sign in sign up
<<
Home , Imprinting (psychology), Inbreeding depression, Westermarck effect

Princeton/Stanford Working Papers in Classics

Evolutionary and the historian

Version 1.0

April 2013

Walter Scheidel

Stanford University

Abstract: New possibilities have been opened up for historians by a new wave of engagement with biology, or more particularly with human biology, for the study of human history, environmental history, health history, and the co-evolutionary history of humans and other species. This paper critically explores the uses and limits of evolutionary psychology for the study of history by focusing on the particularly intensely discussed phenomenon of avoidance.

© Walter Scheidel. [email protected]

“He who is the son says this to his mother: ‘Let us give (ourselves) in bodily intercourse so that we shall not have fear of Hell, and the sins we have committed will go from the account, and at the Cinwad Bridge [of posthumous judgment] we will be (pure in) heart, and a beautiful and seemly place will be ours, and we shall please Ohrmazd and cause harm to Ahriman.’ The mother says this to (her) son if she speaks righteously: ‘I give (myself) to you (in) bodily intercourse,’ even as that son said. The father says this to (his) daughter, the brother says this to his sister, (so) the sister says this to her brother.” 1 In the former Zoroastrian tradition represented in this text from around 900 CE, sexual relations between members of the nuclear were not only to be encouraged but meant to result in progeny: “And there are those children who are born of a father and his daughter. Light flashes and glows and can be seen in course of time; and very fortunate and pleased is someone who has a child of his child (…). It is fitting that such sweetness and joy result from the son of a man whom he begets from his own daughter, a child who is a brother of the same mother. And he who is born of a son and his mother is also the brother of his own father: this is a way of much joy and praise, and the harm never exceeds the advantage, and neither do flaws exceed the beauty.” 2 At first sight, the global historical record appears to furnish us with actual instances of comparable behavior, most notably in ruling from ancient Egypt and to Inca Peru and more recently Laos, Hawai’i, and central Africa. In Egypt under Roman rule, even commoners are reported to have married and procreated with their full siblings. Yet, as we shall see below, reproductive within the nuclear family are rarely unequivocally documented, and when placed in context even the breathless exhortations quoted above betray uneasiness about this conduct. 3 Why is that so? Notwithstanding notable differences in perspective, Émile Durkheim, , Bronislaw Malinowski, Talcott Parsons and Claude Lévi-Strauss all concurred in interpreting constraints of incestuous behavior as the result of a powerful “” that had emerged through cultural . The taboo’s perceived purpose was to prevent socially disruptive sexual relations among very close kin which, thought to be latently desired, were otherwise expected to occur. 4 This model presents historians with potentially serious challenges: if the avoidance of and/or procreation by parents and children or full siblings arose from a cultural construct, global adherence to this principle would seem hard to reconcile with otherwise protean contingencies of cultural diversity and change over time. Conversely, observable exceptions to this norm ought to have triggered problems that the taboo was supposed to forestall, such as destabilization of the family unit or weakening of group solidarity. And in so far as sexual attractions existed within the nuclear family, socially sanctioned release from the taboo – as provided by Zoroastrian doctrine – might be expected to have become a popular measure. All these logical implications of a purely cultural model of “incest” avoidance are problematic to varying degrees. 5 Evolutionary Psychology (EP) offers a complementary way of accounting for universal constraints on reproduction within the nuclear family by positing a biological contribution to actual avoidance behavior. This approach is complementary rather than alternative in that it is fully compatible with the notion of cultural specific prohibitions (or the lack thereof, as in the Zoroastrian record): instead, it seeks to provide a more complete understanding of historically observed behavioral propensities. Among historians, engagement with the expanding field of EP has ranged from minimal to non- existent. Even those who tend to be indifferent to biology might be prepared to make an exception for the

This paper was prepared for a forum on Biology and History. 1 The Pahlavi Rivayat Accompanying the Dadestan i Denig 8g1 (late ninth/early tenth century CE), translated by Alan V. Williams, The Pahlavi Rivayat Accompanying the Dadestan i Denig, Part II: Translation, Commentary and Pahlavi Text (Det Kongelige Danske Videnskabernes Selskab 1990), vol.2, 14. 2 Denkard 3.80 (early tenth century CE), adapted from the French translation by J. de Menasce, Le troisième livre du Denkart (C. Klincksieck 1973), 87-88. 3 For discussion and references, see below, nn.41-50. 4 Jonathan Turner and Alexandra Maryanski, Incest: Origins of the Taboo (Paradigm Publishers 2005), 14-52 provide a convenient survey of their and other theories. 5 The final implication conflicts with the Zoroastrian record discussed below (at nn.44-46).

