sign in sign up
<<
Home , Eusociality, Herbert Gintis, Sociality

Fall Focus on Books

Clash of the Titans

The Social Conquest of Earth. Edward E. O. Wilson defends the notion manipulation (Reeve and Keller 1997). O. Wilson. W. W. Norton (Liveright), that are eusocial by recalling Therefore, he writes in Social Conquest 2012. 352 pp., illus. $27.95. (ISBN that early hominins adopted collec- on page 144, 978-0-87140-413-8 hardcover). tive child rearing in a fixed location defended by some group members The origin and while others dispersed to forage, scav- of eusocial can be dward O. Wilson is an expert on enge, or hunt (Burkart et al. 2009, viewed as processes driven by Esocial insects, but early in his Chapais 2009, Hrdy 2009, Gintis and individual-level natural selec- career, he addressed . van Schaik 2012, Pagano 2012). All tion. It is best tracked from His book (1975) initi- known eusocial species went through queen to queen from one gen- ated a field that has become a major a similar stage of collective brood eration to the next, with the research area in the behavioral sci- rearing. With (the pro- workers of each pro- ences and included a controversial vision of by individuals duced as phenotypic exten- chapter on human social evolution. other than the recipient’s biological sions of the mother queen.… In The Social Conquest of Earth, E. O. parent) in a fixed location established, The defending worker is part Wilson updates his earlier ground- E. O. Wilson argues, there is strong of the queen’s phenotype, as breaking treatment of Homo sapiens. selection on traits that improve coop- teeth and fingers are part of He stakes out a strong position on erative behavior, which would culmi- your own phenotype. the levels-of-selection debate, rejects nate in . inclusive-fitness theory, and redefines To supporters of inclusive-fitness eusociality to include Homo sapiens. Clash of the Titans theory, these statements are outrageous The book is based on several articles E. O. Wilson’s theory thus diverges and strike a blow at population genet- published in professional journals, from the widely held view that ics itself. E. O. Wilson and his coau- including those of E. O. Wilson and human sociality can be explained by thors simply could not be believed. Hölldobler (2005); D. S. Wilson and invoking Hamilton’s (1964a, 1964b) Because of its visibility in the journal Wilson (2007); E. O. Wilson (2008); inclusive-fitness theory. Inclusive- , Nowak, Tarnita, and Wilson’s and, most prominently, Nowak, Tar- fitness theory shows that a person- (2010) critique of inclusive-fitness the- nita, and Wilson (2010). ally costly behavior can spread if the ory has drawn the collective ire of a E. O. Wilson’s work has stirred up behavior offers sufficient benefits to host of prominent population bio- a vigorous—and sometimes highly genetically related individuals. E. O. logists. Nature subsequently published intemperate—debate that has shaken Wilson now rejects this theory, writ- several brief communications in which the very foundations of population ing in The Social Conquest of Earth Nowak, Tarnita, and Wilson’s (2010) biology. In this review, I address the (p. 51) that claims were vigorously rejected, one of fundamental issues involved and sug- which had 134 signatures (Abbot et al. gest a resolution that may be accept- The foundations of the gen- 2011, Boomsma et al. 2011, Strass- able to both sides of this clash of the eral theory of man et al. 2011). Rousset and Titans. based on the assumptions of (2011) accused Nowak, Tarnita, and E. O. Wilson’s treatment of euso- have crumbled, Wilson (2010) of saying nothing new ciality in The Social Conquest of Earth while evidence for it has grown and of using “rhetorical devices.” They is a major innovation. A species is equivocal at best.… Inclusive- then attacked the journal Nature itself, customarily defined as eusocial if it fitness theory, often called kin arguing (p. 1390) that is composed of colonies in each of selection theory, is both math- which there is a reproductive divi- ematically and biologically The publication of this article sion of labor, as in the cases of social incorrect. illustrates more general con- and , in whose cerns about the publishing a single or a few individuals (queens E. O. Wilson favors the more tradi- process.… Nature’s extravagant and their mates) reproduce and other tional individual-fitness-maximization editorial characterization of the castes do not. In humans, by contrast, model in understanding advanced paper as “the first mathemati- adults are generically reproductives. eusociality, which is defined as occur- cal analysis of inclusive-fitness ring if the workers are sterile or if theory” recklessly tramples on worker is suppressed nearly 50 years of accumulated doi:10.1525/bio.2012.62.11.8 by policing (Frank 1995) or queen knowledge. www.biosciencemag.org November 2012 / Vol. 62 No. 11 • BioScience 987 Fall Focus on Books

