Psychological altruism, evolutionary origins, and moral rules

Philosophical Studies

Volumn 89, Issue 2-3, 1998, Pages 283-316

  Kitcher, Philip ^a  

^a UNIVERSITY OF CALIFORNIA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0000237711     PISSN: 00318116     EISSN: 15730883     Source Type: Journal    
DOI: 10.1023/a:1004241214825     Document Type: Article

References (57)

1. The present essay is a fragment of a much longer work-in-progress. Parts of this work have been presented to audiences at the University of California at Irvine, University College London, Princeton University, Stanford University, Washington University, the University of California at San Diego, and the University of Illinois, as well as at the Pacific Division Meeting of the American Philosophical Association. Many people have given me helpful advice and constructive criticisms. I am particularly grateful to Richard Arneson and Brian Skyrms for their suggestions, and to David Brink for helpful written comments on the penultimate draft.
2. I should note explicitly that it is this last condition that seems to pose most difficulties for an evolutionary explanation of psychological altruism. It is also worth pointing out that I want to adopt a liberal conception of B's "presence", taking B to be present when A recognizes that the available actions have an impact on B's preferences. The third condition may be easiest to satisfy when, although B is not physically present, A can perceive that B would benefit from certain options available to A: imagine cases in which A has the opportunity to summon B to share a good.
3. Thus I intend the notion of valuational structure to subsume the concepts of utility matrix and of the representation of games in normal form.
4. The minimal condition for A to be altruistic is that A should take B's preferences into some account in evaluating options - as my account of weak altruism requires. But we are typically interested in much stronger types of altruism, those which play a systematic role in the readjustment of preferences so that there is a difference to action. The ensuing discussion will show how these can be captured within the framework I am proposing.
5. For a relatively simple version of a case like this, see O. Henry's short story The Gift of the Magi. A much more complex and interesting example occurs in the closing pages of Henry James' The Golden Bowl. Martha Nussbaum gives a brilliant discussion of the nuances of the Ververs' perceptions of their mutual accommodation in her essay " 'Finely Aware and Richly Responsible' : Literature and the Moral Imagination" in Love 's Knowledge, New York: Oxford University Press, 1990, pp. 148-167.
6. I used this term in "The Evolution of Human Altruism", Journal of Philosophy, 90, 1993,497-516. It is important to note that, in that essay, I had a rather different conception of the relationship among the psychological states of altruists, so that although the earlier account is similar to that presented here it is not the same.
7. I mention this putative principle, because it reveals the kinds of questions that my account of psychological altruism ought to address. This essay concentrates only on simple cases, and the relations among dimensions of altruism will be explored in future work.
8. In such instances, the altruist will think of the beneficiary's actual preferences as not according with the beneficiary's interests, either because the beneficiary intends to sacrifice her own interests to promote those of the altruist or because she fails to have a correct conception of what her interests are. If one subscribes to a view of a person's good on which it is not reducible to what the person would want under epistemically improved (or ideal) circumstances, then one may also think that there are further cases in which the altruist takes into account the beneficiary's real interests and not her preferences (or even her suitably refined preferences).
9. In "The Evolution of Human Altruism", my notion of paternalistic altruism was only partial. I am grateful to Richard Arneson for helping me to see this.
10. The recent literature in primatology supplies abundant evidence for the first two of these claims, and can, I think, be used to defend the third. See especially Dorothy Cheney and Robert Seyfarth How Monkeys See The World (Chicago: University of Chicago Press, 1990) Chapters 3 and 8; Jane Goodall The Chimpanzees of Gombe (Cambridge MA.: Harvard University Press, 1986);
11. C. Bachmann and H. Kummer "Male Assessment of Female Choice in Hamadryas Baboons", Behavioral Ecology and Sociobiology, 6, 1980, pp. 315-321 ; and R. Byrne and A. Whiten (eds.) Machiavellian Intelligence (Oxford: Oxford University Press, 1988), particularly the essay by Nicholas Humphrey.
12. Many, though not all, of the essays in Machiavellian Intelligence adopt this perspective, and, for a more pronounced articulation of the theme that intelligence is a tool for calculating egoists, see James Barkow, Leda Cosmides and John Tooby (eds.) The Adapted Mind, Oxford: Oxford University Press, 1992. Interestingly, some of the papers that inspired the tradition explicitly allow that the Hobbes-Machiavelli picture of ruthless, self-interested agents may be inadequate to account for the psychology of social animals. See, for example, Nicholas Humphrey "The Social Function of Intellect" (Machiavellian Intelligence pp. 1326) especially p. 23.