2 notion that human behavior is influenced by inherited traits – by treating it with outright suspicion (or worse). Regardless of whether or not this attitude is justified, this particular approach should at the very least not be dismissed for the wrong , based on generic polemics and reflexive misapprehensions about its agenda.6 Reduced to essentials, EP draws an analogy between the evolution of the brain and the rest of the body. Just like all organisms, the human body has been shaped by selective pressures. In response to these pressures it has developed thousands of , traits that were conducive to its survival and reproductive opportunities in the context of the physical and social environment which generated these pressures. 7 That is why we see, walk upright, and talk. There is no good why our brain, which accounts for 2 percent of our body mass but absorbs 20 percent of its energy, should have been exempt from this adaptive process: it is a product of selection just as any other body part. EP elaborates on these mainstream positions by hypothesizing that selection molded the brain in a way that created innate 8 propensities for adaptive behavior, defined as behavior that promotes transgenerational genetic survival. Because the evolution of such traits take time, any such adaptations would have emerged in response to conditions in what EP terms the “Environment of Evolutionary Adaptedness” (EEA), which is not to be thought of as a single concrete physical environment but rather as “the statistical composite of the enduring selection pressures of cause-and-effect relationships that pushed the underlying an systematically upward in frequency until they became species-typical or reached a frequency-dependent equilibrium.” 9 In this respect, such traits would not be different from other adaptations in the human body, which likewise evolved under changing environmental constraints. This theory raises two simple questions: is it true that such inherited psychological adaptations exist, and if they do, how can they be identified? In a narrow sense, the only real counterfactual to the first assumption would be a “blank-slate” model of the brain that has long been discredited. What is open to debate is merely the degree of specificity of such behavioral traits: proponents of EP hypothesize a comparatively high degree, postulating the evolution of universal behavioral propensities that are directly germane to life-and-death matters such as mating behavior, parenting, and group living. Here it is important to note that even if such traits could be shown to exist, they would in no way result in “genetic determinism” in any strict sense of the term: inherited psychological adaptations would constrict human behavior no more than the anatomy of the human hand confines its use to tasks which sustained its evolution. Such traits can only be identified by testing hypotheses broadly and cross-culturally in order to control for the effects of environmental (including cultural) influence. Due to most EP researchers’ background in psychology and anthropology, existing work privileges interpretation of observed behavior over investigation of the underlying physiological mechanisms and processes down to the biomolecular level, even though the latter is in principle part of the overall program. EP-inspired scholarship has generated a large number of case studies designed to provide empirical support for and theoretical fine- tuning of its premises, which cannot be surveyed here even in the most perfunctory manner. 10 Suffice it to

6 For the ferocity of the earlier debate about “sociobiology”, which was both broader (by including behavior) and narrower (by focusing on social behavior) in scope than EP, see Ullica Segerstrale, Defenders of the Truth: The Sociobiology Debate (Oxford University Press 2000). More recent criticism of EP is addressed in the careful rebuttal by Edward H. Hagen, “Controversial issues in Evolutionary Psychology,” in David M. Buss (ed.), The Handbook of Evolutionary Psychology (Wiley 2005), 145-73. For an even-handed critique, see David J. Buller, Adapting Minds: Evolutionary Psychology and the Persistent Quest for Human Nature (MIT Press 2005). 7 Which is definitely not the same as saying that every feature is adaptive: cf. Stephen J. Gould, The Structure of Evolutionary Theory (Belknap Press 2002). 8 “Innate” propensities would be -based but expressed through interaction with exposure to environmental conditions, that is, epigenetic processes (see the paper by ###) and . The latter is a key feature of the case study presented here. 9 John Tooby and Leda Cosmides, “Conceptual foundations of evolutionary psychology,” in Buss, Handbook 22. 10 The best surveys of the state of the field are now Buss, Handbook and Robin I. Dunbar and L. Barrett (eds.), The Oxford Handbook of Evolutionary Psychology (Oxford University Press 2007). Laura Betzig (ed.), Human Nature: A Critical Reader (Oxford University Press 1997) gathers some case studies. The flagship journal is Evolution and

3 note that published work has gradually moved from sometimes impressionist and speculative arguments to formal scientific experimental designs and sophisticated statistical analysis, bringing much-needed rigor to the process of hypothesis-testing. (This rising expectation of scientific rigor has had the side effect of dampening interest in historical applications of EP, which were occasionally attempted mostly in the 1990s but have since abated. 11 ) Ideally, EP hopes to document innate behavioral adaptations that can meaningfully be related to strong selective pressures, that manifest cross-culturally in observable behavior that cannot be equally well explained in cultural-constructivist terms, and that can be linked to specific physiological processes. In the real world, such clear-cut findings may be hard to obtain. In order to illustrate both the potential and the limitations of this approach, I focus briefly on a phenomenon that has long been considered a particularly promising case, what is often (if sloppily) referred to as incest avoidance. I choose this case for several reasons: the plausible existence of a powerful impetus for selection; a rich tradition of intense debate that draws on experimental as well as historical data for both human and animal populations; recent advances in relevant biological research; and connections to historical scholarship, as noted above. “Incest” is a culturally specific concept, covering relations of varying genetic proximity and beyond.12 It merely overlaps with “,” which is a biological phenomenon and technically refers to the likelihood of inheriting identical copies of from each parent. Inbreeding is harmful in as much as it causes deleterious recessive genes (which are not expressed in the ) to become homozygous (i.e., to be paired) in the offspring of closely related parents, who are much more likely to carry identical copies of such mutant alleles than unrelated parents. Homozygosity of deleterious genes is responsible for genetic disorders that can lead to death or severe congenital and developmental disabilities. The statistical probability of this outcome is a direct function of the degree of parental genetic relatedness. 13 The scale of the health costs of human inbreeding (technically known as “”) has been measured with extraordinary care. 14 Negative consequences of reproduction among relatives outside the nuclear family are relatively minor, especially when viewed in the context of historically high attrition rates from other causes: the latest meta-survey documents a 3.7 percent increase in total pre-reproductive mortality for the offspring of first cousins, which is dwarfed by overall rates of

