What Is a Human Universal? Human Behavioral Ecology and Human Nature

What is a Human Universal? Human Behavioral Ecology and Human Nature

Elizabeth Cashdan, University of Utah In Arguing About Human Nature, S. M. Downes and E. Machery, eds. Routledge. 2013

In 1991 the cultural anthropologist Don a function of exposure to sunlight (tanning) is a Brown bucked anthropology’s tabula rasa norm of reaction, as is the percentage of male and tradition by identifying over 300 “human female leopard geckos hatched at different incu- universals”—individual and sociocultural traits bation temperatures (see figure 1). Although that are found in every known human society gecko families vary in the the strength of their (Brown 1991; universals enumerated in Pinker response to temperature, the patterning of the 2003). The items he identified include both psy- response is similar: it is part of leopard gecko chological traits (e.g., wariness around snakes, nature for more males to be born when the tem- sweets preferred, sexual jealousy) and sociocul- perature is warmer, within this range of temper- tural ones (e.g., territoriality, females do more atures. Figure 2 shows another example, from a direct childcare, food sharing). Because they are brilliant study of soapberry bug mating behav- universal, it is plausible that these traits have ior by Carroll & Corneli (1999). A male soap- a biological basis and that they are evolved fea- berry bug may stay attached to his mate after tures of a universal human nature. mating in order to guard her against other suit- What, then, of the many domains where cul- ors. If there are many more males than females tures differ? The assumption is often made that (high sex ratio), this makes adaptive sense: he human nature is found solely in its universals – gains more by keeping other males away than in the traits found in every society. A trait that he loses in forgone mating opportunities. But is found in some societies and not others is then if eager suitors are less numerous (lower sex ra- assumed to be culturally constructed and with- tio), he gains little by mate-guarding and should out an evolutionary foundation. instead leave after mating and search for another mate. The upper figure shows the mate- Behavioral ecologists hold a different view: guarding response of male Oklahoma soapberry because human nature evolved to be flexible in bugs (each line is a family of half-sibs) to exper- predictable ways, the task of understanding hu- imental changes in the sex ratio. Mate-guarding man nature requires that we understand how increases as the sex ratio increases. Although evolution shaped that variation. The assump- each family has a slightly different reaction norm, tion is not just that we evolved to respond flex- they are similar enough to indicate a general ibly, but that selection shaped the nature and feature of what we might call Oklahoma soap- direction of that flexibility. To a behavioral ecol- berry bug nature. An observer will see differ- ogist, then, the predictable, patterned nature of ent behaviors in different environments, but the that response is the universal we must under- responsiveness—the shape of the reaction norm stand. In this view, we cannot understand our curve—is “universal.” At least, it is univer- universal human nature without understanding sal in Oklahoma. The lower figure shows that the variability in its expression. soapberry bug families from Florida do not al- The concept is clarified by viewing variation ter their mate-guarding appreciably, irrespective as a norm of reaction—the pattern of expression of experimental changes in the sex ratio. Why of a genotype across a range of environments. not? In Florida, the climate, hence the sex ra- The increase in a person’s skin pigmentation as

1 Figure 1: Sex ratio reaction norms as a function of incubation temperature in leopard geckos. Each line connects the sex ratio for a half-sib family (oﬀspring sired by a single male) that was divided between two incubation temperatures. (Rhen et al., 2010)

