STEPHEN  JAY  GOULD’S  FINAL  VIEW  OF  EVOLUTION

 A review of “The Structure of Evolutionary Theory” by Stephen Jay Gould, (The Belknap Press, Cambridge, MA: Harvard University Press, 2002), Quarterly Review of Biology, Volume 78, Number 4, December, 2004.

By: William F. Zimmerman, Department of Biology, Amherst College, Amherst, MA 01002, email:  wfzimmerman@amherst.edu

Despite its size and title, this book, written over a period of twenty years by the most widely read popularizer of evolutionary biology, is not a wide-ranging discussion of the logic and history of evolutionary explanations in the life sciences. Rather---and not surprisingly for those who know the writings and persona of Stephen Gould--- it is a narrow account of what its author thinks is wrong or incomplete in Darwinian evolutionary theory (then, since and now); how he and a few like-minded colleagues have significantly revised and “plurified” it; how many of his critics have been some combination of wrong, hypocritical, unprofessionally unfair and motivated by envy; and how, in spite of the criticism he has endured, his ideas have triumphed and earned him a place in contemporary culture and in the history evolutionary thought. Near the beginning of his 1343 page monstergraph Gould did describe it as “partisan” (p.136) and himself as “…the most arrogant of literati..” (p.47), but neither of these warnings quite prepared me for the magisterial way in which they are later justified. Indeed, calling his book “partisan” is one of the few understatements in it: about ten percent of the approximately 1,000 references in the bibliography are to articles or books he authored alone; there are some 50 references in the text to 5, mostly popular books by his best-published critic and therefore favorite nemesis, Richard Dawkins; there is one reference in the text to the work of William Hamilton, and there is not one reference anywhere to Robert Trivers. And yes, Gould’s membership among the “arrogant literati” is assured by this book: the writing is sometimes so verbose, convoluted, and digressive that sentences have to be re-read in order to understand their content (or lack thereof). The following sentence could be used in a writing course as an example of how to kill a reader’s interest:

“In further extensions, macroevolution itself must be reconfigured in speciational terms, with attendant implications for a wide range of phenomena, including Cope’s rule (structurally ordained biases away from a lower size limit occupied by founding members of the clade, rather than adaptive anagenesis toward organismal benefits of large size), living fossils (members of clades with persistently minimal rates of speciation, and therefore no capacity for ever generating much change in a speciational scheme, rather than forms that are either depauperate of variation, or have occupied morphological optima for untold ages), and the reinterpretation of cladal trends long interpreted as triumphs of progressive evolution (and now reevaluated in terms of variational range in species numbers, rather than vectors of mean morphology across all species at any time---leading, for example, to a recognition that modern horses represent the single surviving twig of a once luxurious, and now depleted, clade, and not the apex of a continuously progressing trend).(p. 78)

In the introductory chapter, Gould uses an illustration of a fossil coral (p.18) to symbolize and introduce the interwoven themes of the book, themes that have pre-occupied his professional thinking for some 30 years. The supportive trunk of the coral is Darwin’s theory of evolution by natural selection, and the three branches or “legs” are the principal corollaries of the theory that Gould was intent on revising, each labeled with a single word: (1) “agency”, that natural selection acts on small, randomly-arising, hereditary differences between individuals that enable some of them to have more offspring than others, (2) “efficacy”, that as a result of this process heritable differences accumulate in populations over time, thus leading to changes in, or losses of adaptations and to the production of new species, and (3) “scope”, that this processes is gradual, has occurred over immense periods of time, and is responsible for the adaptive design, history and variety of living things.

        The first section of the book, some 400 pages, provides historical background to these ideas---how Darwin formulated and documented them, the influence of  predecessors and contemporaries on his thinking, and his influence on that of his intellectual successors. Particularly interesting is the discussion of the influence of Adam Smith on Darwin, and altogether I found many incentives to read further in the sources cited.

The second section of the book, Towards a Revised and Expanded Evolutionary Theory, starts with a chapter on the “agency” leg of Darwinian theory, that is the level(s) at which natural selection acts: “Species as Individuals in the Hierarchical Theory of Selection”. Here Gould argued for an ‘interactor” conception of natural selection, according to which it can act simultaneously at several hierarchical levelscell lineages, genes, organisms, demes, species and clades. Here is the argument for species as “agents” of selection:

“ In particular, and continuing to use species as a “type” example of individuality at higher levels, all evolutionary criteria apply to the species as a basic unit of macroevolution. Species have children by branching (in our professional jargon, we even engender these offspring as “daughter” species). Speciation surely obeys principles of heredity, for daughters, by strong constraints of homology, originate with phenotypes and genotypes closer to those of their parent than to any other species of a collateral lineage. Species certainly vary, for the defining property of reproductive isolation demands genetic differentiation from parents and collateral relatives. Finally, species interact with the environment in a causal way that can influence rates of birth (speciation) and death (extinction).” (p.612)

 Although 148 pages are devoted to defining and defending this idea there is no discussion of why it is not subject to the same criticisms of group selection spelled out 37 years ago by George Williams in his book “Adaptation and Natural Selection” (Princeton University Press). One of those criticisms is that the efficacy with which natural selection differentially sorts replicating units is directly proportional to their generation time (faster replication, more frequent selection events) and to their fidelity of replication (the more accurate it is, the more easily selection gets a hold of differences between them). Comparison of genes and species with regard to generation time and replication fidelity makes it clear that the influence of species selection on evolution is likely to have been insignificant.

