Leonard D. Katz
Toward Good and Evil
Evolutionary Approaches to Aspects of Human Morality
Moral Origins and Adaptation
People are curious about how things began and especially about the origins of things they deem important. Besides satisfying such curiosity, accounts of origin may acquire broader theoretical or practical interest when they go beyond narrating historical accident to impart insight into more enduring forces, tendencies or sources from which the phenomena of interest more generally proceed. Accounts of evolutionary adaptation do this when they explain how and why a complex adaptation first arose over time or how and why it has been conserved since then, in terms of selection on heritable variation. Thus the eye was shaped and conserved over evolutionary time on account of its usefulness for sight and if we want to explain why its lens is transparent but its pupil black we shall tell how transmitting light to the retina but preventing scatter behind it both help us see. In such cases evolutionary accounts of origin may provide much of what early Greek thinkers sought in an arche, or origin a unified understanding of somethings original formation, source of continuing existence, and underlying principle (Huffman, 1993, pp. 7892; McKirahan, 1998).
We should not expect to find so single and unified an account of human morality as of the eye or visual system because both the goals concerned and the means to these are likely much more varied. Our eye contributes to our survival and fitness mainly through the single function of sight; a long history of adaptation under uniform constraints of optics and physiological possibility have led to a fairly uniform result. To the extent that human morality is the result of specific adaptations for human social life, which has long varied between groups, times and environments, we may expect their contributions to our capacity for this social life to be more various and facultatively variable. We may wonder with Darwin how exactly to divide the credit for morality between natural selection, culture and learning, but suspect like him that, especially at the later stages of the evolution of morality, culture and learning, both individual and social, had the larger roles (1871, pp. 8081, 166, 394, 404). Still, as Noam Chomsky says, it certainly seems reasonable to speculate that the moral and ethical system acquired by the child owes much to some innate human faculty and is rooted in our nature (Chomsky, 1988, p. 153). Minimally, morality has roots in our mammalian social nature, which goes back to the motherchild bond (Cicero, De Finibus Bonorum et Malorum, sects. 623, reporting Stoic views; Darwin, 1871, pp. 801; Midgley, 1991, p. 8) and this may be as deeply rooted on both sides as anything in our evolved psychology is.
It is, however, much less likely that morality is the product of only one or very few informationally encapsulated specialist mental modules than that language is. Moral considerations are typically accessible to consciousness and deliberation, as the intermediate products of syntactic and phonological processing are not. And moral competence involves the appropriate use of diverse kinds of information to govern the deployment, restraint and balancing of different motives and emotions, such as compassion and indignation, in situationally sensitive and open-ended ways. Even where the perception of moral demands of diverse kinds may be psychologically mandatory and clear, their weighting against each other and against other reasons may be painfully problematic rather than automatic. Consonantly, whereas language seems to depend on discrete brain systems vulnerable to local lesions that may leave other functions largely unimpaired, the brain damage that most specifically impairs moral competence and development seems to impair cognitiveaffective connections minimally, the cognitive conditioning of fear and anxiety and the situation-sensitive generalization of behavioural restraint built on these more generally (cf. Damasio et al., 1990; Damasio, 1994; Anderson et al., 1999). And studies of amoral criminally- inclined psychopaths or sociopaths (who may sometimes be so disposed more by psychological adaptation to opportunity-poor or socialization-poor environments than by originally abnormal temperament based in physiology) seem compatible with this view (see Wilson, 1993, pp. 1058; Lykken, 1995; Mealy, 1995; Black, 1999, Ch. 6). These considerations suggest that morality may be the outcome of interactions between various systems that support and regulate social and affective learning and responding more generally and others that do much else besides, rather than of any single or of very few dedicated strictly modular faculties (cf. Darwin, 1871; Midgley, 1991; Flack and de Waal, Principal Paper and Response, and Railton, below).
Morality, however, may be a product of natural selection without having been itself selected as a functional adaptation (Flack and de Waal, Response, p. 74), by being an outcome of tendencies and capacities that were. Individual learning may organize moral competence from these, by adaptation to a human environment shaped by social learning and cultural evolution as these adaptive processes, operating at different levels and timescales (Brian Skyrms, Response, p. 335), interact. The integration of inherited quasi-modular capacities with human representational, inferential, epistemic, and communicative capacities (Peter Railton, below, p. 58) may be largely self-organizing, as elsewhere in cognitive development it seems to be (Carey, forthcoming). But moral development may also adaptively respond to the demands of life stages, from early obedience, through dealings with peers, to nurturing, as Dennis Krebs suggests (below, p. 320).