13. See "The Evolution of Human Altruism", where I offered both the kin selection example and a much more extensive argument for the possibility that golden-rule altruism can originate and be sustained under natural selection. For reasons that will become apparent later, the selection regime envisaged there, involving iterated optional Prisoner's Dilemma can only be a small part of the story of the evolution of psychological altruism. I should also note that my earlier discussion did not respond adequately to the second type of worry about altruistic dispositions, since if has seemed to some people that reciprocal altruism could be fuelled by explicit calculation.
14. See Mary Midgley Beast and Man, Ithaca: Cornell University Press, 1978 and my Vaulting Ambition.
15. For a particularly lucid formulation of this point, see Joel Feinberg "Psychological Egoism", reprinted in Feinberg (ed.) Reason and Responsibility, Belmont CA.: Wadsworth.
16. Goodall The Chimpanzees of Gombe, Cambridge MA.: Harvard University Press, 1986.
17. Frans de Waal, Good Natured, Cambridge MA.: Harvard University Press, 1996.
18. De Waal, Good Natured, pp. 131-132.
19. My usage of this term is intended to resonate with the suggestions of David Hume and Adam Smith, both of whom I take to have been broadly right about our sentiments. As we shall see, my picture of human agency diverges from theirs in some important respects.
20. This was worked out in John Batali and Philip Kitcher "Evolution of Altruism in Optional and Compulsory Games", Journal of Theoretical Biology, 175, 1995, pp. 161-171.
21. See "The Evolution of Human Altruism", and "Evolution of Altruism in Optional and Compulsory Games".
22. Axelrod's pioneering work is summarized in The Evolution of Cooperation, New York: Basic Books, 1984, which served as the stimulus for the work that Batali and I undertook. Interestingly what we showed is not quite what we (and our predecessors) set out to prove, to wit the stability of cooperation. Rather, we demonstrated the instability of the absence of cooperation, and indeed, the absence of high levels of cooperation. (In our simulations, the frequency of cooperation in any generation is either extremely low or else very high). In "The Evolution of Human Altruism", I argued further that the iterated optional PD regime would favor the emergence of golden-rule altruism. I would now put the conclusion a little differently: the formal relationship among preferences corresponds to the golden-rule value, θ = 0.5, but it is possible that this relationship is set up as the result of explicit calculation of expected benefits (in defiance of condition (4) on weak psychological altruism). This conclusion hasn't yet been tested on computer simulations.
23. See, for example, Jane Goodall's detailed story of a hunt, The Chimpanzees of Gombe pp. 288-289.
24. See De Waal Chimpanzee Politics and Peacemaking Among Primates. Of course, it is possible that the high frequency of grooming is an artefact of the captive situations in which primates have been most closely observed, although studies of wild populations do support the idea that grooming occurs far more often than is needed for purely hygienic reasons.
25. Most obviously, if, in addition to the hygienic benefits, there's a sizeable social payoff that can only be achieved through cooperative behavior, then the game being played is no longer optional PD - indeed, the largest payoff by far may be associated with cooperation with animals who are longstanding allies.
26. This is not to deny that the phenomena investigated by Axelrod and his successors (including Batali and me) don't have a place in the understanding of social relations, cooperative behavior, and altruism, but rather that these can only be properly understood when we have a clear view of the social matrix in which they are set. An analogous point applies to the kinds of interactions explored brilliantly by Brian Skyrms (Evolution of the Social Contract, Cambridge: Cambridge University Press, 1996). I develop this point in a commentary on Skyrms' book (to appear in a symposium in Philosophy and Phenomenologie-al Research).
27. See Peter Rodman and John Mitani "Orangutans: Sexual Dimorphism in a Solitary Species" in Barbara Smuts et al. (eds.) Primate Societies, Chicago: University of Chicago Press, 1987, pp. 146-154. I have heard from several people that the "standard view" outlined in this article has been challenged by further studies of female associations, but I do not know of any published sources.
28. See Donna Leighton "Gibbons: Territoriality and Monogamy", and Kelly Stewart and Alexander Harcourt "Gorillas: Variation in Female Relationships", both in Smuts et. al. Primate Societies (pp. 135-145, and 155-164, respectively).
29. It may be pertinent to note that some human groups of hunter-gatherers tend to undergo fission when their size exceeds 150 to 300. One example is the Yanomamo; see Napoleon Chagnon "Fission in a Yanomamo Tribe", The Sciences, 16, 1976, pp. 14-18.