In passing sentence on The Social cooperate with other genes in promot- only replicators—the sole entities that Conquest of Earth, Dawkins (2012) ing the success of the genome as a are preserved in the evolutionary pro- concluded, “To borrow from Dorothy whole. Of course, the same argument cess. Individuals, in Dawkins’s view, Parker, this is not a book to be tossed applies to inclusive fitness at the level are simply vehicles that genes build to lightly aside. It should be thrown with of the individual in an advanced euso- shuttle themselves successfully from great force.” cial colony. one body to another. Unlike genes, By contrast, some supporters of genotypes and groups of genotypes Hamilton’s rule and its limitations Hamilton’s rule take the notion that are transitory in sexually reproducing Kin selection theory holds that an genes maximize inclusive fitness as the species, he argues; they are broken up allele for helping another will spread very core of evolutionary biological and reorganized continually by sexual if a mathematical inequality known theory. For instance, West and col- recombination. as Hamilton’s rule holds. The rule says leagues (2011, p. 233) explained that But, in fact, genomes are replicators that the allele will spread if the benefit “since Darwin, the only fundamen- as much as genes are, and the pheno- from the help, in terms of fitness to tal change in our understanding of types that genomes control include the recipient, multiplied by the recipi- adaptation has been Hamilton’s devel- metazoan individuals and social ent’s degree of relatedness to the bene- opment of inclusive-fitness theory.… groups, such as E. O. Wilson’s eusocial factor, is greater than the cost to the The idea [is] that organisms can be colonies and human societies. This is benefactor. Degree of relatedness is usu- viewed as maximizing agents.” so because a gene promotes the fitness ally understood as the probability that of its host only if it is complemented an allele is identical in the benefactor by a well-orchestrated configura- and the recipient by common descent. tion of genes with which it interacts In box 1, I derive a generalization of (Sober and Lewontin 1982, Hammer- Hamilton’s rule that relaxes key, usu- stein and Leimar 2006, Noble 2011). ally unstated (and often implausible) Species-defining loci in the genome assumptions about the costs and ben- are therefore necessarily homozygous efits of helping behavior to relatives and identical in both members of a versus nonrelatives that are essential pair so that crossing over and in the derivation of the standard form meiosis do not destroy the integrity of of Hamilton’s rule. My generalization these complementary loci. Of course, makes clear that an allele that causes this assertion must be qualified to deal helping can spread even if it reduces with introns and exons, dominance, the mean fitness of the population. maternal–paternal conflicts, auto- It follows that Hamilton’s rule does The technical literature says other- somal and sex-linked chromosomes, not explain eusociality and, indeed, is wise. The conditions under which evolu- fitness-neutral base replacements, and focused on what may be a minor subset tionary dynamics lead to inclusive-fitness the like. But these adjustments do not of the biological conditions that favor maximization have been carefully stud- weaken the integrity of the genome. the fitness of a gene and promote coop- ied by Alan Grafen and his fellow We may call the genes that support eration in metazoan cells and social researchers, who have shown that this phenotypic commonality the spe- species (Hammerstein and Leimar Darwinian population dynamics entail cies’ core genome. The core genome is a 2006); the absence of inclusive-fitness- inclusive-fitness maximization at the replicator and, therefore, subject to the maximizing behavior even when Ham- individual and gene levels, but only laws of . ilton’s rule is satisfied is widely observed assuming that fitness effects are addi- in social species (Queller and Strass- tive (Grafen 1999, 2006, Gardner and Levels of selection mann 1998, Nonacs 2011). The cooper- Welsh 2011, Gardner et al. 2011). But, In urging a harmonious resolution ation of genes in metazoan cells (Leigh of course, the additivity assumption to the clash of the Titans, Okasha 1971) and in social groups (Bowles and implies that a = = 0 in the general- (2012) suggested that the level of selec- Gintis 2011) is therefore effected by ized Hamilton rule, which is expressed tion is a matter of personal choice or selection at a higher level than the gene in equations 3 and 5 in box 1, in ease of analysis, because mathemati- itself. E. O. Wilson’s critique of Hamil- which case Hamilton’s rule is fully cal models show that the result is not ton’s rule is justified. acceptable. affected by this choice: All frameworks give the same answer. Many biologists Genes survive because they Core genomes as replicators agree with this assessment (Queller cooperate In his critique of The Social Conquest 1992, Dugatkin and Reeve 1994, Genes survive the evolutionary dyna- of Earth, Dawkins (2012) repeated an Wilson DS and Sober 1994, Kerr and mic, then, not because they maxi- argument that he first developed in Godfrey-Smith 2002, Lehmann et al. mize inclusive fitness but because they (1976): Genes are the 2007, Wilson DS and Wilson 2007).