tio, is less variable than in Oklahoma. In Ok- This perspective on human nature derives lahoma, a facultative (plastic, flexible) response from the anthropologists’ knowledge of human makes adaptive sense, because an individual bug variation and the ecological focus of the par- could find himself in a variety of environments ent discipline of behavioral ecology. Most hu- and he will reproduce better if he can respond man behavioral ecologists were trained as cul- to those changes. For the Florida bugs, living tural anthropologists, which gives them an un- in their equable environment, there is no advan- equaled knowledge of the breadth and regulari- tage to such flexibility, and in that population, ties of human variation. They know, better than the flexible response did not evolve (Carroll & anyone, that people living in developed, industri- Corneli, 1999). alized states (the usual subjects of human social science) represent only a very small part of the Soapberry bugs occupy a range of environ- range of human variation, and that those soci- ments, but the range is nothing compared to that eties are in many respects quite unusual. The na- of humans, who live in every part of the planet ture of the parent discipline of behavioral ecology and whose environments include novel ones that also shapes their perspective of human nature. they constructed for themselves. The principle Human behavioral ecology’s modus operandi is illustrated in the previous example, therefore, is to model optimal outcomes by considering the even more important when discussing human na- costs and benefits of different strategies and how ture. To a human behavioral ecologist, then, hu- they trade off against one another. Doing this man nature is not limited to human universals, forces an explicit consideration of the ecological as that phrase is usually understood. The hu- factors that shape those costs and benefits, and man universal is the shape of the response, and how they vary across environments. the task is to understand how selection pressures shaped it. In what follows I will consider examples

2 Figure 2: Mate guarding reaction norms as a function of sex ratio in soapberry bugs in Oklahoma (top) and Florida (bottom). (Carroll & Corneli, 1999)

3 which illustrate the following implications of this havioral ecologists who work with humans also view of human nature: (1) human and non- typically hold the working assumption (not be- human animal behavior can be understood us- lief) that the human behavioral differences they ing the same evolutionary theoretical perspec- observe are the facultative expression of a largely tive, and, in some cases, models (hence human shared genotype. nature is seen as part of the evolved natural Finally, human behavioral ecologists as- world), (2) viewing behavioral variation as a re- sume that human nature is adaptive (fitness- action norm provides guidance on how policy can enhancing), and that selection will lead to op- address the darker side of human nature, and timal outcomes. These optimal outcomes are (3) the individual-maximizing process of natu- modeled as the best outcome possible given re- ral selection has created a remarkably altruis- source constraints and tradeoffs between com- tic, cooperative human nature. I will not, in peting demands. The assumption of optimality this essay, attempt to review the field of human is less a matter of belief about human nature behavioral ecology generally, which has been than it is a useful working assumption. Behav- done admirably elsewhere (Borgerhoff Mulder, ioral ecologists know, as well as anyone, that peo- 2003; Cronk, 1991; Laland & Brown, 2011; Smith ple sometimes do maladaptive things (although et al., 2001; Winterhalder & Smith, 2000). How- they do not expect maladaptive outcomes to be ever, several points should be made first about common), and they also know the reasons why the assumptions under which behavioral ecolo- evolution sometimes leads to sub-optimal out- gists operate. comes. But it is a reasonable and powerful work- Human nature, broadly speaking, encom- ing assumption in generating and testing hy- passes the ways in which people think, feel, and potheses about the functions of, and selection act. However, thoughts and feelings are them- pressures on, human behavior—hence how hu- selves “invisible” to natural selection, since they man nature came to be. This is best demon- can only affect survival or reproduction by mo- strated by example, so we will consider several, tivating behavior. An emotion, no matter how beginning with mating and marriage. strongly felt, is irrelevant to evolution if it does not cause an observable change. For this rea- Mating and Marriage: Biological mod- son, human behavioral ecologists are largely un- els can explain a lot about human cul- concerned with psychological mechanisms, and tural behavior focus instead on the behavioral outcomes that selection can act upon. Marriage is universal, but variable, across human For somewhat different reasons, human be- societies, and our understanding of what forms havioral ecologists also are largely unconcerned are natural or normal have both moral over- with genetic mechanisms. Although biologists tones and policy implications. The ethnographic studying the behavioral ecology of other species database makes it clear that human nature en- are increasingly interested in the genetic basis of compasses marriages that are both monogamous flexible responses, these are difficult to study in and polygynous, and even, under very rare and people, and we know very little about the ge- special circumstances, polyandrous. What be- netics underlying human nature and behavior. havioral ecology adds to this pluralistic view of Human behavioral ecologists therefore typically mating and marriage is the specification that adopt a “phenotypic gambit” that assumes there men and women will adjust their mating and has been sufficient genetic variation and time for marriage choices to environmental circumstances competing selection pressures to have resulted in in predictable ways, and that those choices will the evolution of better-adapted phenotypes. Be- be optimal from the perspective of enhancing re-

