Virtue Signaling
Page 5
The virtue ethics perspective from philosophy is admittedly not doing much explanatory work in this paper. I do not assume that the ‘virtues’ historically identified by philosophers will equal the moral adaptations that can be identified in humans using standard adaptationist criteria of special design. Nor do I assume that the idealistic reasons for advocating certain virtues in normative ethics will have anything to do with the nitty-gritty selection pressures that may have actually shaped those virtues phylogenetically.
So why mention virtue ethics at all? Four reasons. First, virtue ethics provides a useful counter-balance to the traditional consequentialist (utilitarian, payoff-based) ethics have historically influenced previous evolutionary theories of altruism (such as kin selection and reciprocal altruism). Second, as I will argue in a later section, virtue ethics shifts the level of analysis usefully from isolated altruistic acts to stable personality traits. Third, many virtue ethicists write carefully and insightfully about our emotional and cognitive responses to other people’s virtues and vices, and their work can be construed as a useful first draft of the qualitative, descriptive moral psychology that may prove useful in understanding the ‘receiver psychology’ of moral signaling. Fourth, virtue ethics offers a new route whereby evolutionary theory can influence the contemporary humanities and social sciences – we can ask not what virtue ethics can do for us, but what we can do for virtue ethics. Thus, my allusions to virtue ethics are intended in the spirit of maximizing the interdisciplinary relevance of adaptationist research.
Costly Signaling Theory, Fitness Indicators, and Moral Virtues
Costly signaling theory has intellectual roots in many traditions and academic fields, some of which construed human moral virtues as costly signals. In Friedrich Nietzsche’s The Genealogy of Morals in 1887, pagan virtues were considered attractive signals of health and power. In Thorstein Veblen’s The Theory of the Leisure Class in 1899, conspicuous consumption and conspicuous charity were seen as hard-to-fake signals of wealth and social status. In a seminal 1975 paper, biologist Amotz Zahavi viewed many animal signals and pro-social behaviors as hard-to-fake indicators of animal fitness.
Since about 1990, costly signalling theory has revolutionized the study of both sexual selection and human altruism. Most animal communication is relentlessly narcissistic, advertising the signaler’s own individual species, sex, age, health, fertility, social status, phenotypic condition, and/or genetic quality. However, animals often have incentives to lie about their own qualities, to attract more mates, solicit more parental investment, or deter more predators and rivals. Costly signaling theory offers a solution to this problem of lying: if a signal is so costly that only high-health, high-status, high-condition animals can afford to produce it, the signal can remain evolutionarily reliable.
Almost any fitness-related cost will work: matter, energy, time, or risk. For example, a peacock’s tail is burdensome in all four senses: its growth and maintenance require several hundred grams of mass, many calories, much time to grow, and much risk (its costs undermine immunocompetence and parasite-resistance). Often, the most complex, elaborate, and puzzling signals observed in nature are the result of sexual selection through mate choice (as Darwin argued back in 1871). These sexual ornaments almost always impose high costs on the bearer, guaranteeing their reliability as indicators of condition and fitness.
This paper argues that many human virtues evolved through sexual selection as costly signals. This hypothesis has been advanced by a few previous researchers (such as Darwin and Zahavi), and its empirical testing has been one of the most active areas of evolutionary psychology and evolutionary anthropology in the last few years. Indeed, many pro-social behaviors that were assumed to arise through kinship or reciprocity are now thought to have emerged as costly signals of individual fitness, favored by social and sexual selection.
For example, it was often assumed that risky big-game hunting evolved because the best hunters could better feed their own offspring. However, most hunted meat from big game is distributed too widely in hunter-gatherer clans for this paternal provisioning theory to work. Rather, recent research suggests that the most successful hunters who provide the pro-social ‘public good’ of hunted meat also tend to attract more high-quality female mates. Meat-provisioning may not be a conscious sexual strategy, and may not even be the causal mediator of good hunters’ increased reproductive success (good hunting and high attractiveness may both be caused by an underlying trait such as high genetic quality). Nevertheless, such research raises the possibility that altruistic meat-provisioning was favored, at least in part, by sexual selection. Likewise, our mate preferences for other moral virtues may be explained by costly signaling theory. If a young woman places a single’s ad stating “SHF, 26, seeks kind, generous, romantic, honest man”, we might translate this in evolutionary terms as “single Hispanic female, 26, seeks a healthy male of breeding age with a minimal number of personality disorders that would impair efficient coordination and parenting in a sustained sexual relationship, and a minimal number of deleterious mutations on the thousands of genes that influence the development of brain systems for costly, conspicuous, altruistic displays of moral virtue.” Of course, this hypothetical single’s ad itself is not good psychological evidence or a costly signal in its own right – it is cheap and easy to fake. Rather, the ad identifies some desired moral virtues that would be hard to fake consistently during a lengthy courtship.
Good Genes, Good Parents, and Good Partners
Sexually-selected costly signals typically advertise two classes of traits: good genes or good parenting abilities. Different moral virtues might advertise one or the other, or both. They might also advertise the capacity to be a good partner in a long-term sexual relationship – someone reliable, trustworthy, adaptable, agreeable, and efficient at coordinating joint activities.