Human Behavior (founded in 1980 as and Sociobiology ); see also Human Nature (1990-) and Evolutionary Psychology (2003-). 11 Examples include Laura Betzig, Despotism and Differential Reproduction: A Darwinian View of History (Hawthorne 1986) and her numerous articles; J. Kroll and B. S. Bachrach, “Medieval dynastic decisions: evolutionary biology and historical explanation,” Journal of Interdisciplinary History 21 (1990), 1-28; S. B. Johnson and R. C. Johnson, “Support and conflict of kinsmen in Norse earldoms, Icelandic families, and the English royalty,” Ethology and Sociobiology 12 (1991), 211-20, with debate in the 1993 and 1995 issues; E. Hill, “Lineage interests and nonreproductive strategies: an evolutionary approach to medieval religious women,” Human Nature 10 (1999), 109-134; Jörg Wettlaufer, “Von der Gruppe zum Individuum: Probleme und Perspektiven einer ‘evolutionären Geschichtswissenschaft’,” in S. Selzer and U.-C. Ewert (eds.), Menschenbilder – Menschenbildner (Berlin 2002), 25-52; Walter Scheidel, “Sex and empire: a Darwinian perspective,” in Ian Morris and Walter Scheidel (eds.), The Dynamics of Ancient Empires (Oxford University Press 2009), 255-324. The eminent medieval historian David Herlihy showed interest in an evolutionary perspective in his “Biology and history: the triumph of monogamy,” Journal of Interdisciplinary History 25 (1995), 571-83. 12 E.g., P. Bonté (ed.), Epouser au plus proche: inceste, prohibitions et stratégies matrimoniales autour de la Méditerranée (Paris 1994); Philippe Moreau, Incestus et prohibitae nuptiae: conception romaine de l’inceste et histoire des prohibitions matrimoniales pour cause de parenté dans la Rome antique (Les Belles Lettres 2002); Geert Jan van Gelder, Close Relationships: Incest and Inbreeding in Classical Arabic Literature (I. B. Tauris 2005); Elizabeth Archibald, Incest and the Medieval Imagination (Clarendon Press 2001). 13 First cousins share 12.5 percent of their genes through common descent, which means their progeny will inherit identical gene copies from both of them for 6.25 percent of all gene loci. For full siblings or parents and children, the rates are 50 and 25 percent, respectively. Accumulation over generations may produce significantly higher averages. 14 Results drawn from millions of people all over the world have now been synthesized by Alan H. Bittles, in Context (Cambridge University Press 2012).

4 pre-reproductive death of 30 to 50 percent in historical high-mortality populations. Moreover, marital fertility is often somewhat higher in . 15 This means that we cannot postulate significant selective pressures against consanguineous unions at this level of genetic proximity, an intuition that is fully borne out by the global empirical record. About half of the societies surveyed in the Ethnographic Atlas permitted or even favored cousin marriage. About one-sixth of the current world population lives in countries where between 20 and more than 50 percent of all couples are related as the level of second cousin or closer, and one in ten persons on earth is part of such as union. The incidence of consanguineous marriage varies massively both geographically (with heavy concentrations in the Middle East and South India) and over time, as most societies have witnessed declines as a corollary of urbanization and development. 16 Its popularity is highly context- specific, as it may appeal to rural and uneducated groups (by reducing the need for dowries) as well as propertied elites (whom they help maintain privileged status). All of this demonstrates the complete lack of universal constraints that might be associated with inherited behavioral adaptation. Reproductive relations within the nuclear family – between parents and children and between full siblings – differ dramatically in genetic terms. Predicted harm is four times that for first-cousin unions, and while the empirical evidence is inevitably sparse and fraught with potential biases (given the handicaps of many incestuous parents) it consistently reveals serious consequences: a deliberately conservative assessment puts the odds of premature death at about 15 percent and envisions a total excess rate for death and major disability of around 30 percent.17 Inbreeding depression on this scale would have mattered even in high-mortality regimes, raising overall attrition to 50 to 85 percent and necessitating extremely high rates of replacement fertility even before the possible effects of increased child mortality and surviving offspring’s disabilities on parental are taken into account. The costs of nuclear- family incest are massive by any standard. From a PE perspective, these costs provide a powerful incentive for the evolution of a behavioral adaptation to prevent mating at this level of genetic proximity, not just in humans but also in other species that are vulnerable to severe inbreeding depression. Animal behavior is relevant here because observed avoidance mechanisms cannot readily be ascribed to cultural evolution. Reproductive relations among first-order relatives cause similarly severe attrition in mammals as in humans, resulting in comparable incentives for behavioral adaptations.18 behaviors have been observed in many species, most notably among : the principal mechanisms are dispersal upon sexual maturity and