The four principal papers presented here, with interdisciplinary commentary discussion and their authors responses, represent contemporary approaches to an evolutionary understanding of morality of the origins from which, and the paths by which, aspects or components of human morality evolved and converged. Their authors come out of no single discipline or school, but represent rather a convergence of largely independent work in primate ethology, anthropology, evolutionary biology, and dynamic systems modelling on related problems, conjectures and tentative conclusions. In inviting contributions I deliberately made no attempt to define morality more sharply than common language and understanding have left it, including our ordinary responses to right and wrong, but not all of the very diverse thinking about this and other practical concerns and about what we should make of all these that ethics encompasses. This was not because of any lack of past definitions of morality, but because the history of controversy over these makes it advisable not to prejudge the question pending our inquirys results. From these we may see some outlines of plausible answers begin to emerge, only a brief sketch of some of which I can attempt here, sampling the hypotheses, scientific support, critical discussion of this, and refinement of positions in authors responses.
Toward a Natural Prehistory of Morality and Politics
In our first principal paper, Jessica Flack and Frans de Waal describe aspects of the social behaviour and inferred psychology of monkeys and apes, the living animals most closely related to us, in which they find building blocks for the later evolution of human morality. Following Darwin, they, like Christopher Boehm in his following contributions, seek psychological traits we share with our closest relatives and presumably also with our common ancestor, from which, given the course of human cognitive evolution, and the need to adjust individual and shared interests in a social life, morality would naturally emerge. They find in the simple rules, expectations, order, and sense of social regularity implicit in the social interactions of group-living primates elements of rudimentary moral systems arising out of the interaction of sympathetic, cooperative and competitive sentiments and the interpenetration of cognitive judgement and emotion, in ways that prefigure human morality and that can be used as clues to its evolution and dynamics. But as they make clear in responding to Kummer, Railton and others (p. 68, below), it is in the sentiments and capacities underlying the social behaviours they discuss, rather than in the behaviours themselves, that they find the raw material for the evolution of human morality.
The following commentary and controversy lines up partly along generational lines, in keeping with the generation-long progressive relaxation of norms against ascribing cognitive and conscious states to nonhuman animals in behavioural science. Human developmental psychologist Jerome Kagan, the firmest sceptic, asserts that animals have no conscious intentions and feel no guilt for past failures to live up to internalized standards, which they lack. Two veteran primatologists voice lesser degrees of scepticism about the cognitive and moral capacities of apes. Irwin Bernstein doubts that questions about animal intention and awareness of norms are experimentally tractable and apparently believes that less cognitive explanations of behaviour should therefore be preferred. Flack and de Waal in their Response oppose to such cognitive parsimony a principle of evolutionary parsimony, a heuristic assumption of similarity between closely related species (such as humans and chimpanzees) for the guidance of hypothesis formation and research (pp. 723, below), which would tell also against Bernard Thierrys suggestion that if elements of rudimentary moral systems are attributed to chimpanzees they might as well be attributed to the less cognitively sophisticated macaque monkeys with which he works, on the sole ground of similar behaviour. Hans Kummer also resists some of the cognitive and motivational interpretation for which de Waal especially is known, emphasizing that moral judgement is at the core of morality. But he also suggests relevant experiments, as does the younger experimentalist Josep Call, for investigating non-human primates understanding of intention and of social norms.
Flack and de Waals discussion of calculated reciprocity, conflict resolution and community concern, especially among chimpanzees, leads naturally to our second principal paper, in which cultural anthropologist Christopher Boehm leads us down a path moral evolution may have taken starting from the ancestor we had in common with chimpanzees and bonobos, which presumably like these descendants had inherited tendencies to dominate others but also to dislike being dominated. Boehm hypothesizes that an ancestral regime of despotic social dominance and fearful submission led in the human line to successful collective action to intentionally suppress tyrants and bullies in nomadic bands like Chomsky, seeing opposition to such forms of oppression, hierarchy, domination and authority as characteristic of morality (Chomsky, 1988, p. 154). The distributed consciously exercised social control thus established was then used for other ends besides, constituting morality as a sense of right and wrong that is born out of group-wide systems of conflict management based on shared values that constrains individual behaviour through a system of approval and disapproval as Flack and de Waal express their agreement with this aspect of Boehms view in their Response (p. 69, below). While people are thus able through human culture, cognition and language to collectively suppress social dominance hierarchy in a way not possible for chimpanzees and bonobos and to more effectively avoid and manage intragroup conflict, in so doing they draw on a potential for concerted collective action that is occasionally observed at work in chimpanzees and that is presumably based in capacities inherited from a common ancestor.
