persons such that it was in their interest to die, it would not follow that it was anyone’s duty to kill them; it would not even follow that it was anyone’s duty to let them die. For any such conclusion to follow, various other premisses would have to be true as 37 To me, the most persuasive other argument in defence of abortion is the natural intuition that Thomson was perhaps trying to spell out: the thought that it is outrageous to demand of a woman that she allow her body to be invaded and occupied internally by another human body that she does not wish to be there. I sympathise; I find this thought entirely understandable. But how does the thought about invasion justify deliberately killing a human being, as a matter of standard procedure? (The point of that last italicised clause is that I can see how such an assault might indeed justify a drastic response in a one-off case: but in our society we are not talking about a one-off case, we are talking about a mechanised, bureaucratised, mass-production practice.) What the thought might justify, again as a matter of common practice rather than one-off emergency, is attempts to transplant an embryo from one woman’s womb to another’s: many women, after all, would love to be pregnant, but are not. But it is a striking fact about the remarkable gynaecological technologies that we have developed that embryonic transplant technology has not been one of them. I think that tells us that this thought about invasion is not the main objection that people have had to leaving unwanted pregnancies alone. At most, that thought has been operating in combination with the (I have argued) completely unpersuasive thought that the embryo is morally insignificant. 35 well. Most notably, it would have to be true that there was no moral objection, in the circumstances, to killing or allowing death—which is precisely the question at issue here. Similar remarks apply to the claim that some lives are “not worth living”. Even if this claim is true, it is a further grave step to say that any such life should be ended, either by our action or by our non-intervention. The claim anyway involves the same confusion as I noted before (7.4), between a person and that person’s life. Views about the value of a life are not the same thing as views about the value of a person unless we treat the value of the person as nothing over and above the value of the life. But, I have argued, we should not allow that conflation. A second and deeper problem with the interests argument is that its premiss is false. What could it mean to say that it is in someone’s interest to die? Of course it could be someone’s preference to die, which is enough for a preference utilitarian like Singer. But preferences are not reasons, nor necessarily even grounds of reasons; for preferences can be mistaken, even informed and considered preferences. Intuitively, what typically shows preferences to be mistaken is their lack of match with a person’s interests; from which it plainly follows that interests are not preferences. What then are interests? The value ethicist’s answer to that will be based on the theory of the goods (7.1-3): a person’s interests are his potential participations in the various goods by the exercise of his agency, and what is in his interest is whatever characteristically tends to promote or protect such participation. But death is naturally understood38 as the end of the person. It cannot therefore be understood as promoting or protecting either his agency or his participation in the goods. And that means that it cannot be in his interest. But can’t someone sometimes be “better off dead”: in such a sad state of existence that it would be preferable for her not to (continue to) exist at all? Once again we should note that, even if this could happen, it would not in itself give anyone a licence to kill. But in any case the answer to the question, strictly speaking, is No. We cannot compare a person’s level of welfare39 in two possible situations, only one of which is a situation where that person exists. For obviously, in the other situation, there is nothing to compare. The idea that there is something to compare between these two situations is created by the assumption that, where the person does not exist, her welfare level can so to speak be set at zero—which is of course better than a negative welfare level. But this is just a mistake. A situation where someone does not exist is not a situation where her welfare level is zero, or equivalent to zero. Rather, it is a situation where her welfare level is equivalent to nothing. And it would be an elementary fallacy to take this occurrence of “nothing” as meaning “zero”. The claim that someone can be better off dead does not presuppose the ability to measure a negative welfare score against a zero welfare score; it presupposes the ability to measure a negative welfare score against “Not Applicable”. It is not merely pedantic or playing with words to point out that there is no such ability. But unless we assume that there is, it is hard to assign any more than an 38 Prescinding here from how death might be supernaturally understood. In this paragraph I grant for the sake of argument that it makes sense to talk about levels of welfare in a numerical fashion. Outside the paragraph, I do not grant this. 39 36 illusion of meaning to the common talk of being better off dead, and of lives not worth living. “Maybe you’re right that death can’t promote someone’s interests. Still, there