American Scientist

THE SUPERORGANISM: The Beauty, Elegance, and Strangeness of Insect Societies. Bert Hölldobler and Edward O. Wilson. xxi + 552 pp. W. W. Norton, 2009. $55.

Bert Hölldobler and Edward O. Wilson, whose book The Ants won a Pulitzer Prize in 1991, have now come out with a magnificent book about social insects in general, The Superorganism. It represents evolutionary natural history at its very best, providing a rich description of how such creatures as ants and bees engage in the struggle for existence. The details are often astonishing, but the real strength of the work lies in how those details are put together. Toward the end of the book, fascinating historical accounts reveal how the various groups of ants originated and how and why they have diversified through time. Such narrative history is possible only because of the contributions of systematic biologists.

Although well written and adorned with many remarkable photographs, the book is really a scholarly one aimed at a broad, but erudite, audience. I would not discourage younger students and nonscientists from reading it, but they should be forewarned that they will encounter daunting technical passages. For although the authors have made an admirable effort to explain the technicalities in plain English, there is no getting away from the fact that they are addressing some of the most profound and difficult questions that have ever challenged evolutionary biologists: What are the basic units into which life is organized? What role do such units play in the economy of nature? How are societies constructed? What do we mean by a society in the first place? At the heart of their analysis is the concept of the superorganism—the idea that a colony of social insects is the equivalent of an organism, made up not of cells and tissues but of “closely cooperating animals.” The superorganism is one step up from the organism in the hierarchy of biological organization.

One of the most striking features of insect societies is that they contain “neuter castes” of organisms that do not reproduce (worker bees, for example). That created a problem for Darwin, who conceptualized his theory of natural selection in terms of one individual outreproducing other members of its species. He solved the problem by saying that it is individual “families” (in this case, individual colonies), not just individual organisms, that reproduce differentially. Darwin treated groups composed of organisms—families, tribes, colonies—as units that get selected. In the case of the neuter castes, he reasoned, it is an advantage to such communities to have sterile members who spend their time and energy working for the prosperity of the colony as a whole rather than bearing offspring. The very Darwinian idea that selection can take place at supraorganismal levels as well as among individual organisms provides much of the philosophical substrate for Hölldobler and Wilson’s book.

What Darwin referred to as the selection of families is one version of what has been called group selection. The term leaves something to be desired, because multicellular organisms are groups of cells, and biological species are groups of organisms. Life exists as a hierarchy of wholes made up of parts. What goes on at any given level is both similar to and different from what goes on at the other levels. Only species speciate, but their multiplication resembles that which goes on at lower levels. Thus we find fission and budding occurring in speciation, just as they do in cell division and in the formation and fragmentation of colonies.

It is a commonplace, alas often neglected, that organisms do not evolve; rather, evolution is something that happens at the populational level. But organisms do have ontogenies. Functional specialization and division of labor between the parts of larger wholes are found at many levels. So if we want to understand the general principles and laws of nature that apply to the living world, it helps to make comparisons between levels, not just within them.

There is an ancient tradition of comparing all sorts of things, including the world as a whole and human society, to organisms. It has played a major role in religious and political belief systems. In objective natural science, however, such comparison should be carried out in a critical fashion. The mere fact that both an atom and a eukaryotic cell have a nucleus is not good evidence for the notion that all matter has life (hylozoism). Nonetheless, analogies—even far-fetched ones—may play a useful, heuristic role in science if they lead us to ask interesting questions. One important question we might ask is whether treating social insects as superorganisms provides us with an appropriate or “natural” classification of life in general.

The rationale for the superorganism view is straightforward enough. A unicellular organism can transform itself into a multicellular one by incorporating additional cells into its body, and a multicellular organism can transform itself into a multiorganismal one by investing itself with additional organisms. (To be consistent, then, perhaps we should call multicellular organisms “supercells.”)

A serious objection to treating societies as organisms in more than just a metaphorical sense is that the organism is the basic unit of physiological autonomy, whether it consists of one cell or many. Furthermore, with minor exceptions, multicellular organisms are genetically homogeneous, whereas insect societies are genetically heterogeneous. This fact turns out to be much more important than was once believed.

In a chapter on the division of labor, Hölldobler and Wilson list a number of “functional parallels” between insect societies and organisms. Some of their analogies are rather far-fetched, and the list itself contains some clumsy logical mistakes. Cells are appropriately compared to colony members, but organs, such as gonads, are compared to castes rather than to the colony members that are instances of those castes. The different castes are like the different types of cells within an organism; to get the real equivalent of an organ, the caste would have to be organized into “teams”—functional units composed of more than one member of the caste. Another item on the list is sensory organs, the functional parallel for which is the “combined sensory apparatus of colony members.” But I cannot find anything equivalent to an eye or an ear in the admirably detailed descriptions that the authors provide. They place the organ systems in the organism in parallel with the corresponding functions that are carried out by the components of the colony. For instance, the equivalent of the nervous system is not a system of parts that are specialized for conduction and processing of information, but rather “communication and interactions among colony members.” The comparison of embryonic growth to sociogenesis (the growth and development of the colony) is fairly successful, because both of these are processes in which new properties are realized. However, that kind of ontogeny might also be compared to economic development. The skin and skeleton are compared to the nest. As an invertebrate comparative anatomist, I am hard pressed to find anything in the nest of an ant that corresponds to the integument of a tapeworm or the coelomic cavity of an echiuran. Such nests may have some walls and buttresses, giving protection and mechanical support here and there, but the edifices seem rather more like the barns and storehouses of a family farm.