15 Bittles, Consanguinity 135, 101-7. 16 George P. Murdock, Ethnographic Atlas (University of Pittsburgh Press 1967); Bittles, Consanguinity 3, 58-62. Cf., e.g., L. Holy, Kinship, Honour and Solidarity: Cousin Marriage in the Middle East (Manchester 1989); P. G. Reddy, Marriage Practices in South India: Social and Biological Aspects of Consanguineous Unions (Madras 1993). 17 Bittles, Consanguinity 191. This is lower than the composite 55 percentage rate derived from five studies (188-91) and disregards the possibility of disproportional increases at very high levels of proximity. Negative results are corroborated by outcomes in sects which attained comparable levels of inbreeding through transgenerational accumulation (e.g., Old Order Amish: Victor McKusick, (ed.), Medical Genetic Studies of the Amish (Baltimore 1978) or inbred dynasties (e.g., the Habsburgs: G. Alvarez, F. C. Ceballos and C. Quintero, “The role of inbreeding in the extinction of a European royal dynasty,” Public Library of Science One 4 (2009), e5174). This also disregards the fact that these are one-off occurrences: in a counterfactual environment without incest avoidance, many such unions would have been preceded by similar ones, resulting in significantly higher rates of attrition. The most relevant contrast is therefore not between baseline and one-time incestuous matings but between baseline and recurrent incestuous matings. 18 K. Ralls, J. D. Ballou and A. Templeton, “Estimates of lethal equivalents and the costs of inbreeding in mammals,” 2 (1988), 185-93; Anne Pusey and Maria Wolf, “Inbreeding avoidance in ,” Trends in Ecology and Evolution 11 (1996), 201-6; M. J. E. Charpentier, A. Widdig and S.C. Alberts, “Inbreeding depression in non-human primates: a historical review of methods used and empirical data,” American Journal of Primatology 69 (2007), 1370-86; Bittles, Consanguinity 94.

5 sexual aversion to co-resident kin. 19 Avoidance behavior appears to be generated in the first instance by proximity in childhood. In the pithy summary of a leading expert’s summary of research up to the early 2000s, “[m]ating among adults is most inhibited among maternal relatives in species in which these have close associations. The extent to which mating is inhibited among close paternal relatives is more variable and appears to depend largely, though perhaps not completely, on the closeness of association during maturity.” 20 The key mechanism, therefore, is association as such rather than actual (“positive”) .21 This raises the question whether humans display comparable forms of avoidance behavior that cannot simply be ascribed to cultural learning – in this case, the “incest taboo” – and whether they can be linked to comparable proximal mechanisms. Owing to the considerable challenges of abstracting a putatively innate trait from variegated cultural influences in accounting for observed outcomes, the answer to the first question remains contested. Empirical research has focused on adolescent and adult sexual avoidance behavior connected to close early childhood association with non-kin or more distant kin (or, inversely, on the association between the lack of such behavior and early separation). This would seem the best way to confirm the “,” named after Edward Westermarck’s 1891 claim that extended exposure in childhood leads to sexual aversion later in life. Evidence is furnished by “natural experiments” where individuals were raised as if they had been first-order kin without being related in this fashion or being expected to assume that social role and subsequently exhibited the kind of avoidance behavior usually found within nuclear families. The most extensively documented and analyzed case concerns the so-called “minor” marriages in Taiwan in the first half of the twentieth century in which young girls were adopted by the parents of their future husbands, raised as their sisters and later married to these biologically unrelated “brothers.” Based on meticulous analysis of some 20,000 marriages, Arthur Wolf has found significantly lower fertility rates and dramatically higher rates of divorce (and adultery) for such couples than for other unions. More specifically, and consistent with the hypothesis of early childhood association, the strength of this effect is inversely correlated with the girl’s age at adoption (and with the boy’s age in those rare cases where he was younger than the adopted girl) and thus positively correlated with the intensity of early childhood co- socialization. This effect peters out around age ten, when co-socialization was no longer significant: that marriages with girls adopted at higher ages were as successful as others speaks against social stigma as a cause for marital stress in “minor” marriages; nor was wealth a factor. Adoption per se was likewise not a significant variable as other types of marriages that involved adopted women were not affected by these problems. 22 The only remaining alternative explanation, resentment of abuse of adopted “sister-wives” by their new families, does not properly address the age-specific patterning. These data therefore provide fairly strong evidence for an age-dependent against later sexual relations. Another study observed that several thousand individuals who had been communally raised in Israeli kibbutzim in a way that mimicked upbringing among siblings did not later enter into marriage with

19 E.g., Anne E. Pusey, “Sex-biased dispersal and inbreeding avoidance in and mammals,” Trends in Ecology and Evolution 2 (1987), 295-9; “Mechanisms of inbreeding avoidance in nonhuman primates,” in J. R. Feierman (ed.), Pedophilia: biosocial dimensions (Springer 1990), 201-220; Pusey and Wolf, “Inbreeding avoidance.” 20 Anne Pusey, “Inbreeding avoidance in primates,” in Arthur P. Wolf and William H. Durham (eds.), Inbreeding, Incest, and the Incest Taboo: The State of Knowledge at the Turn of the Century (Stanford University Press 2004), 71. 21 More recent work is consistent with this, e.g. Zhu Yong, Li Jin-Hua, Xia Dong-Po, Chen Ra and Sun Bing Hua, “Inbreeding avoidance by female Tibetan macaques Macaca thibetana at Huangshan,” Acta Zoologica Sinica 54 (2008), 183-90; Elise Huchard, Leslie A. Knapp, Jinliang Wang, Michel Raymond and Guy Colishlaw, “MHC, mate choice and in a wild social ,” Molecular Biology 19 (2010), 2545-61. Cf. also more generally Anja Widdig, “Paternal kin discrimination: the evidence and likely mechanisms,” Biological Reviews 82 (2007), 319-34. 22 Arthur P. Wolf, and Early Childhood Association: A Chinese Brief for Edward Westermarck (Stanford University Press 1995); “Explaining the Westermarck effect or, What did select for?” in Wolf and Durham, Inbreeding , 76-92.