4 productive success. wife of a wealthy man she can end up with more What is the optimal number of wives (assum- resources (after division among wives) than she ing a man is legally allowed to have more than could by being the sole wife of a pauper. This one)? It depends. Both time and resources are argument (formalized as the “polygyny thresh- limited and subject to tradeoffs, and these trade- old” model) was initially developed to explain offs apply across species. If a male spends time mating patterns in birds, but was also used suc- guarding one mate, he has less time available to cessfully to explain who married whom among pursue others. If he spends resources (energy or the agro-pastoral Kipsigis. Borgerhoff Mulder money) trying to attract and monopolize addi- (1990) showed that having co-wives imposes re- tional mates, he has less to invest in his current productive costs on Kipsigis women, which they offspring. Like the soapberry bugs, therefore, a try to minimize by judicious marital choices. human male with few alternative mating oppor- The Kipsigis women she studied (or their par- tunities may do better to stay with his mate (not ents acting on their behalf) chose men who had only to keep other suitors away but also, in the the most acres available after division among ex- case of humans, to invest in his offspring and isting wives (consistent with the female-choice enhance their reproductive success). polygyny threhold model), not the most acres This argument has been used to explain dif- overall (the latter would suggest instead that ferences in mating and marriage among two the wealthiest men were able to control the out- groups of South American foragers, the monoga- come). mous Hiwi and the polygynous Ache. Among the These examples illustrate both limitations Hiwi, a comparative shortage of reproductive- and contributions of human behavioral ecology aged women and lower fertility promotes monog- to an understanding of human nature. Because amous pair bonds and men directing more of it follows the phenotypic gambit, human be- their resources to provisioning their family, even havioral ecology is agnostic about the mecha- though that effort has a modest effect on chil- nisms that lead to these patterns, so has little dren’s survival compared to that of Ache men. to say about them. Does the choice of mate Ache men, in contrast, have relatively more op- arise from conscious consideration of the pros portunities for new matings and added paternity, and cons of each option? Or from innate pref- and so they can gain more reproductive success erences, such that fitness-enhancing mates look by searching for new mates, and marriage is less sexier and more attractive? Or (most likely) durable than among the Hiwi (Hurtado and Hill both? For the answer to that question, you will 1992; see also Blurton Jones et al., 2000). Cross- need to ask the evolutionary psychologists – it is national data (Trent & South, 1989) and histori- simply not a focus of human behavioral ecology. cal trends (Guttentag & Secord, 1983; Pedersen, Behavioral ecology does show that (1) marriage 1991) also find that high sex-ratio societies, in choices are consistent with models of adaptive which men more than women must compete for behavior (i.e., behavior that enhances the fitness mates, are associated with more stable marriages of the people making them), (2) the diversity of and lower divorce rates. mating and marriage patterns is, therefore, not Mating and marriage decisions are complex, divorced from nature but rather is a predictable not least because women and men often want dif- manifestation of it, and (3) models from biology, ferent things and the outcome must consider the which were developed for other species, are sur- strategic decisions of both. Where women can prisingly successful in explaining these cultural choose their own mates, mating patterns may patterns. The last point underlies the fact that be driven more by female choice. Polygyny can human nature is one manifestation of animal na- be in a woman’s interests if by being the second ture.