988 BioScience • November 2012 / Vol. 62 No. 11 www.biosciencemag.org Fall Focus on Books

Box 1. A generalized Hamilton’s rule.

The derivation of Hamilton’s rule is clean, intuitive, and simple. Suppose there is an allele at a locus of a haploid organism that leads individual A to incur a fitness cost c that bestows a fitness gain b on individual B. Suppose the frequency of this allele in the population is q, and the probability that B has a copy of the allele is p. Then, if the size of the population is n, there are qn individuals with the helping allele, and they increase the number of members of the population from n to n + qn(b – c), but they increase the number of helping alleles from qn to qn + qn(pb – c). Therefore, the frequency of the allele from one period to the next will increase if qn + qn(pb – c) b(p – q) ∆q = – q = – c > 0. (1) n + qn(b – c) 1 – q

Now suppose that B has a degree of relatedness r to A, meaning that B has a copy of the helping allele with probability r by common descent from an ancestor. Then we can write

p = r + (1 – r)q,

where the second term comes from the fact that if B does not have the helping allele by descent from a common parent, B may still have the allele with the populationwide probability q. Substituting the above expression for p in equation 1, we arrive at Hamilton’s rule for the allele to spread:

br > c. (2)

We can improve the realism of this equation by adding diploidy, , dominance, mutual helping, and the like. Uyenoyama and Feldman (1980), Michod and Hamilton (1980), Queller (1992), Rousset and Billard (2000), and many others have shown that something resembling equation 2 will still hold, although in general the frequency q will also appear in the revised equation 2. Suppose, however, that in performing the helping act, A induces a fitness cost on nonrelatives and a fitness cost on the colony as a whole. For example, > 0 can occur if A redirects brooding care from nonrelative to relative larvae, and > 0 if A’s help reduces its contribution to colonywide tasks. If the population size is n in the current period, population size n′ in the next period will include n + qn(b – c) because of the helping behavior but will be reduced by n q and, because the number of nonrelatives is n(1 – q), will be additionally reduced by n(1 – q) . The number of relatives of the helping allele in the current period is qn, which is increased by the helping behavior by qn(pr – c) and decreased through lower colony efficiency by qn q. Therefore, the new population size is

n′ = n(1 – (1 – q) – q) + qn(b – c), (3) and equation 1 becomes

qn(1 – βbq) + qn(pb – c) ∆q = – q > 0, (4) n(1 – αq(1 – q) – βq) + qn(b – c) which simplifies to

b(p – q) + (1 – q)q > c (1 – q).

Substituting p = r + (1 – p)q, we get what I will call the generalized Hamilton’s rule:

br + q > c. (5)

Equation 5 shows that the cost to the colony of the helping behavior ( ) is ignored in the equation for the fitness of the helping allele and that transferring resources from nonrelatives to relatives ( > 0) can lead to an increase in the frequency of the helping allele even if it in no other way benefits relatives (b = 0). The effect on population fitness of an increase in the helping allele is then given by equation 6, which is derived from equation 3:

dn´ = n(b – c – α(1 – 2q) – β). (6) dq

There is, of course, nothing to prevent this effect from being negative even when the Hamilton condition br > c is satisfied or from being positive if even if Hamilton’s rule is violated.