6 their cohort peers, except for a few who had been separated for significant amounts of time early in life. 23 Critics have pointed out that mandatory army service introduced them to large mating pools, curtailing the likelihood of peer-group unions. 24 More recently, a number of such individuals reported frequent sexual attraction in adolescence and abstention due to group expectations, a contrast to earlier observations that may be linked to intervening changes in rearing practices but cannot fully be explained by it. 25 New work on third-party attitudes to sexual relations with peers is not readily consistent with these self-reports (see below). A possible challenge to the sensitization model is posed by the nineteenth-century utopian Oneida community in New York where children were reared communally and many of them married peers regardless of alternative opportunities and did not experience lower fertility or any divorce. 26 This objection has been countered by more detailed observations to the effect that age and gender segregation during rearing prevented close proximity and hence putative sensitization for many of these couples. 27 The co-socialization approach has been expanded by assessing the impact of early childhood association on the success and appeal of consanguineous marriages. Findings include lower success rates for marriages of first cousins raised in sibling-like intimacy in Lebanon, successful first-cousin unions with substantial spousal age differences in Pakistan, and the curtailment of later marital appeal by childhood same-room sleeping of cousins in Morocco and more generally by close childhood propinquity among cousins on Sumatra. 28 Conversely, desensitization of actual first-order kin is suggested by increased self-reported sexual activity with procreative potential among siblings who had been extensively separated in early childhood. 29 A new line of research concentrates on third-party attitudes to incest, showing that childhood co-socialization among both siblings and non-kin produces disgust of incest. Among non-kin individuals communally reared in kibbutzim, the degree of sexual aversion to opposite-sex peers, the degree of disapproval of third-party peer sexual behavior including incest, and the extent of childhood co-sozialization have all been found to be positively correlated.30

23 Joseph Shepher, “Mate selection among second-generation kibbutz adolescents and adults: incest avoidance and negative ,” Archives of Sexual Behavior 1 (1971), 293-307; Incest: A Biosocial View (Academic Press: New York 1983). 24 John Hartung, Review of Shepher, Incest , American Journal of Physical Anthropology 67 (1985), 169-71. 25 Eran Shor and Dalit Simchai, “Incest avoidance, the incest taboo, and social cohesion: revisiting Westermarck and the case of the Israeli kibbutzim,” American Journal of Sociology 114 (2009), 1803-42, and their exchange with Alexandra Maryanski, Stephen K. Sanderson and Raymond Russell in the same journal (117 [2012], 1503-13). 26 D. R. Taves, “Rethinking Oedipus: an evolutionary perspective of incest avoidance: comment,” American Journal of Psychiatry 151 (1994), 297. 27 David M. T. Fessler, “Neglected natural experiments germane to the Westermarck hypothesis - The Karo Batak and the Oneida community,” Human Nature 18 (2007), 355-64. Lack of co-socialization before age six for women means that the apparent success of adoption-based marriages in New Guinea (similar to the Chinese “minor” marriages) cannot be used as an argument against the Westermarck effect: P. B. Roscoe, “Familiar partners? The Mountain Arapesh and the Westermarck effect,” Journal of Anthropological Research 51 (1995), 347-62. 28 J. McCabe, “FBD marriage: further support for the Westermarck hypothesis of the incest taboo?,” American Anthropologist 85 (1983), 50-69; C. McC. Pastner, “The Westermarck hypothesis and first cousin marriage: the cultural modification of negative sexual imprinting,” Journal of Anthropological Research 42 (1986), 573-86; A. Walter, “The evolutionary psychology of mate selection in Morocco: a multivariate analysis,” American Ethnologist 7 (1997), 300-17; Geoff Kushnick and Daniel M. T. Fessler, “Karo Batak cousin marriage, cosocialization, and the Westermarck hypothesis,” Current Anthropology 52 (2011), 443-8. 29 Irene Bevc and Irwin Silverman, “Early separation and sibling incest: a test of the revised Westermarck theory,” Evolution and Human Behavior 21 (2000), 151-61. 30 Siblings: Daniel M. T. Fessler and C. David Navarrete, “Third-party attitudes toward incest: evidence of the Westermarck effect,” Evolution and Human Behavior 25 (2004), 277-94; Debra Lieberman, John Tooby and Leda Cosmides, “Does morality have a biological basis? An empirical test of the factors governing moral sentiments regarding incest,” Proceedings of the Royal Society B 270 (2003), 819-26. Non-kin: Debra Lieberman and Thalma Lobel, “Kinship on the kibbutz: coresidence duration predicts altruism, personal sexual aversions, and moral attitudes among communally reared peers,” Evolution and Human Behavior 33 (2012), 26-34. Both: Jan Antfolk,

7

Whilst of uneven quality and inevitably open to criticism, 31 behavioral observation studies offer non-trivial support for the postulate of avoidance-inducing mechanisms that are unrelated to biological kinship and cultural learning. 32 The next question is which physiological mechanisms might account for this phenomenon. Animal studies highlight the importance of the major histocompatibility complex (MHC), a highly polymorphic group of genes that serves as a matching system used by the immune system to discriminate between self and others, in influencing mating preferences by conveying information via olfactory cues. When it can be documented, preference tends to be for genetic difference, thereby facilitating inbreeding avoidance. 33 In mice, the importance of associate reference (in which an indirect referent for genetic difference is established through prior sensitization to the body odor of others around them) has been demonstrated by cross-fostering experiments in which mice removed from their natal litter after birth and reared by unrelated females subsequently exhibited avoidance of the MHC type of the foster mother but not of their actual kin. 34 Sensitization in general is most likely linked to critical periods in life such as early childhood.35 Although hopes that discovery of a similar mechanism could explain the Westermarck effect in humans have so far remained unfulfilled, promising leads do exist.36 In several studies, human subjects in blind trials preferred the body odor of MHC-dissimilar individuals, even though a comparable bias has only occasionally been established for married couples.37 The most pertinent study focuses on an