30. See R. Wrangham "On the Evolution of Ape Social Systems", Social Science Information, 18, 1979, pp. 334-368;
31. "An Ecological Model of Female-Bonded Primate Groups", Behaviour, 75, 1980, pp. 262-300;
32. "Social Relationships in Comparative Perspective", in R. Hinde (ed.) Primate Social Relationships: An Integrated Approach (Oxford: Blackwell, 1983), and "Evolution of Social Structure" in Barbara Smuts et al. (eds.) Primate Societies (Chicago: University of Chicago Press), pp. 282-296. Wrangham bases his approach on the idea that the principal determinant of female reproductive success will be her access to food, and that the principal determinant of male reproductive success will be the ability to copulate as frequently as possible with an estrous female. On his account orangutans pursue their relatively solitary lives because females can most efficiently forage for fruit by working alone, and males have the physical abilities to defend a home range including the smaller home ranges of several females. As will be clear below, I'm going to abstract from some of the particularities of Wrangham's discussion, offering a more general model of which his explanation would be a special case.
33. "Evolution of Social Structure" p. 290. Compare Hobbes: "the weakest has strength enough to kill the strongest, either by secret machination, or by confederation with others that are in the same danger with himself" (Leviathan p. 183). However, Hobbes would not have thought that this could apply to the brutes, because, without speech "there had been amongst men neither Common-wealth, nor Society, nor Contract, nor Peace, no more than amongst Lyons, Bears, and Wolves" (Leviathan p. 100). Hobbes underrated the lions and the wolves, and knew nothing of the chimpanzees and bonobos.
34. For similar suggestions to the effect that the most important issues about evolution of cooperation and sociality need to move beyond PD, see R. Noe "Alliance Formation among Male Baboons: Shopping for Profitable Partners", in A. Harcourt and F. de Waal (eds.) Coalitions and Alliances in Humans an Other Animals, Oxford: Oxford University Press, 1992, pp. 285-321. Several of Noe's points about baboons are analogues of claims I have made about chimpanzees. See, for example, his finding that coalitions and alliances can persist even when one of the parties never receives what is apparently the main prize.
35. Note that the fitness values that occur in the payoff matrices for these games have to be computed by taking into account the consequences of the actions for the underlying alliances and coalitions to which the agent belongs. This articulates the point made earlier that the structure of animal interactions cannot be understood in isolation from the demands of the most fundamental game, here identified as the coalition-formation game.
36. I hope to present these in a book that will work out the project begun in this essay.
37. See, for example, John Maynard Smith, Evolution and the Theory of Games, Cambridge: Cambridge University Press, 1982, p. 108.
38. The argument is non-constructive because it says that there will be coalitional structure without saying what it is. As will become clear, I don't know how to determine the structure, even given full information about the initial conditions. So the claim is roughly as useful as the mean-value theorem, which tells us that there is a maximum of a function within a particular interval, without specifying where it is. That sort of information was good enough for a celebrated application of Reaganomics - and, in the present context, we'll see that the absence of a constructive argument has philosophical implications.
39. Chimpanzee Politics, Baltimore: Johns Hopkins, 1984, p. 126. My subsequent discussion will explain, and partially endorse, this reading of de Waal's. It's important to note, however, that the picture of cunning political agents, given in de Waal's first book is softened in his second (Peace-Making Among Primates, Cambridge M.: Harvard University Press, 1989), and replaced by an emphasis on animal sympathy and altruism in his most recent (Good Matured). Although I cannot fully document it here, I think that my analysis of the psychological altruism found in chimpanzees and bonobos explains the complex findings recorded by de Waal, offering a perspective from which all his observations fit together.
40. Chimpanzee Politics p. 128.
41. Peacemaking Among Primates Chapter 2. De Waal makes the important observation that Lull's desire to remain with his troop was so strong that it was difficult to remove him, even after he had been severely wounded.
42. I should acknowledge that there are obvious concerns, given our ignorance of the nature of the dependence, about the testability of attributions of conflicting dispositions. A general response to these concerns would note that we should not be bound by any operational imperative; more specifically, I think that there are forms of behavior that provide evidence for the claim that an animal has a disposition to produce conflicting valuations (see the text below). The approach taken here has some kinship with Walter Mischel's emphasis on the failure of cross-situational consistency in people who have stable personality profiles (see W. Mischel and Y. Shoda "A Cognitive-Affective System Theory of Personality: Reconceptualizing Situations, Dispositions, Dynamics, and Invariance in Personality Structure", Psychological Review, 102, 1995, pp. 246-268; I am grateful to George Mandler for drawing my attention to the problem of trait consistency, and suggesting that I read Mischel's work). There are also connections between the approach I am adopting here and some of Anthony Damasio's work on decisionmaking and its neural disruption; see his Descartes' Error, New York: Putnam's 1994. For reasons of space, I can't elaborate the affinities here.
43. This provides a start on an answer to worries about testability. We can sometimes suppose that conflicting valuations are reflected in different signals sent to various parts of the body - one valuation would be relevant to the state of one bodily system but not to another, and conversely, so that the overall state of he body is a mosaic. In general, we might test the presence of attitudes not by focusing on behavior but on readiness for behavior.