Bruce Knauft, a cultural anthropologist who has long been developing theories related to Boehms, stresses, as does the primatologist Bernard Thierry, the difference language makes in enabling the sharing of explicit norms, but argues that symbolic communication had been coevolving with brain enlargement even during our Homo erectus past. This permitted the symbolically mediated internalization of socially prescribed norms as part of a larger suite of features including constraints on sexual behaviour, a rudimentary division of labour, the gendering of morality, socialized control over junior males [especially with regard to their access to mates], increased group size, [and] increased home range as well as the reduction of bullying behaviour, and a reverse dominance hierarchy that enforces egalitarianism among fully adult males posited by Boehm (p. 138, below). Ethnographer Robert Dentan also remarks the generally unequal status of women and younger males and observes that the moral order established by the local arbiters of this may not serve equally the interests of all members of even simple societies. Donald Black advocates a pure sociology that dispenses with all psychology, individual or collective, and seeks to explain the social control aspects of morality in terms of social structure and relations alone. Boehm in his Response defends his psychological explanation of the arising of egalitarian morality in small, autonomous groups (in which all people lived in Paleolithic times and some still do today) as the product of deliberate collective action and the permanent threat of the same, with reference to the ethnographic evidence. And evolutionary psychologist Dennis Krebs discusses the need for cooperators to curb cheating and free-riding, and to curb dominance as a means to this end, and the mixed motives that lead people to use morality to restrain each other even as they seek to evade the same restraints themselves. His genetic individual selectionist approach, which draws especially on the work of R.D. Alexander, contrasts with those presented in the remaining two principal papers.
How Kindness and Fairness Can Evolve and Survive
Elliott Sober and David Sloan Wilsons book, Unto Others: The Evolution and Psychology of Unselfish Behavior (1998), which is summarized and commented on next, and Brian Skyrms paper, which draws in part on his book, Evolution of the Social Contract (1996), operate mainly at a more theoretical level. Their concern is more with showing how individually unselfish adaptations are possible in principle, given reasonable but idealized assumptions, than with any specific adaptation or historical scenario. Wilson, an evolutionary biologist, and Sober, a philosopher of biology, argue for functional adaptations on the level of groups, of which the group- benefitting and self-denying behaviour of humans following groupish moralities are presumptive examples. Skyrms shows the advantages of computational modelling using evolutionary rather than classical game theory for producing explanatory theories in the social and behavioural sciences. One class of the adaptive dynamical models Skyrms discusses, those based on correlated evolutionary game theory, model the evolutionary biology used in Part One of Unto Others to show how individually self- denying behaviour can evolve and survive in a Darwinian world.
The consciously Darwinian plots of some Thomas Hardy novels bring out the problem these address almost as starkly as does game theory. In The Trumpet Major and again in The Woodlanders, Hardy presents loyal, conscientious, self-sacrificing heroes whose virtues reward is the loss of love and life. Nature may not craft plots as telling as Hardys, but she is just as unremitting and has much more time. Any heritable trait or strategy that advances the propagation of competitors more than it does its own (net of the cost to itself) incurs a fitness disadvantage relative to otherwise similar ones that benefit from the sacrifices of the first without reciprocating, and will thus inexorably tend to be selected out in favour of these others over time. It seems that nice traits and nice guys finish last. But where, then, could what we take to be our own sometimes altruistic behaviour, and the motives and attitudes that seem to be sometimes behind this and to let us join Hardy in identifying with these heroes, have come from? Or are we just fooling ourselves, the better to fool others, when we see ourselves in this light, as Krebs commentary on Boehm may suggest?
Hardys trumpet major gives up on the woman he loves in his brothers favour. Since brothers tend to share heritable traits (although not all the heros estimable ones in this case), such evolutionary sacrifice is less than it at first appears, and may even be selected for in harsh environments when it is the best one can likely do. So much has long been uncontroversial in evolutionary biology. David Wilsons work has emphasized that what is essential here is not kinship but that the benefits conferred by self-denial are not spread randomly through the population (as early models tended to assume) but rather biased toward those that have the altruistic trait. His group selection approach catches what is general in this: a trait that is locally disadvantageous may evolve by conferring sufficient benefits sufficiently concentrated within trait groups in which it is sufficiently overrepresented with respect to the total (or meta)population for its frequency in the metapopulation to grow or remain stable even as its share in each of the disproportionately productive trait groups declines. The nonrandomness of the interactions relevant to the traits fitness is expressed mathematically by a correlation coefficient, as also in Skyrms correlated evolutionary game theory models of the same phenomena. As Stevens commentary on Unto Others makes clear, such group selection appears to be at work in actual biological cases, even where cognition and preference play no role.