5 Life history and parenting: what the nancy is a linear function of the probability of human behavioral ecology perspective future marriage (data from England and Wales, implies about problematic behavior Lycett & Dunbar 1999). Views of human nature have implications for policy, since behav- It is part of human nature for a mother to love ior that is pathological is addressed by trying to her children, but it is also part of human nature change the individual, while a normal but un- that such love is not unconditional. Humans are desired aspect of human nature (i.e., one that unusual primates in this respect. The primatol- lies within the norm of reaction of most of the ogist Sarah Hrdy (2009) has pointed out that in population) is more profitably addressed (if one most other primates a mother gives birth only wishes to) by changing the circumstances that when her other offspring are independent, and favor it. By viewing behavior such as child ne- unconditional nurture for each new arrival is the glect and even infanticide as a fitness-maximizing norm. This is not true for humans, who differ response to resource constraints and competing from other primates in having a long period of demands, the behavioral ecologist would be hes- childhood dependency together with interbirth itant to label such behavior as pathological, and intervals that are short compared to those of our more likely to suggest that we ameliorate the sit- closest primate relatives. One result of having uation by addressing the constraints and com- to take care of several children simultaneously peting demands that made such choices adap- is that human mothers face allocation decisions tive. She would also be aware that a single most other primate mothers do not, and with- mother who struggles without kin help is in an drawal of investment, including even infanticide, unnatural situation, although a common one in is part of the human condition. Another con- modern industrial societies, and would suggest sequence is that human mothers, unlike other policies that change the circumstances to resem- primate mothers, need and get help with child- ble those to which we are normally adapted. rearing from other relatives (the industrialized While models in human behavioral ecology may world is an exception to this otherwise universal seem cold-blooded, therefore, the policy implica- feature). tions of its approach are likely to be both humane and progressive. Behavioral ecologists have shown that a mother’s reproductive decisions are sensitive to factors that would maximize the fitness returns Generosity on her investment: the condition of the infant, the mother’s social and economic support, and When Dawkins (2006) coined the term “selfish her other options. Because a woman’s reproduc- gene” he was describing the fact that any gene tive options diminish with age, the probability that promotes reproductive success in the body that she will terminate investment in an offspring it finds itself in will help itself to spread in the also declines with age. Figure 3 shows this pat- population. Genes may be selfish in this sense, tern in a population of forager-farmers, the Ay- but they can spread by promoting altruistic as oreo (Bugos & McCarthy, 1984), and modern well as selfish behavior. From Dawkins’s gene- Canada (Daly & Wilson, 1988). The opportu- centric perspective, a gene that is shared by two nity costs are greater and the age effect steeper individuals through common descent will be fa- for single women than for married women, since vored if those individuals behave altruistically men are less likely to marry a woman with chil- toward each other. Kin selection, which is built dren sired by another man, other things equal. on this idea, is the reason many species exhibit This tradeoff is illustrated in figure 4, which behavior that is altruistic (in the technical sense shows that the probability of terminating a preg- of an act that favors the recipient at some cost to

6 Figure 3: Risk of infanticide as a function of mother’s age among (a) the Ayoreo (n=141 births) and (b) Canada (Daly & Wilson, 1988).

7 Figure 4: Age-speciﬁc probability of abortion plotted against the probability of future marriage during the reproductive life span ((ages 16-40 years) of single women. England and Wales 1991. (Lycett & Dunbar, 1999).