However, this is true only for account- Because there are replicators at vari- to an evolutionary dynamic, selec- ing purposes; consideration of causal ous levels of biological organization tion must also occur at various levels. processes tells a different story. and because replicators are subject A segregation distorter, for instance, www.biosciencemag.org November 2012 / Vol. 62 No. 11 • BioScience 989 Fall Focus on Books operates at the level of the gene, parental reproductives can lead work- Conclusions because selective forces operative at ers to behave in ways that reduce E. O. Wilson’s assertion that the higher levels need not be invoked in the mean fitness of the colony, but inclusive-fitness approach is “both its explanation (Sober and Lewontin such behavior is not an expression mathematically and biologically incor- 1982, Burt and Trivers 2006). Where of the workers’ inclusive fitness but, rect” should be revised. A more accu- the core genome specifies structured rather, a transcription error. Such rate assessment is that Hamilton’s rule interactions among related individu- “outlaw” workers are not “policed” is incomplete, and with a few changes, als, kin selection occurs and is there- but are, rather, eliminated or con- it can maintain its exalted position fore a form of . trolled through an error correction in sociobiology. These changes lead Where the causal relations inscribed process within the colony. Like cogs in us to the generalized Hamilton’s rule, in the core genome occur between or a machine, a worker in such a colony which points beyond itself to the need among metazoan individuals, analysis cannot “defect,” although it can “fail” for many additional tools in the biolo- in terms of group selection at a level as a part in a machine (Hammerstein gist’s repertoire to explain complex higher than the individual is appro- and Leimar 2006). . priate. The group could merely affect could have explained all this in his The evolutionary forces that deter- fitness payoffs to its members, but famous book The Extended Phenotype mine the nonlinear interactions generally, this will entail that group (1989); he might have saved a lot of among loci in metazoans and among structure be inscribed in the core futile debate. individuals in sociobiological species genome. Social groups, then, are sub- must be studied using tools such as ject to exactly the same evolution- Homo sapiens is not a eusocial the phenotypic gambit (Grafen 1984), ary forces—replication with mutation, species (DS Wilson 1977, 1980, selection, and heredity—as are indi- Humans evolved in cohesive groups Taylor 1992, 1996), agent-based mod- vidual genes and metazoan individuals. involving many unrelated eling (Gintis 2009), and the physiol- Whatever the level at which selection (Boyd and Silk 2002) and lack a repro- ogy of suppressor genes (Leigh 1977). operates, the resulting dynamic can, of ductive division of labor (but see Foster E. O. Wilson and the coauthors of his course, at least in principle, be trans- and Ratnieks 2005). Moreover, human recent publications are to be lauded for lated into changes in gene frequencies. sustains a high level of coop- expanding our horizons, not scorned eration among genealogically distant for their small mistakes and tendency No in advanced eusocial individuals (Gintis 2000, Richerson to overstate their position. So The species and Boyd 2004, Bowles and Gintis Social Conquest of Earth is, despite Biological altruism is defined as a 2011). By contrast, the reproductive some weaknesses, a book that I recom- behavior that lowers the personal fit- division of labor is central to sustain- mend. It is an inspiring account of a ness of the focal individual but raises ing cooperation in eusocial species, significant advance in our understand- the fitness of other individuals (Foster just as sequestration of the germline ing. Contemporary inclusive-fitness 2008). In the case of advanced euso- is key to the emergence of complex theory is a very big boat, and it can ciality, worker fitness is zero (i.e.,c = 0 metazoans (Buss 1987). Thus I argue, only be rocked by giants (Gadagkar in equations 2 and 5 in box 1). Altru- contra E. O. Wilson, that it is prudent 2010). ism in advanced eusocial species is to exclude humans from the eusocial therefore precluded in principle. Even fold. Acknowledgments the generalized Hamilton’s rule cannot Much of E. O. Wilson’s critique of I would like to thank Samuel Bowles, explain altruism in advanced eusocial the inclusive-fitness model of eusocial- Steven Frank, and David Queller for species, because it does not exist. ity does not apply directly to humans advice in preparing this article. In an advanced eusocial colony, or to primitively social species. In our E. O. Wilson maintains, the workers work, my colleagues and I routinely References cited “are extensions of the queen’s pheno- deploy inclusive-fitness models, along Abbot P, et al. 2011. Inclusive fitness theory and type… robots she has created in her with higher-level competition among eusociality. Nature 471: E1–E4. image that allow her to generate more groups, to explain the emergence of Boomsma JJ, Beekman M, Cornwallis CK, queens and males than would be pos- altruism in humans (Fehr and Gin- Griffin AS, Holman L, Hughes WOH, Keller L, Oldroyd BP, Ratnieks FLW. 2011. Only sible if she were solitary” (p. 143). This tis 2008, Bowles and Gintis 2011). full-sibling families evolved eusociality. is correct. Workers have no active evo- In human society, individuals who Nature 471: E4–E5. lutionary role in an advanced eusocial attempt to maximize their inclusive Bowles S, Gintis H. 2011. A Cooperative Species: species. They are therefore analogous fitness at the expense of the group are Human Reciprocity and Its Evolution. to robots, just as cells in the meta- generally penalized not by the “parlia- Princeton University Press. Boyd R, Silk JB. 2002. How Humans Evolved, 3rd zoan body are—however complex ment of genes” but by social sanctions, edition. W. W. Norton. they may be—basically analogous to which are parliaments on yet a higher Burkart JM, Hrdy SB, Van Schaik CP. 2009. robots. Of course, mutations in the level. Cooperative breeding and human cognitive

