Book Fails to Advance Main Theses
Reviewed in the United States on March 25, 2017
It is with some regret that I write a (mostly) negative review of this work by E.O. Wilson, because his earlier writings such as the tour de force "Sociobiology" were among the books that inspired me to pursue a career in biology.
First, let me state the strong points of "The Social Conquest of Earth" - like most of Wilson's books, it is very well written and contains a great deal of carefully researched information, ranging from the author's own expertise in the biology of ants and other social insects to various findings in biological and cultural anthropology. However, much of this consists of ground that was previously covered (albeit in now outdated form) in "On Human Nature" and "Conscilience."
Unfortunately, the main weak points of "The Social Conquest of Earth" are also the book's central theses. One of these points, the claim that human beings are "eusocial," is perhaps a matter of semantics rather than a fatal flaw. Wilson asserts that humans have come to dominate the Earth's macrofauna for the same reason that ants and other social insects have come to dominate the world's microfauna: both are the result of social hierarchy and efficient communication systems/divisions of labor. This general point is indisputable, there is no denying that both humans and ants have achieved unparalleled forms of social integration that allow us (and them) to extract and process resources more efficiently than any other living organisms of similar size and metabolism. However, the definition of eusociality (promoted by Wilson himself in earlier key works like "The Insect Societies" and "Sociobiology") requires a reproductive division of labor, with sterile or nearly sterile worker castes and reproductive castes. This clearly isn't the norm in human societies, since even the most stringent social hierarchies (e.g. caste system in India, social classes in Medieval Europe) do not involve a strict reproductive division of labor. In humans, even the lowest castes and social classes generally have the opportunity to reproduce. The only vaguely analogous examples of reproductive division of labor in human societies may be the role of homosexual berdasche in certain American Indian tribes, but this is far too culture-specific an example to permit generalization as an example of "eusociality" in humans. Perhaps a somewhat closer analogy for strongly hierarchical human societies would be reproductive skew models for nearly eusocial insects such as paper wasps, or closer still, the reproductive skew in pack mammals such as timber wolves, where alpha males and females are the principal, but not sole or permanent, reproductive individuals in the social group. The key to human and ant ecological dominance is not eusociality or (reproductive) division of labor per se, but rather high levels of social integration. Referring to humans as eusocial simply leads to confusion, because it isn't reproductive division of labor per se that leads to social dominance but ergonimic division of labor, which in insects is coupled to reproductive division of labor while in humans is (largely) decoupled.
However, the point about human eusociality is a semantic one that can be excused, in spite of the possible confusion it may cause. The most important issue raised in "The Social Conquest of Earth" is Wilson's critique of kin selection and inclusive fitness models as the theoretical foundation for understanding human (and insect) sociality. It is this issue that distinguishes this book from the author's earlier works, and it is on this issue that the book's merits should stand or fall. In my opinion, the book fails as an effort to refute kin selection and inclusive fitness as explanatory models, for the following reasons:
Wilson contends that kin selection is inadequate as an explanation for the origin of the high degree of sociality seen in humans. Strictly speaking, this it is true that human cooperation extends far beyond kin groups, since human socities exhibit a high level of altruism among non-kin. However, this issue has been addressed extensively by many authors in the past, e.g. theories of reciprocal altruism, psychological theories which posit non-kin groups (religions, nationalities, even sports team loyalties) as emotional surrogates for tribal kin groups, whereby altruism to non-kin is analogous to keeping of pets or adopting children as emotional surrogates for one's own (i.e. as a surrogate emotional outlet for instincts that make Darwinian sense) Wilson discusses altruism among non-kin as unambiguous evidence for group selection and against kin selection without seriously considering, much less refuting, alternative hypotheses.
Furthermore, there exists several decades' worth of literature that extends kin selection/inclusive fitness model to exceptional cases in where the simple additive model presented by Hamilton fail to apply - including strong selection. Specifically, Wilson correctly notes that Hamilton's rule assumes linearity and weak selection, but fails to note that alternative models for strong selection and non-additive effects have been considered in the recent primary literature. Similarly, while it is true that for completely eusocial insects with sterile worker castes it is not meaningful to model inclusive fitness in terms of altruism or parent-offspring conflict (because the worker castes have no potential for reproduction to sacrifice), this point confounds the end result with the evolutionary history that lead to that endpoint. To continue with Wilson's analogy of cells in a multicellular organism, once the germline/soma is in place, it isn't meaningful to speak of standard cost/benefit models among cells in terms of Hamilton's rule, and the same is true for advanced eusocial insects where workers are constutively incapable of reproduction. However, the ancestors of multicellular organisms and the primitively eusocial (and subsocial) ancestors of advanced eusocial insects surely experienced evolutionary scenarios where Hamiltonian cost/benefit analyses and evolutionary game models were quite applicable. This had to occur in order to establish the soma/germline division or the worker/queen castes from precursors where such a reproductive division of labor was only weakly defined. To contend otherwise would be to assert that eusociality arose de novo.
I also find Wilson's contention that group selection is an alternative to kin selection quite puzzling, and I say so as one who is quite sympathetic to hierarchical selection models across all levels (having done some work in this field). Indeed, in "Sociobiology", Wilson, in my opinion correctly, treats kin selection as a special and important case of trait-group selection. This point was perhaps best articulated by his sometime coauthor David Sloan Wilson (no relation, as far as I'm aware). D.S. Wilson has long argued that for kin selection to favor altruistic behaviors, organisms must be partitioned into kin groups, just as other forms of group selection require a mechanism for trait group partitioning in order to establish a covariance between group level fitness (e.g. inclusive fitness for kin) and phenotype. Why E.O. Wilson now treats kin selection as an (allegedly) refuted and discarded alternative to broader models of group selection isn't clear. These broader group selection models may indeed explain such phenomena as co-founding of colonies by unrelated queen wasps and facets of human altruism among non-kin, but these are open questions that require a careful evaluation and refutation of alternative models (including more restrictive kin selection scenarios). Unfortunately, Wilson's book fails to make this case, and the technical paper that Wilson references (Nowak, Tarnita, Wilson 2010 Nature) merely articulates a point made by D.S. Wilson decades earlier: that kin selection can be modeled as a special instance of group selection by using an alternative form of genetic book-keeping (i.e. trait groups rather than Hamilton's inclusive fitness). Nowak et al simply present a more cumbersome, albeit more general, alternative to Hamilton's rule.
There may very well be much to be gained by applying generalized, rigorous models of group selection to our understanding of human and insect sociobilogy, but unfortunately "The Social Conquest of Earth" fails to make a case for doing so. Readers are encouraged to read Wilson's earlier works such as "Sociobiology" in its place, as well as E. Sober and D.S. Wilson's "Unto Others."
23 people found this helpful