However stimulating the exercise of viewing insect societies as superorganisms may be, there are excellent grounds for viewing them as Darwin did: as families. Those grounds are to be found in the science of economics. The economic reason for having more than one organism participate in an enterprise is not far to seek. By combining their resources, a pair of organisms can save on fixed costs. Just as human roommates might share a toilet and a telephone, the foundresses of an insect society might share the entrance to a burrow. But when such an arrangement gets set up, there can be problems with free riders and absconders. Any costs of competing among the members of the group will have to be subtracted from the productivity of the whole and the reproductive success of its parts. The most straightforward way to achieve reproductive success as a component of a superorganism is to be the gonad and have somebody else serve as the kidney. But that raises the question of how the “somatic organs” (that is, the workers) are induced to forgo reproduction. One possibility is kin selection, which means that by aiding relatives they attain reproductive success of a less straightforward kind.

Thirty-five years ago, when I first commented on such matters, a particular version of kin selection was a very popular explanation for the reproductive behavior of social Hymenoptera (the order that includes ants, bees and wasps). These animals are haplodiploid: Unfertilized eggs, which are haploid (having just a single set of chromosomes, all derived from the mother), become males, and fertilized eggs, which are diploid (getting one set of chromosomes from the mother and another from the father), become females. If the mother in a colony mates only once, then the members of the family are all very closely related. That circumstance was thought to further the evolution of self-sacrifice by the members of the colony. To me it was obvious from the outset that something was wrong. Most important, termites (Isoptera), although social, are diploid. Furthermore, most lineages of Hymenoptera are solitary. So haplodiploidy is neither necessary nor sufficient for the evolution of sociality.

Therefore something else has to be involved. One such thing is the principle of parental exploitation, which states that when Mom packs the lunch box, Junior eats what is good for Mom. It involves coercion, and plenty of evidence for coercion is presented in The Superorganism. Another possibility is group selection, which Hölldobler and Wilson now prefer to kin selection, although they seem not to be of one mind about it. Hölldobler has maintained that kin selection and group selection are really the same thing from a different perspective.

But the mere fact that organisms and societies can evolve by various selective mechanisms is not the whole story, for it does not tell us the fundamental reasons that ants and termites live social rather than solitary lives. Those reasons are to be found in economics—in the science that concerns itself with how resources are utilized and allocated. The Superorganism goes a long way toward providing the answers; I was particularly pleased to find the division of labor so often invoked and so richly documented.

The hymenopteran emphasis leads to a certain amount of parochialism, though, and to the neglect of some relevant literature. Adam Smith might well have been cited on the point that the division of labor saves time that might be spent in switching from one task to another. And his idea that the division of labor is limited by the extent of the market might have been applied to discussions of colony size. Furthermore, there are advantages to combining labor as well as to dividing it. Again, Adam Smith drew attention to the combination of teaching and research. Darwin’s ideas about the division of labor owed much to Henri Milne-Edwards, to whom he dedicated one volume of his great monograph on the Cirripedia. In On the Origin of Species, Darwin writes

The advantage of diversification in the inhabitants of the same region is, in fact, the same as that of the physiological division of labour in the organs of the same individual body—a subject well elucidated by Milne Edwards.

This extrapolation from political economy to the economy of nature created the modern science of ecology. I might also mention Babbage’s idea that the division of labor provides opportunities for putting diverse talents to work.

It is a virtue of good scientists to change their minds as knowledge advances. Wilson has changed his mind substantially with respect to both the concept of the superorganism and group selection. In Sociobiology (published in 1975), he predicted that that new discipline would swallow up other disciplines, including economics. The Superorganism makes it abundantly clear that the opposite is happening. Sociobiology is becoming a minor branch of bioeconomics.

Hölldobler and Wilson are admirably restrained about drawing human implications from their work; they merely suggest that some interesting and important lessons might be learned. Given the role that the superorganism notion has played in the literature of political ideology, though, they might have warned their readers not to take such analogies too seriously.

Michael T. Ghiselin is Senior Research Fellow at the California Academy of Sciences. He is the author of Metaphysics and the Origin of Species (State University of New York Press, 1997), and he and his collaborator Guido Cimino are completing a book on the evolution of chemical defense in opisthobranch gastropods.