Good genes indicators advertise general ‘genetic quality,’ which probably reflects having a low ‘mutation load’ – fewer than average errors in DNA replication. By favoring mates with a lower-than-average number of harmful mutations, sexually reproducing organisms can increase the expected survival and reproductive prospects of their offspring – even if their mate contributes nothing as a parent after fertilization.
Moral virtues may function as good genes indicators by being difficult to display impressively if one has a high mutation load that impairs the precision of body and brain development. For example, displaying a sophisticated, empathetic social intelligence requires the development of a complex ‘Theory of Mind,’ which might be easily disrupted by a variety of mutations associated with autism, schizophrenia, mental retardation, social anxiety, and language impairments. Thus, a conspicuously expert level of empathy may function as a sort of neurogenetic guarantee. For moral virtues to function as good genes indicators, they must show at least moderate degrees of genetic variance, heritability, and positive genetic correlations with other fitness-related traits.
By contrast, good parent indicators advertise phenotypic traits that help care for offspring, such as feeding them, grooming them to remove parasites, protecting them from predators, resolving sibling rivalries, and teaching life-skills through play and practice. So, a genuinely empathic personality may also function as a good-parent guarantee, testifying to the likely patience, kindness, protectiveness, playfulness, and conscientiousness that helps children thrive. For moral virtues to function as good parent indicators, they need not show genetic variance, heritability, or genetic correlations with other fitness-related traits; they need only show reliable phenotypic correlations with parenting-relevant abilities.
Finally, good partner indicators advertise phenotypic traits that promote efficient coordination and high mutual benefits in long-term sexual relationships. In game theory terms, such relationships (such as marriages) are iterated, mixed-motive games with very complex conflicts and confluences of interest, many possible equilibria, and incomplete information about the other player’s pos
sible tactics and preferences.
Some of a potential mate’s moral virtues could function as signals that maximize one’s payoffs and minimize one’s risks in such relationship games. For example, moral capacities for conscientiousness and patience may signal a partner’s likelihood of playing mutually beneficial strategies given the iterated (repeated-interaction) nature of long-term relationships. Moral capacities for empathy and sympathy may signal that a partner attaches positive utility to one’s own happiness in addition to their own, which makes it much more likely that a Pareto-optimal (mutually beneficial) equilibrium will be maintained in the relationship. A moral preference for romantic commitment over violent aggression may signal that a partner will seek to sustain a cooperative relationship through promises rather than threats. In each case, the moral trait as a good partner indicator may seem intuitively attractive for its own sake, since conscientious, empathic, committed partners just make life easier. However, as noted earlier, whenever there are incentives to act like a better partner during courtship than after reproduction, the problem of trait stability arises.
Mate preferences for human moral virtues, whether as good genes, good parent, or good partner indicators, may have originated in other social preferences concerning kin selection, reciprocal altruism, social commitment, progeny choice, or other domains. For example, psychological adaptations for detecting, remembering, and avoiding cheaters in the domain of social reciprocity could have been extended (‘exapted’) to the domain of mate choice as a preference for the moral virtues of honesty and reliability. Indeed, mate preferences for moral virtues may have originated as non-functional by-products of these other social preferences, and would have been more appropriately analyzed using ‘receiver bias’ models rather than costly signaling models of sexual selection. However, given the central adaptive importance of mate choice, it seems likely that such non-functional preferences would have been rapidly eliminated (if they had net fitness costs), or modified and specialized for adaptive mate choice (if they had net fitness benefits).
Sexually Selected Signals and Sex Differences
From a costly signaling perspective, sexual selection is not restricted to explaining sex differences; it can also explain sexual similarities (non-dimorphism) in extravagant traits when mutual mate choice is at work. Humans are unusual among mammals in showing intensive offspring care by both mothers and fathers, which favors roughly equal levels of male and female mate choice for long-term socially monogamous relationships. This has resulted in low-dimorphism physical ornamentation in humans such as long head hair, relatively hairless bodies, and everted lips, and low-dimorphism psychological ornamentation such as cognitive abilities for language, art, music, humor, and ideology.
Thus, a sexual selection account of moral virtues does not imply that males evolve all the conspicuous virtues and females play the passive role of virtue-assessment. Given mutual choice, both human sexes should show conspicuous, sexually-attractive moral virtues during mate attraction and retention.
However, human males face higher variance in reproductive success, so are predicted to allocate somewhat more energy, time, and risk to mating effort, including costly, dangerous, public displays of moral virtue. For example, this model naturally explains why males are over-represented among pro-social heroes who risk their lives to save unrelated strangers. It may also explain why males remain over-represented in high-risk, under-paid, altruistic, romantically attractive professions such as the police, fire, and military services (whereas women remain over-represented in low-physical-risk, under-paid, altruistic, romantically attractive professions such as nursing and school teaching). The biased sex ratios in these high-risk versus low-risk professions are undoubtedly influenced by social norms regarding gender, danger, and sex roles. Nonetheless, human gender norms seem unlikely to be the whole explanation, given the cross-species ubiquity of sex differences in risk-taking, and the evidence that human sex differences in risk-taking are largely mediated by sex differences in personality traits such as sensation-seeking, aggressiveness, and extraversion, rather than gender roles concerning risk per se.