the altruist). Altruism toward kin is an evolved Both of these explanations, and others, have part of human nature also, but humans are un- been evaluated by behavioral ecologists, particu- usual in the scope of their unselfish behavior, larly in the context of food sharing. Supportive which is extended far beyond the usual explana- evidence for the role of reciprocity comes from tory reach of kin selection. Understanding how evidence that food sharing among both Ache and selection has favored such behavior is a major Hiwi is contingent on past behavior (people share part of the recent hunan behavioral ecology re- more with those who have shared with them) search agenda. (Gurven, 2006), and that Ache who produced and shared more than average also received more Generosity to non-kin can confer long-term food when they were injured or sick (Gurven benefits (even though at a short-term cost) if the et al., 2000). There is also evidence for sharing as altruist can be sure that a favor given now will be advertisement. Some foragers, especially young reciprocated at some future time when he needs males, target hard-to-get and large game, which the help (reciprocal altruism). Such a system is is shared more widely than other resources. The vulnerable to cheaters, so people are expected to behavior is costly, but pays indirect fitness ben- be careful about who they are generous to, lim- efits, chiefly by enhancing mating opportunities iting their generosity to people who have shared for good hunters, even where the sharing is not with them in the past or who have a reputation contingent (Hawkes & Bliege Bird, 2002; Smith for having been generous with others. Exper- & Bird, 2000; Smith et al., 2003). In the case imental games have shown that decision rules of both reciprocity and showing off, the generos- of conditional generosity can out-compete self- ity is real but ultimately enhances the fitness of ish behavior. A second way in which generosity the donor. Food sharing is one of Don Brown’s to non-kin can confer benefits to the altruists is human universals, but, like most universals, it through advertising the donor’s resources or abil- is exhibited strategically, and we will not under- ities (e.g., the philanthropist who gets his name stand human nature without understanding that on a building).

8 Figure 5: Sharing in the anonymous “dictator game”: Mean oﬀer (as a percent of total stake) by degree of market integration of ﬁfteen traditional societies and the US. (Henrich et al., 2010)

variability and the reasons for it. possible only in groups where norms of fairness Explaining variation in generosity is a major and trust among strangers have spread through focus of research in behavioral ecology. Figure 5 social means. Most of the behavioral ecology shows how results of an economic sharing game research on this topic continues to be agnostic (the “dictator” game) vary with the market inte- about mechanisms, but this is starting to change. gration of the society. In this game, two anony- Behavioral ecologists and evolutionary psycholo- mous players are given a sum of money and one gists are working together to understand the evo- of them (the “dictator”) is given the right to al- lution of the cognitive specializations that makes locate whatever fraction of it he wants to the us such an unusually cooperative species. Co- other (the recipient). The recipient can do noth- operation on a larger scale, involving such phe- ing to punish the dictator for meager offers, yet nomena as ethnocentrism, fairness to strangers in no society do people, on average, fail to give in market societies, and large-scale warfare, are something (the reason for this persistent but un- likely to require the understanding of new mech- economical behavior is still being debated). In- anisms involving cultural as well as genetic trans- terestingly, the two groups with the lowest (least mission, and this is another new and growing generous) offers are Hadza foragers and Tsimane area that is expanding the field of human be- forager-farmers, both of whom depend heavily havioral ecology. It is also an area of active on sharing in their daily lives and are far more collaboration between human behavioral ecolo- dependent on it than people in the U.S., who an- gists and other social scientists, especially ex- chor the high end by offering nearly a 50-50 split. perimental economists. Human behavioral ecol- Group size is also a strong predictor of generosity ogists sometimes grumble that these sister dis- in this game, with large groups being more gener- ciplines are largely “ecology free” and give in- ous. Henrich et al. (2010) reach beyond the usual sufficient attention to the real range of human evolutionary mechanisms and models of behav- variation, and that evolutionary psychologists in ioral ecology to explain this counter-intuitive re- particular give inadequate attention to trade-offs sult, arguing that large-scale market exchange is among competing aims, but the trend is clearly