A second difficulty with species selection is the lack of evidence for any differences in adaptations between species or between any higher taxa that require explanation by processes other than differential change in the frequency of alternative alleles at single gene loci, due their effects on the reproduction of individuals carrying them. This is what “natural selection” is generally understood to mean, and if similar processes occur at the level of species we would expect to find adaptations in them that are exclusively designed for out-reproducing other species and that are separate from and additional to the adaptations engendered by natural selection acting on individuals and genes. I found no convincing evidence for such a process, or for such adaptations in Gould´s discussion and concluded that “species selection” is a confused reification of natural selection acting on individuals. That is, some kinds and lineages of organisms have formed species or gone extinct more readily than others (differential birth and extinction of species, due to competition and/or catastrophes), and this has resulted in the survival of some adaptations instead of others in the long term. But those adaptations arose in the first place, and may have later disappeared, because of generation-to-generation interactions between individuals and their biotic and physical environments, not because of interactions between species as discrete and bounded units.

As a hypothetical example of species selection Gould imagined a pond containing a dominant species of fish whose gills function optimally in well-aerated water but vary little. A second, less abundant, “middling” species has inferior but more variable gills, so that when the pond is reduced by drought the fish with “optimal” gills goes extinct, but some individuals of the middling species, able to breathe in stagnant water, survive and repopulate the pond when it refills. Here is Gould’s take on what happened:

“Can we explain the persistence of the middling species, and the death of the optimal form, only by organism selection? I don’t think so. The middling species survives, in large part, as a result of the greater variability that allowed some members to hunker down in the muddy pools..... The middling species survived qua species because the gills varied among its parts (organisms), not because all its members gained advantage when the environment changed. ......We may represent this story at the organismal level by discussing the gills of the few middling fishes that carried the species through the crisis. But the middling species prevailed by species selection on variability---for this greater variability imparted an emergent fitness to the interaction of the species with the changed environment.” (p.666).


Gould´s thinking on the middling species is muddled. There is an alternative explanation that avoids this group selectionist fallacy: just as the gills that vary little but work well in the first species was an individual adaptation to consistently well-aerated water, so more variable, but less efficient gills can be explained as an individual adaptation (based on a genetic/developmental polymorphism) that enables some fish to live in marginal parts of the pond and some to survive extreme reduction of the pond by drought.

Later in this chapter Gould inveighed against the view that genes are the units of selection:

“I shall show in this section that, while genes may be appropriately designated as fundamental replicators (under a defendable but nonexclusive strategy of research), replicators simply aren’t units of selection or, for that matter, causal agents at all under our usual notions of mechanism in science. The misidentification of replicators as causal agents of selection---the foundation of the gene-centered approach--- rests upon a logical error best characterized as a confusion of bookkeeping with causality.”

        I know of no writings in which the evolutionary biologists to whom he attributed this “gene centered” view argue that genes directly cause their own differential reproduction; it would be like contending that rain drops are “causal agents” in their fall to the ground. Saying that genes are units of selection amounts to no more than saying that that changes in the kind of information they contain has to be the long term product or effect of selection processes. Most evolutionary biologists take an “interactionist” view of how genetic information comes to be differentially replicated: genes participate in a hierarchy of interactions through their cellular environments with other genes, other cells and the outside environment---that result in the development of individuals, some of which leave more descendants than others because of the ways in which the genes they are carrying effects and affects their interactions with each other and the environment. And low and behold something like this is the interactionist view championed by Gould after his dismissal of “gene selectionism”! Many examples of this style of arguing are found in the book, and Gould raised it to an art form in his writings as a public intellectual: construct a flimsy, confused, windmill of an argument, attribute it to a fictional tribe of sclerotic, orthodox “Darwinian fundamentalists”, charge forth with much literary fanfare to destroy it, and build something more intellectually and morally acceptable in its place, using what appears to be his own revolutionary thinking, but is actually borrowed from those to whom he attributed the false views in the first place.