44. T. Schelling "The Intimate Contest for Self-Command", Chapter 3 of Choice and Consequence, Cambridge MA.: Harvard University Press, 1984, pp. 57-82. The discussions on pp. 60-61 are especially pertinent.
45. For details of peace-making strategies, see de Waal Peacemaking Among Primates.
46. I deliberately use terminology of Harry Frankfurt ("Freedom of the Will and the Concept of a Person", Journal of Philosophy, 68, 1970, pp. 5-20)
47. and of Allan Gibbard (Wise Choices, Apt Feelings Cambridge MA.: Harvard University Press, 1990). It seems to me that Frankfurt's and Gibbard's insights can be connected with the evolutionary perspective I am developing. I suspect that Gibbard would approve of this, although Frankfurt might view it as misguided and unnecessary. 44 A number of people (most forcefully Bas van Fraassen) have expressed surprise that the author of a book critical of human sociobiology might want to speculate in this area. I am sensitive to this concern. My best attempt to lay it to rest runs as follows: human morality is a natural phenomenon, that must fit, somehow, into our natural history; the last chapter of Vaulting Ambition identifies some ways in which I think that it does not fit, but my criticisms there will never be fully convincing unless I can devise a better account; any such account should not aspire to do more than show how human morality could have begun (in standard terminology, it is a "how possibly" explanation - see Vaulting Ambition p. 72 ff.); but finding any account that coheres with what we know about other primates and about ourselves is very hard, and it seems to me that standard philosophical accounts of our moral psychology are in constant danger of conjuring illusory faculties and abilities that are quite at odds with what we take ourselves to know about animal behavior; thus it seems worthwhile offering an avowedly conjectural picture and exploring how the problems of moral philosophy look from its perspective. (Although this is my favorite response to a common concern, it does not satisfy me on every day of the week.)
48. Edward Westermarck's comparative survey of social rules in "primitive" societies, The Origin and Development of the Moral Ideas, London: MacMillan, 2 vols., 1906-1908, is a mine of comprehensive information.
49. For an argument that this was Darwin's strategy see my essay "Darwin's Achievement" (in N. Rescher [ed.] Reason and Rationality in Science, University Press of America, 1985).
50. Moore Principia Ethica, New York: Prometheus Books, 1988, p. 188 (section 113).
51. See Vaulting Ambition Chapter 11, and "Four Ways of Biologicizing Ethics", in Elliott Sober (ed.) Conceptual Issues in Evolutionary Biology (second edition), Cambridge MA.: MIT Press, 1994.
52. For representative versions of the kinds of naturalism I have in mind, see Alvin Goldman Epistemology and Cognition (Cambridge MA.: Harvard University Press, 1986), Fred Dretske Naturalizing the Mind (Cambridge MA.: MIT Press, 1994), and my own The Nature of Mathematical Knowledge (New York: Oxford University Press, 1983) and The Advancement of Science (New York: Oxford University Press, 1993).
53. This, of course, is Bernard Williams' famous "one thought too many". See
54. "Persons, Character and Morality", in Amelie Rorty (ed.) The Identities of Persons,
55. Berkeley: University of California Press, 1976, pp. 197-216, especially pp. 213-215.
56. Although (1) and (3) represent cartoon versions of what goes on in our moral lives, the cartoons are influential. A presently unpublished lecture by Barbara Herman has convinced me that Kant's actual views are much more nuanced and intricate, and Hume's moral psychology is also more sophisticated (as his discussions of artificial virtues make clear). But neither Hume nor Kant nor the philosophers they have influenced seem to escape the idea that we have a coherent set of preferences, and hence do not recognize the need to address the four types of moral agency as equally fundamental.
57. Partly, this is for reasons offered by Schelling in Choice and Consequence, which show how systems of normative control can be inept or incomplete. Further problems result from the fact that we can recognize dispositions in the agent that produce valuations clashing with the system of normative control as genuine expressions of the agent's character and desires. The philosophical tradition is heavily inclined towards the view of the system of normative control as constituting the wishes of a "real" self, but psychodynamic conceptions stemming from Freud should remind us of the repressive possibilities of a system of normative control. See, for two very different examples, Freud's The Ego and the Id, and George Ainslie's Picoeconomics (Cambridge: Cambridge University Press, 1992). Samuel Scheffler has offered a penetrating account of the implications of some aspects of Freud's work for our understanding of central issues in moral psychology, and I think that the picture I sketch here could be articulated to capture his insights. (See Scheffler Human Morality, New York: Oxford University Press, 1992, Chapter 5.)