9 toward an integration of the human evolution- Dawkins, R. (2006). The selfish gene. Oxford ary sciences. Substantive differences in the disci- University Press. plinary views of human nature stem from histori- cal differences in the populations studied and the Gurven, M. (2006). The evolution of contingent models and methods employed, and are likely to cooperation. Current Anthropology, 47(1), be short-lived as these complementary strengths 185. are pooled through collaboration. Gurven, M., Allen-Arave, W., Hill, K., & Hur- tado, M. (2000). it’s a wonderful life signaling References generosity among the ache of paraguay. Evo- lution and Human Behavior, 21(4), 263–282. Blurton Jones, N., Marlowe, F., Hawkes, K., & Guttentag, M. & Secord, P. (1983). Too many OConnell, J. (2000). Paternal investment and women?: The sex ratio question. Sage Publi- hunter-gatherer divorce rates. In L. Cronk, N. cations. Chagnon, & W. Irons (Eds.), Adaptation and human behavior: An anthropological perspec- Hawkes, K. & Bliege Bird, R. (2002). Showing tive (pp. 69–90). Aldine de Gruyter. off, handicap signaling, and the evolution of men’s work. Evolutionary Anthropology: Is- Borgerhoff Mulder, M. (1990). Kipsigis women’s sues, News, and Reviews, 11(2), 58–67. preferences for wealthy men: evidence for female choice in mammals? Behavioral Ecology Henrich, J., Ensminger, J., McElreath, R., Barr, and Sociobiology, 27(4), 255–264. A., Barrett, C., Bolyanatz, A., Cardenas, J., Gurven, M., Gwako, E., Henrich, N., et al. Borgerhoff Mulder, M. (2003). Human behav- (2010). Markets, religion, community size, and ioral ecology. In Encyclopedia of Life Sciences. the evolution of fairness and punishment. Sci- Nature publishing group. ence, 327(5972), 1480.

Brown, D. (1991). Human universals. Temple Hrdy, S. (2009). Mothers and others: The evolu- University Press. tionary origins of mutual understanding. Belk- nap Press. Bugos, P. & McCarthy, L. (1984). Ayoreo infanticide: A case study. Infanticide: Comparative Hurtado, A. & Hill, K. (1992). Paternal eﬀect and evolutionary perspectives, (pp. 503–520). on oﬀspring survivorship among ache and hiwi hunter-gatherers: Implications for modeling Carroll, S. & Corneli, P. (1999). The evolution pair-bond stability. In B. S. Hewlett (Ed.), of behavioral norms of reaction as a problem Father-child relations: Cultural and biosocial in ecological genetics: Theory, methods, and contexts (pp. 31–55). Aldine de Gruyter. data. In S. A. Foster & J. A. Endler (Eds.), Geographic variation in behavior: perspectives Laland, K. & Brown, G. (2011). Sense and non- on evolutionary mechanisms (pp. 52–68). New sense: Evolutionary perspectives on human be- York: Oxford University Press. haviour. Oxford Univ Pr.

Cronk, L. (1991). Human behavioral ecology. Lycett, J. & Dunbar, R. (1999). Abortion rates Annual Review of Anthropology, 20, 25–53. reﬂect the optimization of parental investment strategies. Proceedings of the Royal Soci- Daly, M. & Wilson, M. (1988). Homicide. Al- ety of London. Series B: Biological Sciences, dine. 266(1436), 2355–2358.

10 Pedersen, F. (1991). Secular trends in human ﬁts of costly signaling: Meriam turtle hunters. sex ratios. Human Nature, 2(3), 271–291. Behavioral Ecology, 14(1), 116.

Pinker, S. (2003). The blank slate: The modern Smith, E., Mulder, M., & Hill, K. (2001). Con- denial of human nature. Penguin books. troversies in the evolutionary social sciences: A guide for the perplexed. Trends in Ecology Rhen, T., Schroeder, A., Sakata, J., Huang, V., & Evolution, 16(3), 128–135. & Crews, D. (2010). Segregating variation for temperature-dependent sex determination in a Trent, K. & South, S. (1989). Structural deter- lizard. Heredity, 106(4), 649–660. minants of the divorce rate: A cross-societal analysis. Journal of Marriage and the Family, Smith, E. & Bird, R. (2000). Turtle hunting (pp. 391–404). and tombstone opening:: public generosity as costly signaling. Evolution and Human Behav- Winterhalder, B. & Smith, E. (2000). Analyzing ior, 21(4), 245–261. adaptive strategies: Human behavioral ecology at twenty-ﬁve. Evolutionary Anthropology: Smith, E., Bird, R., & Bird, D. (2003). The bene- Issues News and Reviews, 9(2), 51–72.