        Gould was led to his ideas about species-level selection by Niles Eldredge’s and his proposal, first made more than 30 years ago, that the pattern of evolutionary change shown by the fossil record is most often one of “punctuated equilibria”---short periods of rapid morphological change associated with speciation, followed by long periods of little or no change (stasis)---rather than “phyletic gradualism”, which he characterized as the “Darwinian orthodox” view that morphological change goes on continuously and incrementally within species lineages, and is not tightly linked to speciation events. Thus, “phyletic gradualism” is the second or “scope” leg of orthodox Darwinism that Gould wished to revise, and a book-length chapter (279 pages) with the presumptuous title, “Punctuated Equilibrium and the Validation of Macroevolutionary Theory”, is devoted to it. Some interesting examples of fossil evidence are presented that seem to support the theory, but there is little to no discussion of how many such arguable examples there are and not enough discussion of why punctuated equilibria has generated so much criticism and controversy. One reason is the difficulty of inferring a linkage between rapid evolutionary change and speciation from fossil evidence alone. Even where the record of sedimentary deposition is continuous over the period of time required to measure   change in what appears to be a single phyletic lineage, it is very difficult to conclude that rapid morphological change is associated with speciation. Such change may reflect an entirely phenotypic response (adaptive or not) to changes in the organisms physical, chemical or biotic environment, or it may reflect a sudden immigration of individuals or genes from a distant, previously isolated population of the species that had undergone gradual change over a long period of time.

A related difficulty is that because the linkage between morphological changes that paleontologists must use to infer speciation and the genetic changes that establish reproductive isolation between species varies between organisms it is very difficult to know when, along an axis of morphological change over time, reproductive isolation was completed. Most of the morphological differences between closely related (but reproductively isolated) living species are very small and would not be evident in fossils, a finding that weighs heavily against linkage between speciation and significant morphological change in the first place. Also, rapid morphological change need not be accompanied by speciation: a future paleontologist, happening upon a continuous fossil record of the evolution of domestic dogs from wolves, might define many of them as species and infer punctuated equilibrium, even though they all interbred (at least up to 2003 A.D.).

        Another general reason why the theory of punctuated equilibria has generated criticism is that, contrary to Gould´s implied claims, it is not an original idea and does not represent a significant revision of “orthodox Darwinism”. Evolutionary biologists (including Darwin) have long known from the fossil record, field observations and laboratory experiments that rates of evolution vary widely and do so because environments and/or ecological opportunities can change rapidly. And they have long agreed that some linkage between rates of evolutionary change and rates of speciation is to be expected because the changes in genotype that underlie the morphological divergence of populations will directly or indirectly initiate and reinforce reproductive isolation between them. Finally, as others have pointed out, the morphological rates of evolution Gould calls rapid are only relatively so---hundredths or thousandths of a percent change per generation. “Punctuated equilibrium” could as well be called “punctuated gradualism”.
               
        In the next chapter, Integration and Adaptation (Structure and Function) in Ontogeny and Phylogeny: Historical Constraints and the Evolution of Development,  Gould sought to challenge the “scope” leg of orthodox Darwinism, "…the functionalist attribution of effectively all substantial evolutionary change to natural selection" (p.1027) by showing that "….internal factors [historical constraints] restrict the freedom of natural selection to establish and control the direction of evolutionary change"(p.1028). In spite of all of the subsequent discussion I remain mystified by his conclusion that what he means by “constraints”---evolved properties of developmental systems that limit the expression of genetic variation---can be called a “cause” or “force” of evolution in the same way that natural selection is. Do we say that the absence of geological conditions favorable to mountain building is a "cause" of the evolution of local landscapes in the same sense that volcanism and crustal folding are? Besides, given how hard it is to understand why evolution has taken the paths it has and how little we understand about how structure and form develop, discussion of how developmental systems have constrained evolution from taking the paths that it has not seems at best premature. Gould found the recently discovered homologies between the invertebrate and vertebrate “homeobox” genes involved in body segment and head development to be further evidence for a great antiquity of constraints imposed by developmental mechanisms. It is equally valid to marvel at the extraordinary variety of forms and structures that can be generated by natural selection working with variants of genes that have conserved base sequences, spatial distribution on chromosomes and temporal sequence of expression. If everything from squids to spiders and fruit flies, and from lungfish to bats and elephants can be produced by historically conserved developmental systems, is this evidence of “constraints”, and if so, what could the products of unconstrained developmental systems be like?

           Gould also concluded that because developmental systems set limits on the kinds of genetic variations that can be expressed, the variation presented to selection is not random. He found this contrary to Darwinian orthodoxy: "…for the logic of pure selectionism does presuppose nondirectional variability, and the existence of strongly preferred channels, based on the architecture and history of development, does require an important restructuring (not just a minor nuancing) of Darwinian logic." (p. 1090). I doubt that even a minor nuancing is required. It has been known for some 60 years that mutations are not random chemically: the structure and bonding properties of DNA bases make some kinds of nucleotide substitutions more likely than others. It has also been long known that the structure of the genetic code constrains the effects of nucleotide substitutions by making them more likely to be silent (no change in amino acid) or conservative (a functionally similar amino acid is substituted). And it is not new that developmental systems restrict the kinds of mutations and recombined genes that can be expressed. The only sense in which the non-directionality or randomness of genetic variation is central to Darwinian logic, then and now, is that it is random with respect to what is good or bad for the reproductive success of an individual’s descendants. All of the evidence to date suggests that this is so.