More generally, sexual selection is probably relevant somehow whenever there are cross-culturally stable sex differences in the display or judgment of human moral virtues. By contrast, most traditional theories of human moral evolution through kinship, reciprocity, group selection, and equilibrium selection are sex-blind. Therefore, they have trouble accounting for any observed sex differences in the means, variances, heritabilities, genetic correlations, and life-history profiles of specific quantifiable moral virtues and any associated social, political, or religious attitudes.
Evaluating Moral Persons Versus Moral Acts
Costly signaling theory portrays human moral actions in a new light, as reliable cues of personal moral traits. This may seem a peculiar idea to most moral philosophers, who have traditionally focused on judging the morality of isolated acts rather than the moral virtues of whole people. Recently, as act ethics has been supplanted by virtue ethics, attention has returned to the moral-person level of description – just the right level of description to consider in costly signaling models of moral evolution. It is the level that unifies the study of quantitative traits in evolutionary genetics, mate choice in evolutionary psychology, person perception in social psychology, personality traits in behavior genetics, parole decisions in criminal justice, and voter choice in democratic elections.
It seems unlikely that our prehistoric ancestors made moral judgments about isolated behavioral acts. Rather, as in other domains of person perception, they probably interpreted the behavioral acts as cues of stable individual traits (virtues or vices). In small-scale hunter-gatherer bands, morality came in person-sized units, not act-sized units. Ancestral hominids had to choose their lovers, friends, and allies as integrated moral packages – good or bad, hero or villain, lover or stalker, reciprocator or cheat.
Moreover, it would not make adaptive sense to judge isolated moral acts in tightly-knit prehistoric social groups. Individual actions must be assessed in the context of individual qualities, such as age, sex, status, physical health, mental health, personality, intelligence, and genetic relatedness. We tolerate theft by our own toddlers more than theft by unrelated adults. We forgive unkind words spoken during high fevers by the sick. We do not expect a keenly empathic Theory of Mind in the severely brain-damaged or autistic. Also, individual actions must be assessed in the context of ongoing relationships, as different social-interaction domains call for different moral-judgment criteria, focused on different virtues. We may favor kindness and fidelity more in mate choice, honesty and conscientiousness more in reciprocity, and genetic similarity, residual reproductive value, and gratitude more in kin-directed altruism. This is why mothers can love psychopathic sons. Only in social reciprocity with unrelated acquaintances do we see the Tit-for-Tat moral accounting that corresponds to traditional act ethics.
The moral-person and moral-act levels of description show some other key differences relevant to this mate choice model. First, ‘morality’ means something different at the person-level compared to the act-level. A moral act may be one that obeys some rationally defensible, universalizable, deontic or consequentialist principle. However, a moral person, from the point of view of a standard prehistoric hunter-gatherer, is someone who embodies pro-social virtues that make them a good mate, friend, relative, or trading partner. In evolutionary terms, a moral person is simply one who pursues their ultimate genetic self-interest through psychological adaptations that embody a genuine, proximate concern for others.
Second, the moral-person level emphasizes that perceived moral virtue is an emergent property of interaction between the moral judgment-maker and the morally judged – just as beauty arises from the sexual ornaments of the displayer interacting with the perceptual adaptations of the beholder. Beauty is neither ‘subjective’ nor ‘objective,’ but what one could call ‘objectively relational’ – it is a real emergent p
roperty of a costly signaling system including both ornaments and preferences. Likewise for moral virtues in this mate choice model – the morality emerges from the interaction of traits and preferences. By contrast, the moral-act level of description tends to downplay the role of the observer in making moral judgments, pretending that there can be direct contact between a moral act to be judged and a universal normative principle such as Kant’s categorical imperative or Mill’s utilitarian principle.
Finally, we generally accept that ‘ought implies can’ when we judge moral acts. We don’t expect the poor to donate to charity, or quadriplegics to jump in front of trolleys to save children. They can’t do the moral thing, so we don’t expect them to. However, when judging the morality of whole persons in real relationships, we are rarely so forgiving. If a potential mate has Tourette’s syndrome and can’t refrain from screaming ‘psycho prick!’ repeatedly during a public first date, there is unlikely to be a second date, no matter how much we understand about verbal disinhibition in neurological disorders. If a potential hunting partner had a severe head injury that renders him too clumsy to hunt effectively, we may pity him, but will still resentfully exclude him from the hunt.
When the fitness stakes are high, we hold people morally accountable even for faults that are not their own. Moral culpability is a slippery idea, since everyone must be a joint product of their genes, their environment, and random developmental events. If our ancestors couldn’t ostracize helplessly evil people, they couldn’t have protected themselves from serial-rapist psychopaths or hot-tempered murderers.