is a sense in which it can protect his interests, in the sense that his death at a given time spares him from terminal indignities that spoil the shape of his life”: so, in part, Ronald Dworkin argues in Life’s Dominion. I say “in part”, because, of course, the paraphrase of his argument that I’ve just constructed talks about death, not about killing people. As noted already, someone might agree that death can indeed come at a time that spares the dier a succession of indignities, and yet not see that as a reason for killing anyone. (Those who wish here to invoke the slogan “Dying with dignity” should perhaps ask themselves about the exact content of the morally-loaded notion of dignity, and about whether it could ever be truly dignified to do the serious wrong to oneself or another that, if my arguments are right, is involved in euthanasia.) More generally, the right response to a medical condition is, obviously, to treat the condition, not to kill the patient. If there is pain, there should be pain relief; if there is suffering there should be therapy; if there is indignity there should be the restoration of dignity. In all these cases, to kill the patient is, apart from anything else, a grotesquely irrelevant choice, reminiscent of Caligula’s cure for Gemellus’ cough—to have his head chopped off.40 It is truly a remarkable fact that killing is all too often the only option for the “treatment” of many terminal patients that philosophers today will take seriously.41 (In this respect philosophers are, fortunately, very different from doctors, and indeed from most of those who have any actual involvement in terminal care.) There is, then, no good reason to think that it is ever in a patient’s interest to die. Obviously, it might be in a patient’s interest to be freed from pain, or from suffering or indignity. Also obviously, the only way to free the patient from pain (or whatever) might be to do something that causes or hastens her death. But what does that show? It shows that the palliative-care approach to those who are terminally ill and in great pain gets things exactly right, whereas the euthanasiac approach gets things crucially wrong. 42 The palliative carer believes that he has reason to free the patient from pain, and in certain circumstances is prepared to accept the patient’s death as a side-effect—possibly an inevitable side-effect—of pain relief. What the palliative carer does not think he has reason to do, unlike the euthanasiac, is aim at the patient’s death: he aims at this neither as a means to releasing her from pain, nor (of course) as an end in itself. If death is never in anyone’s interest, indeed is the negation of all interests, we can see why the palliative carer is right to see his reasons this way. There is no reason to do what is not in anyone’s Or so Robert Graves reports in I Claudius. There is little trace of the story in, e.g., Suetonius’ Life of Caligula. 41 One can find oneself the object of a disturbing undercurrent of anger if one dares to challenge euthanasiac thinking. It is sometimes made to sound as if the opponent of euthanasia were personally responsible for the sufferings of the terminally ill, and as if the only person offering them real help is not the person who wants to offer them comfort and pain-relief, but the person who wants to kill them. This bizarre inversion of values seems evident to me in my debate with Joachim Jung and Richard Taylor in Philosophy Now issues 40 and 41 (Chappell 2003). 42 For information on the palliative-care approach, go to http://www.cicelysaundersfoundation.org/. 40 37 interest; and there is positive reason not to do what negates people’s interests. So there is no reason to perform any act of euthanasia, and every reason not to perform it. (Here is a third advance appeal to the next chapter’s distinctions.) So much for the interests argument for euthanasia. What about the autonomy argument, which says that “we have no business interfering with the choice” of patients who choose to die? This robust assertion that autonomy is paramount has a grand ring to it. But the grand ring has a hollow echo. It makes preferences sovereign, and we have already seen that that is not a tenable position. We don’t generally think that autonomy is paramount. We don’t think that we are bound to respect people’s freedom of choice anywhere and everywhere. If someone’s autonomous choice is to rob banks, we think they should be stopped, not allowed to get on with it. In short, we think that autonomy is paramount only where people are making choices that we already think are not seriously wrong. In the present context of argument, this assumption begs the question; for whether euthanasia should be an exception to the general rule that it is wrong to kill persons is precisely what is at issue. I conclude that no reason has been shown to see the cases of self-defence, abortion, and euthanasia as exceptions to the rule against killing persons. As already noted, my discussion and rejection of these three putative exceptions to the rule against killing persons has led me to appeal to two familiar distinctions: between foreseeing and intending, and between acting and omitting to act. These distinctions are an integral part of the moral theory that I am developing anyway, but in this chapter their importance to the overall case for value ethics has come out particularly clearly. It is therefore high time for a full and explicit account of these distinctions. To that task I turn in Chapter 8. 38