        The discussion of constraints lead Gould, near the end of the book, to revisit the subject of “spandrels”, an architectural term borrowed by him and Richard Lewontin to characterize structures or behaviors of organisms that have no adaptive function, but rather are consequences of the development of structures or behaviors that do. The search for spandrels was, then, part of their critique of the “adaptationist program” of orthodox Darwinism---seeing everything about organisms as a product of natural selection. However, their critique was accompanied by a much larger, rhetorically embellished message---a public proclamation that modern Darwinism was intellectually flawed, corrupt and in dire need of revision. This attack from within on the livelihood of many evolutionary biologists elicited, besides dismay and anger, the reply that is still pertinent to Gould´s revisit to the whole subject: formulating and testing hypotheses about how natural selection causes the evolution of adaptation is still the central task of evolutionary biology, and understanding how organisms work has to start with understanding what they were designed for. As for his pet spandrels, he finally acknowledged that the female clitoris is not a spandrel of its developmental homolog in males, the penis, but rather has an evolved function in regulating female sexuality. (This is something most of his students of both sexes could have told him, and in any case I always thought the converse “just so” spandrel story was more interesting, maybe even made more sense). He did continue to argue that male nipples are a functionless spandrel of female nipples, and maybe they are. Of course, someone might propose and experimentally test the hypothesis that male nipples have evolved to mark body symmetry and therefore offspring and mate quality. In other words, the subject matter of the new field of research founded by Gould and Lewontin---the study of non-adaptations (spandrels)---will likely turn out to be mostly non-studied adaptations, which always has been the growth industry of evolutionary biology as new theory is developed.

        In thinking about where TSET stands in relation to the totality of Stephen Gould´s work I find it useful to consider that there were several sides to the man as teacher, writer and scholar. One side was the brilliant exponent and raconteur of the fascination of evolution and its history in western thought, and to this he brought erudition, intelligence and masterful writing skills that won him a wide audience of readers and much deserved praise. This is the side of Gould evident in the first section of the book, on the history of Darwinian logic and debate, and historians of evolutionary theory might find this section  alone worth the book’s $30 price.

Another side of Gould was the more academic and scholarly one of historian, critic and reviser of evolutionary theory. Many have argued that in this role he did stir up a great deal of useful debate that lead his peers to critically examine their own ideas when provoked into doing the same for his. This may be true for those specifically interested in macroevolutionary theory, but as a student and teacher of evolutionary explanations across the life sciences during the same time Gould was, I usually found the products of this side of him, including the major part of this book devoted to revising and extending evolutionary theory, to be too exasperating to recommend to anyone else. On the one hand they are replete with interesting insights, connections and turns of phrase, especially the related forays into the history of evolutionary thought. On the other hand, however, they are often confused, ignorant of other relevant areas of evolutionary biology, oblivious to the contributions and criticisms of peers, and just off in a self-centered world of thought in which he constructed fake arguments in order to easily deconstruct them, inflated his minor variations on old ideas by inventing new words for them, and avoided careful, critical analysis of an idea in favor of spinning it into an erudite metaphor, or stuffing everything it reminded him of into a peacock tail sentence the length of paragraphs. Some of the worst of this was his destructive and blatantly political misrepresentations of evolutionary social theory, a field that during his lifetime was established by what are arguably the most important advancements in evolutionary theory since Darwin, but that receives in this book only the space required by him to dismiss its major premises. His public attacks on this work, directed mainly against the explanatory metaphors and excessive claims of popularizers, drew upon such soap box intellectual techniques as quoting out of context, conjuring up right wing political demons and loaded buzzwords to debunk established and testable theory, and much more self-promoting, moralistic posturing than can be laughed off as typical of professors (even at Harvard).

Some might find the first section of The Structure of Evolutionary Theory useful in a graduate or advanced undergraduate seminar on the history of evolutionary theories, but for a similar course on the structure and logic of evolutionary theories I would first assign and discuss the works of the major authors whose thinking Gould found outdated, flawed or incomplete. Or perhaps I would start with two, recently published collections of classic papers that show evolutionary theory at its best, William D. Hamilton’s “Narrow Roads of Gene Land” (Oxford University Press, 2001) and Robert Trivers´ “Natural Selection and Social Theory” (Oxford University Press, 2002). For me, the major part of Gould´s final magnum opus was just too long and too much of a struggle.