Mira Karlsson, Anna Backstrom and Pekka Santtila, “Disgust elicited by third-party incest: the roles of biological relatedness, co-residence, and family relationship,” Evolution and Human Behavior 33 (2012), 217-23. 31 In addition to the above studies, see the critical assessments in Fessler and Navarrete, “Third-party attitudes,” (balanced), Gregory C. Leavitt, Incest and Inbreeding Avoidance: A Critique of Darwinian Social Science (Edwin Mellen Press 2005) (dismissive a priori), and Shor and Simchai, “Incest avoidance,” negative but particularly important for their insistence on distinguishing between claims of “indifference” and “aversion” to incest. In a detailed survey, Markus J. Rantala and Urszula M. Marcinkowska, “The role of sexual imprinting and the Westermarck effect in mate choice in humans,” Behavioral Ecology and Sociobiology 65 (2011), 859-73 note the limits of the evidence and the need for more robust studies. 32 This is not the only mechanism that has been proposed: for a dual system complemented by evolved response to “maternal perinatal association” (i.e. witnessing one’s mother’s care for another child), see now Debra Lieberman, John Tooby and Leda Cosmides, “The architecture of human kin detection,” Nature 445 (2007), 727-31, applied to historical data by Debra Lieberman, “Rethinking the Taiwanese minor marriage data: evidence the mind uses multiple kinship cures to regulate inbreeding avoidance,” Evolution and Human Behavior 30 (2009), 153-60. 33 D. Penn and W. Potts, “How do major histocompatibility genes influence odor and mating preferences?” Advances in Immunology 69 (1998), 411-36; “The scent of genetic compatibility: and the major histocompatibility complex,” Ethology 108 (2002), 1-21. On kin recognition mechanisms more generally, cf. now Dustin J. Penn and Joachim G. Frommen, “Kin recognition: an overview of conceptual issues, mechanisms and evolutionary theory,” in Peter Kappeler (ed.), Animal behavior: evolution and mechanisms (Springer 2010), 55-85. 34 K. Yamazaki, G. K. Beauchamp, D. Kupniewski, J. Bard, L. Thomas and E. A. Boyse, “Familial imprinting determines H-2 selective mating preferences,” Science 240 (1988), 1331-2; Amy Eklund, “The effect of early experience on MHC-based mate preferences in two B10.W strains of mice ( Mus domesticus ),” Behavior Genetics 27 (1997), 223-9; D. Penn and W. Potts, “MHC-disassortative mating preferences revised by cross-fostering,” Proceedings of the Royal Society of London B 265 (1998), 1299-306. 35 Cf. Peter A. Brennan and Keith M. Kendrick, “Mammalian social odours: attraction and individual recognition,” Philosophical transactions of the Royal Society B 361 (2006), 2061-78. For the actual process, see Hideto Kaba, “Neurobiology of mammalian olfactory learning that occurs during sensitive periods,” Current Zoology 56 (2010), 819-33. 36 Cf. M. A. Schneider and L. Hendrix, “Olfactory sexual inhibition and the Westermarck effect,” Human Nature 11 (2000), 65-91 for what is needed. 37 Comprehensively summarized and discussed by Jan Havlicek and S. Craig Roberts, “MHC-correlated mate choice in humans: a review,” Psychoneuroendocrinology 34 (2009), 497-512, to which we may now add Romain Laurent and Raphaelle Chaix, “MHC-dependent mate choice in humans: why genomic patterns from the HapMap European American dataset support the hypothesis,” Bioessays 34 (2012), 267-71. See also Thomas Roberts and Jonathan P. Roiser, “In the nose of the beholder: are olfactory influences on human mate choice driven by variation in immune system genes or sex hormone levels?” Experimental Biology and Medicine 235 (2010), 1277-81.

8 endogamous population of Hutterites in which over 400 conjugal unions did not only show marked preference for MHC-dissimilar spouses but also revealed that those MHC matches that did occur among couples were twice as often inherited from fathers than from mothers. This points to a sensitization process through differential association in childhood, with more intense contact with mothers (in a system with traditional gender roles and strict division of labor) and thus greater sensitization to their MHC signatures. This represents the best evidence suggestive of how mating preference (and hence avoidance behavior) might be acquired through childhood association that is currently available. 38 Studies of foster children’s responses to the odor of their birth and foster families are required to clarify this matter.39 Why is this relevant to historians? A universal innate psychological adaptation that is effectively conducive to learned incest avoidance – the (subconscious) learning being achieved through neurochemical processes mediated by early childhood association – may be expected to produce different historical outcomes than learned incest avoidance that entirely depends on the ongoing transmission of purely cultural adaptations that address this issue, that is, “incest taboos.” 40 A gene-based trait would create a universal background constraint, raising the bar for behavioral exceptions. In this context, it catches the eye that one thing that virtually all well-documented historical cases of non-criminal sexual relations within the nuclear family have in common is that they were regarded, emically, as deliberate deviations from the norm, regardless of cultural context. 41 Such occurrences were largely limited to “royal incest,” which is exceptional by definition: in these cases, transgressive mating behavior was meant to mark out rulers (or at most narrow elites) as special or unique and approximate them to (frequently incestuous) deities or primal ancestors. More pragmatically, however, they were often either embedded in polygamy (which would absorb the costs of inbreeding depression) or merely nominal, non-consummated unions. 42 By contrast, habitual non-royal incest has been extremely rare, confined in the first instance to ancient and early medieval Zoroastrian (Mazdean) culture. 43 As noted at the beginning of this paper, the