In the human case, peoples spending their time and good deeds with those of their kind is frequently and easily done. For example, altruists may be drawn by their altruism to meet as volunteers at the fire house and rescue squad, and in consequence end up befriending each other and helping each other far more than they help random others. Even random differences in group composition and the effects of geography may yield correlation enough, and kinship also plays its role. So the kind being kind more to their kin and other associates of their kind may explain how selection can make and keep a species somewhat kind, by the kind benefitting from the kindness of others who are kind more than unkind others do. As Sober and Wilson explain, this is made easier in human groups by the culturally-mediated enforcement of social norms, which may be less costly to the individual and hence more able to evolve than the primary practice unenforced would be. The cost of enforcement may be small and such joint institutions easily established especially if the moral institution of social control described by Boehm is already in place. And variation in social norms, helped along by cultural drift and social conformity, would reinforce the phenotypic homogeneity within, and heterogeneity between, groups that group selection requires.
Boehm observes that human culture has thus created a niche in which group selection for genes, as well as for heritable cultural traits, could occur. The evolutionary biologists Kevin Laland, John Odling-Smee and Marc Feldman, however, remain skeptical that genetic group selection has actually been important in nature, although they accept the importance of cultural group selection in explaining human unselfish behaviour. Amotz Zahavi goes further, suggesting that the theoretical possibilities for group selection, which none of these biologists contest, notwithstanding, individual-level selectionist explanations of altruistic behaviour will be more important in every case, as in his examples. Herbert Gintis discusses evidence from computer simulations and from experimental economics on the emergence of equality and fairness and their enforcement, while Alex Rosenberg suggests that enforcement of norms would have been too much against the enforcers own interests to function in simple human societies such as those in which we evolved, and concludes that the question requires a wider range of evidence and modelling results of the kinds discussed by Brian Skyrms and his commentators. Among the results of interest that Skyrms discusses is a bargaining game in which equality and fairness fall out automatically from the adaptive dynamics, whereas concerns for these, observed in everyday human life and in experiments, such as some discussed by Gary Bolton, are paradoxical for theories based on the exclusively selfish rational choice commonly assumed in social science (pp. 2768, below). Skyrms suggests that our evolved ends may differ from those of Darwinian fitness (p. 282, below). And Dennis Krebs seeks to build on correlated evolutionary game theory not a group selection theory, such as Sober and Wilsons, but one of overlapping circles of indirect reciprocity in which peoples individual interests in their reputations, because of the rewards of cooperation dependent on these, keep them somewhat honest but perhaps only as honest as their individually-selected genetically-based adaptations (generalized by the limitations of their cognitive and affective discriminations) make them.
Love, Strife and Community Design
One face of our nature directs a compassionate gaze upon the good of others; a second more particularly regards moral relations between persons and characterizes their attitudes and actions toward each other as right and wrong. While it may nod in moral approval, its stare of righteous anger and the enduring internalized fear this inspires are more forceful in morality's most urgent tasks of deterring and avoiding harm. But without the first concern, for well-being, providing guidance, the moral machinery of social control and self-control would seem to lack any rational point. It is thus fitting that our authors address our capacities for both perspectives and their interactions with strategic self-interest. As Boehm especially reminds us, morality consists in large part in the intentional use of social control for a common good. In keeping with this, Sober and Wilson, after presenting their account of how altruistic behaviour may evolve, proceed in Part II of Unto Others to argue for an evolved plurality within human motivation, including some readiness to take another's good as an ultimate end. They argue that this would evolve because it would be more efficient and reliable in, for example, caring for one's children to have ultimate desires directly for their welfare than to have only desires about one's own pleasant and painful feelings.
The attraction that pleasure and the affections rooted in it have for us may lead us to care ultimately about more distal objects such as another's good as easily as about our own (Schlick, 1939 , II.8, pp. 4751). Empedocles, the first natural selection theorist (Aristotle, Physics II.8:198b2432; Furley, 1987, pp. 948), even identified pleasure with Love, the cause of friendly thoughts and conjoint action, which brings together animals from their parts, people into couples and communities, and all things out of many into one in opposition to Strife, which separates them (Fragments 17 and 20; Furley, 1987, pp. 836). But Nature, as we now know, regards ultimately only fitness and not our happiness (Darwin, 1871, p. 298), and does not scruple to use hate, fear, punishment and even war alongside affection in ordering social groups and selecting among them, just as she uses pain as well as pleasure to get us to feed, water and protect our bodies and also in forging our social bonds.