990 BioScience • November 2012 / Vol. 62 No. 11 www.biosciencemag.org Fall Focus on Books

evolution. Evolutionary Anthropology 18: Proceedings of the Royal Society B 266: Richerson PJ, Boyd R. 2004. Not by Genes 175–186. 799–803. Alone: How Culture Transformed Human Burt A, Trivers R. 2006. Genes in Conflict: The ———. 2006. Optimization of inclusive fit- Evolution. University of Chicago Press. Biology of Selfish Genetic Elements. Harvard ness. Journal of Theoretical Biology 238: Rousset F, Billard S. 2000. A theoretical basis University Press. 541–563. for measures of kin selection in subdivided Buss LW. 1987. The Evolution of Individuality. Hamilton WD. 1964a. The genetical evolution populations. Journal of Evolutionary Biology Princeton University Press. of social behavior, I. Journal of Theoretical 13: 814–825. Chapais B. 2009. Primeval Kinship: How Pair- Biology 7: 1–16. Rousset F, Lion S. 2011. Much ado about Bonding Gave Birth to Human Society. ———. 1964b. The genetical evolution of social nothing: Nowak et al.’s charge against inclu- Press. behavior, II. Journal of Theoretical Biology sive fitness. Journal of Evolutionary Biology Dawkins R. 1976. The Selfish Gene. Oxford 7: 17–52. 24: 1386–1392. University Press. Hammerstein P, Leimar O. 2006. Cooperating Sober E, Lewontin RC. 1982. Artifact, cause, ———. 1989. The Extended Phenotype: The for direct fitness benefits. Journal of and genic selection. Philosophy of Science Gene as the Unit of Selection. Freeman. Evolutionary Biology 19: 1400–1402. 49: 157–180. ———. 2012. The descent of Edward Wilson. Hrdy SB. 2009. Mothers and Others: The Strassman JE, Page RE Jr, Robinson GE, Seeley Prospect (24 May 2012). (29 August Evolutionary Origins of Mutual Under- TD. 2011. Kin selection and eusociality. 2012; http://www.prospectmagazine.co.uk/ standing. Belknap. Nature 471: E5–E6. magazine/edward-wilson-social-conquest- Kerr B, Godfrey-Smith P. 2002. Individualist Taylor P. 1992. Altruism in viscous populations: earth-evolutionary-errors-origin-species) and multi-level perspectives on selection An inclusive fitness model. Evolutionary Dugatkin LA, Reeve HK. 1994. Behavioral ecol- in structured populations. Biology and 6: 352–356. ogy and the levels of selection: Dissolving Philosophy 17: 477–517. ———. 1996. Inclusive fitness arguments in the group selection controversy. Advances in Lehmann L, Keller L, West S, Roze D. 2007. genetic models of behavior. Journal of the Study of Behavior 23: 101–133. Group selection and kin selection: Two Mathematical Biology 34: 654–674. Fehr E, Gintis H. 2008. Human motivation concepts but one process. Proceedings of Uyenoyama MK, Feldman MW. 1980. Theories and social cooperation: Experimental and the National Academy of Sciences 104: of kin and group selection: A population analytical foundations. Annual Review of 6736–6739. genetics approach. Theoretical Population Sociology 33: 43–64. Leigh EG Jr. 1971. Adaptation and Diversity: Biology 17: 380–414. Foster KR. 2008. : Altruism. Natural History and the Mathematics of West SA, El Mouden C, Gardner A. 2011. Sixteen Pages 154–159 in Jorgensen SE, ed. Ency- Evolution. Freeman, Cooper. common misconceptions about the evolu- clopedia of Ecology. Elsevier. ———. 