38 Carole Ober, Lowell R. Weitkamp, Nancy Cox, Harvey Dytch, Donna Kostyu and Sherman Elias, “HLA and mate choice in humans,” American Journal of Human Genetics 61 (1997), 497-504, esp. 503, with Fessler and Navarette, “Third party-attitudes” for elaboration. 39 An earlier finding of mutual aversion between fathers and pubertal daughters and between pubertal opposite-sex siblings, which is not readily explicable in terms of differential early childhood sensitization but might point to different avoidance-fostering mechanisms (Glenn E. Weisfeld, Tiffany Czilli, Krista A. Phillips, James A. Gall and Cary M. Lichtman, “Possible olfaction-based mechanisms in human kin recognition and inbreeding avoidance,” Journal of Experimental Child Psychology 85 [2003], 279-95 ) could not be replicated by a recent study: Camille Ferdenzi, Benoist Schaal and S. Craig Roberts, “Family scents: developmental changes in the perception of kin body odor?” Journal of Chemical Ecology 36 (2010), 847-54. 40 Various rationales for this taboo have been proposed in the literature: see above, n.4. William H. Durham, Coevolution: Genes, Culture, and Human Diversity (Stanford University Press 1991), 286-360, based on a cross- cultural survey argues that avoidance behavior is derived from belief in (observed) deleterious effects of very close inbreeding. (I note in passing that cultural acquisition of the incest taboo is also capable of triggering bodily reflexes: E. B. Royzman, R. F. Leeman and J. Sabini, “You make me sick”: moral dyspepsia as a reaction to third- party sibling incest,” Motivation and Emotion 32 [2008], 100-8). 41 The fullest treatment is still Norbert Sidler, Zur Universalität des Inzesttabu: Eine kritische Untersuchung der These und der Einwände (Stuttgart 1971). 42 See esp. R. H. Bixler, “Sibling Incest in the Royal Families of Egypt, Peru, and Hawaii,” Journal of Sex Research 18 (1982), 264-281; and see also, e.g., John M. Goggin and William C. Sturtevant, “The Calusa: a stratified, nonagricultural society (with notes on sibling marriage),” in Ward H. Goodenough (ed.), Explorations in Cultural Anthropology (McGraw-Hill 1964), 179-219, at 202-7; W. H. Davenport, Pi'o: An Enquiry into the Marriage of Brothers and Sisters and other Close Relatives in Old Hawai'i (Lanham 1994); S. L. Ager, “Familiarity : incest and the Ptolemaic dynasty,” Journal of Hellenic Studies 125 (2005), 1-34; Grant Evans, “When brother-sister marriage becomes incest,” Australian Journal of Anthropology 21 (2010), 228-45. 43 The literature is large: the most recent treatments are Paul J. Frandsen, Incestuous and Close-Kin Marriage in Ancient Egypt and Persia (Museum Tusculanum Press 2009), 60-120, and (better) Prods O. Skjaervo, “Marriage