But evolutionary competition for genetic and cultural propagation need not be violent, as relations between human groups under some conditions often are. Dennis Krebs (p. 320, below) asks how this will play out in our increasingly interconnected world, as does Robert Axelrod elsewhere, discussing suggestive agent-based modelling results about the conditions under which multiple large internally homogeneous groups arise (1997, Ch. 7). Perhaps we may look toward such bottom-up modelling for guidance in designing a world of overlapping and open communities supporting individual freedom and intergroup intercourse rather than between-group polarization. The failed top-down utopian plans of the last century have given social design a bad name. But, as Boehm and Knauft remind us, we have been living in intentional communities since humanity began. Our inherited social nature and needs leave us no choice, and the cognitive capacities that coevolved with these may enable us to continue to learn to navigate together, between the emotional isolation of individual selfishness and total absorption in group chauvinism, neither of which wholly satisfies our evolved nature perhaps toward values that this complex nature enables us to judge are better.
I thank Christopher Boehm, Noam Chomsky and William Harms for their helpful comments, the authors for their instruction, and Daniel Dennett and the students in his Tufts University seminar of Fall 1999 for theirs.
Anderson, S.W., Bechara, A., Damasio, H., Tranel, D. and Damasio, A. R. (1999), Impairment of social and moral behavior related to early damage in human prefrontal cortex, Nature Neuroscience, 2 (11), pp. 10327.
Axelrod, Robert (1997), The Complexity of Cooperation: Agent-Based Models of Competition and Collaboration (Princeton: Princeton University Press).
Black, Donald W. with Larson, C. Lindon (1999), Bad Boys, Bad Men: Confronting Antisocial Personality Disorder (New York and Oxford: Oxford University Press).
Carey, Susan (forthcoming), The Origin of Concepts: Evolution vs. Culture (Cambridge, MA: MIT Press).
Chomsky, Noam (1988), Language and Problems of Knowledge: The Managua Lectures (Cambrdge, MA: MIT Press).
Damasio, Antonio R. (1994), Descartes Error: Emotion, Reason, and the Human Brain (New York: Putnam).
Damasio, A. R., Tranel, D. and Damasio, H. (1990), Individuals with sociopathic behavior caused by frontal damage fail to respond to autonomically social stimuli, Behavioural Brain Research, 41, pp. 8194.
Darwin, Charles (1871), The Descent of Man, and Selection in Relation to Sex (London: Murray [reprinted, Princeton: Princeton University Press, 1981]).
Furley, David (1987), The Greek Cosmologists, Vol. 1, The Formation of the Atomic Theory and Its Eleatic Critics (Cambridge and New York: Cambridge University Press).
Huffman, Carl A. (1993), Philolaus of Croton: Pythagorean and Presocratic (Cambridge: Cambridge University Press).
Lykken, David T. (1995), The Antisocial Personalities (Hillsdale, NJ: Erlbaum).
McKirahan, Richard (1998), Arche, in Routledge Encyclopedia of Philosophy, Vol. 1, ed. Edward Craig (London and New York: Routledge).
Mealy, Linda and Commentators (1995), The sociobiology of sociopathy: An integrated evolutionary model, Behavioral and Brain Sciences, 18, pp. 52399.
Midgley, Mary (1991), The Origins of Ethics, in A Companion to Ethics, ed. Peter Singer (Oxford: Blackwell).
Schlick, Moritz (1939, 1962 ), Problems of Ethics, trans. David Rynin (New York: Prentice-Hall; reprint, New York: Dover Publications, 1962 [Fragen der Ethik (Vienna: J. Springer Verlag)]).
Skyrms, Brian (1996), Evolution of the Social Contract (Cambridge and New York: Cambridge University Press).
Sober, Elliott and Wilson, David Sloan (1998), Unto Others: The Evolution and Psychology of Unselfish Behavior (Cambridge, MA: Harvard University Press).
Wilson, James Q. (1993), The Moral Sense (New York: Free Press).
Online Resources Related to Section 4
Our fourth section presents a selective review of aspects of some relevant
mathematical theories and computational modelling results. Skyrms (1996)
overlaps in coverage with his principal paper for this section. William
Harms online simulators to go with Skyrms book may be found at: <http://eame.ethics.ubc.ca/Skyrms/index.htm>.
Axelrods (1997) Appendix B has exercises and advice for starting out in
agent-based modelling, an approach to modelling evolution different from
the adaptive dynamic modelling discussed by Skyrms, but on which, for example,
Harms own results reported in his commentary are based. This Appendix
has an associated Web page:
<http://pscs.physics.lsa.umich.edu/Software/CC/CCAB.html> with links to online resources.