1977. How does selection reconcile tion of cooperation in humans. Evolution Foster KR, Ratnieks FLW. 2005. A new eusocial individual advantage with the good of the and Human Behavior 32: 231–262. vertebrate? Trends in Ecology and Evolution group? Proceedings of the National Academy Wilson DS. 1977. Structured demes and the 20: 363–364. of Sciences 74: 4542–4546. evolution of group-advantageous traits. Frank SA. 1995. Mutual policing and repression Michod RE, Hamilton W. 1980. Coefficients American Naturalist 111: 157–185. of competition in the evolution of coopera- of relatedness in sociobiology. Nature 288: ———. 1980. The Natural Selection of Popula- tive groups. Nature 377: 520–522. 694–697. tions and Communities. Cummings. Gadagkar R. 2010. Sociobiology in turmoil Noble D. 2011. Neo-Darwinism, the modern Wilson DS, Wilson EO. 2007. Rethinking the again. Current Science 99: 1036–1041. synthesis and selfish genes: Are they of use theoretical foundation of sociobiology. Gardner A, Welsh JJ. 2011. A formal theory of in physiology? Journal of Physiology 589: Quarterly Review of Biology 82: 327–348. the selfish gene. Journal of Evolutionary 1007–1015. Wilson DS, Sober E. 1994. Reintroducing group Biology 24: 1801–1813. Nonacs P. 2011. Kinship, greenbeards, and selection to the human behavioral sciences. Gardner A, West SA, Wild G. 2011. The geneti- runaway in the evolution Behavior and Brain Sciences 17: 585–654. cal theory of kin selection. Journal of of social cooperation. Proceedings Wilson EO. 1975. Sociobiology: The New Evolutionary Biology 24: 1020–1043. of the National Academy of Sciences 108: Synthesis. Harvard University Press. Gintis H. 2000. and human 10808–10815. ———. 2008. One giant leap: How insects sociality. Journal of Theoretical Biology 206: Nowak MA, Tarnita CE, Wilson EO. 2010. achieved altruism and colonial life. 169–179. The . Nature 466: BioScience 58: 17–25. ———. 2009. Game Theory Evolving: A 1057–1062. Wilson EO, Hölldobler B. 2005. Eusociality: Problem-Centered Introduction to Modeling Okasha S. 2012. Altruism researchers must Origin and consequences. Proceedings of Strategic Interaction, 2nd ed. Princeton cooperate. Nature 467: 653–655. the National Academy of Sciences 102: University Press. Pagano U. 2012. Love, war and cultures: An 13367–13371. Gintis H, van Schaik CP. 2012. Zoon politicon: institutional approach to . The evolutionary roots of human hyper- Journal of Bioeconomics. Forthcoming. cognition. In Christiansen M, Richerson Queller DC. 1992. A general model for kin selec- HERBERT GINTIS P, eds. . MIT Press. tion. Evolution 46: 376–380. Herbert Gintis ([email protected]) Forthcoming. Queller DC, Strassmann JE. 1998. Kin selection is a professor of at Central Grafen A. 1984. Natural selection, kin selection, and social insects. BioScience 48: 165–175. and group selection. Pages 62–84 in Krebs Reeve HK, Keller L. 1997. Reproductive bribing European University, in Budapest; and JR, Davies NB, eds. Behavioural Ecology: An and policing as evolutionary mechanisms a visiting professor at the University Evolutionary Approach Sinauer. for the suppression of within-group selfish- of Siena, in Siena, Italy. He is also ———. 1999. Formal Darwinism, the individual- ness. American Naturalist 150 (suppl. 1): an external professor at the Santa Fe as-maximizing-agent analogy and bet-hedging. S42–S58. Institute, in Santa Fe, New Mexico.

www.biosciencemag.org November 2012 / Vol. 62 No. 11 • BioScience 991