9 exceptional character of this practice (called xwedodah ) was reflected in its strenuous glorification as an act of religious devotion extolled as “the greatest good work of the religion,” “the best action of the living,” and “the second greatest good deed,” sure to generate “a sublime radiance”, scare away demons and offset mortal sins.44 It was not merely condemned by outside observers from Europe to China – which might simply be regarded as a sign of cultural variation – but also considered a great challenge by its own proponents, as in this previously cited exhortatory treatise: “Zoroaster [i.e., the prophet] said to Ohrmazd [i.e., the supreme god]: ‘In my view (it is) bad and hard and distressing that I should make xwedodah so prevalent among mankind.’ Ohrmazd said: ‘In my view also (so it would be) as in yours, except for this (reason), that it is the (most) excellent thing of all; then let it not be difficult and hard for you. Be diligent in practicing xwedodah , and others, too, will practice diligently.’” 45 Among early Zoroastrians, this practice appears to have served as an honest signal of strong commitment to challenging group ideals. Albeit often portrayed as a form of regular marriage (and as such presumably buffered by polygamy, which was common in elite circles where this practice might have been more common), the most detailed extant account portrays it more as a rare ritual: four copulations were deemed an extraordinary achievement worthy of eternal salvation.46 Thanks to the lack of royal or religious connotations, the only other case on record, brother-sister marriage in Roman Egypt in the first few centuries CE, seems more difficult to explain. According to census documents, a significant proportion of recorded commoners were married to full siblings in unions that produced lots of children.47 However, the very existence of this phenomenon is now in doubt as it has recently been argued that sibling ties between these spouses may have been created by adoption rather than biological kinship. 48 In the absence of more than circumstantial evidence and cross-cultural analogies, this thesis is highly contested. 49 Yet even if these couples were indeed made up of biological siblings, observed spousal age gaps and divorce rates and the possibility of prolonged cross-fostering by wet-nurses might nevertheless allow us to reconcile their existence with the basic premises of the Westermarck effect. 50 Overall, the historical evidence is better consistent with a model of biologically grounded aversion behavior than with the notion of preferences determined solely by culturally contingent norms. This rapid survey has highlighted the formidable complexities of testing even a single EP- inspired prediction. Although the proposition is theoretically plausible in terms of ultimate causation – that the substantial costs of intense inbreeding depression caused by nuclear family incest provide next of kin,” in Encyclopaedia Iranica , http://www.iranicaonline.org/articles/marriage-next-of-kin (last updated June 29, 2011; accessed August 2, 2012). 44 Pahlavi Vistasp Yasht 17; Wizidagiha i Zadspram 26.3; Dadestan i Menog i Xrad 4.4; Denkard 9.41.27. See also above, n.1-2. 45 The Pahlavi Rivayat Accompanying the Dadestan i Denig 8o1. Aversion is corroborated by a resident outsider, Jesubokht, who also refers to congenital disorders, which may however be a stereotypical charge (E. Sachau, Syrische Rechtsbücher , vol. 3 [Berlin 1914], 35, 37). 46 Sayes-ne-sayast 18.4; The Pahlavi Rivayat Accompanying the Dadestan i Denig 8f2 (cf. 8h1 for four years of marriage, prima facie implying the possibility of intercourse just once a year). 47 Roger S. Bagnall and Bruce W. Frier, The Demography of Roman Egypt (Cambridge University Press 1994), 127- 34. Influential cultural-constructivist studies include Keith Hopkins, “Brother-Sister marriage in Roman Egypt,” Comparative Studies in Society and History 22 (1980), 303-54; Brent D. Shaw, “Explaining incest: brother-sister marriage in Graeco-Roman Egypt,” Man 27 (1992), 267-99. For a biologically informed perspective, see Walter Scheidel, Measuring Sex, Age and Death in the Roman Empire (Journal of Roman Archaeology 1996), 9-51. 48 Sabine R. Hübner, ““Brother-sister” marriage in Roman Egypt: a curiosity of humankind or a widespread family strategy?” Journal of Roman Studies 97 (2007), 21-49. 49 Rejected by Sofie Remijsen and Willy Clarysse, “Incest or adoption? Brother-sister marriage in Roman Egypt revisited,” Journal of Roman Studies 98 (2008), 53-61; Jane Rowlandson and Ryosuke Takahashi, “Brother-sister marriage and inheritance strategies in Greco-Roman Egypt,” Journal of Roman Studies 99 (2009), 104-39. 50 Walter Scheidel, “Ancient Egyptian sibling marriage and the Westermarck effect,” in Wolf and Durham, Inbreeding , 93-108. If the phenomenon was real, inbreeding depression may have been considerable: Walter Scheidel, “Brother-sister marriage in Roman Egypt,” Journal of Biosocial Science 29 (1997), 361-71.

10 powerful selective pressures for a countervailing adaptation – and consistent with varied ethological findings, it has proven hard solidly to confirm it with the help of behavioral data. The underlying physiological processes are still only dimly perceptible. The resultant uncertainty qualifies both confident claims about evolved avoidance mechanisms and equally confident dismissals of the very notion. Few other hypotheses in the field of EP have attracted such close scrutiny. If all of them did, how well would the adaptationist program hold up? The answer to this question will not emerge from partisan commitments but only from further and more robust empirical studies. Historians may well want to defer judgment on the merits of this case or others, but they can ill afford to ignore a research agenda that seeks to cast light on the framing conditions for human behavior and the limits of behavioral plasticity. Without an awareness of these constraints, how are historians to make sense of a phenomenon such as nuclear- family incest and the limited appeal of sporadic cultural mutations that rendered this practice socially acceptable? Most importantly, and this is a point very much worth stressing in this Forum, this is by no means a one-way street. We historians are not meant to be a mere audience for the findings of other disciplines. Our intimate familiarity with an enormous wealth of behavioral data and the interpretive challenges they raise puts us in an excellent position to make a genuine contribution to the ongoing debate about the merits of EP approaches. 51 To return one more time to the present example, any assessment of adaptationist interpretations of incest avoidance critically relies on an appreciation of the long-term historical record: it is only by taking proper account of historical instances that may seem to conflict with notions of a universal (and ultimately innate) incest taboo and of the specific circumstances and properties of these instances that EP can hope to avoid narrowly presentist and thus potentially unrepresentative empirical testing of its hypotheses. In fact, this is an area where historians have already been able to make themselves heard. 52 True engagement cannot arise from passive consumption: give-and-take is the only way forward. Historians on the one hand and experimental social scientists and life scientists on the other must work together to develop viable models of the variegated factors that influence human behavior and historical outcomes.

51 Cf. the growing application of EP in the global study of literature: e.g., Joseph Carroll, Literary Darwinism (Routledge 2004); “Literature and Evolutionary Psychology,” in Buss, Handbook 931-52; Jonathan Gottschall, Literature, Science and a New Humanities (Palgrave Macmillan 2008); The Storytelling Animal: How Stories Make Us Human (Houghton Mifflin Harcourt 2012). 52 Reflected in L. Hendrix and M. A. Schneider, “Assumptions on sex and society in the biosocial theory of incest,” Cross-Ctural Research 33 (1999), 193-218; Wolf and Durham, Incest ; Bittles, Consanguinity 184-7.

11