Following the previous post summarizing Gould (1980), I decided I should probably also cover Gould (1982). As you may recall, the point of the last review was to analyze the claim that Gould called for the overthrow of the Modern Synthesis or linked punctuated equilibria with saltationist mechanisms. We may ask the same things about the paper from 1982, though I hope that such will not be the only use of this post.
To begin with, here is the summary:
The essence of Darwinism lies in the claim that natural selection is a creative force, and in the reductionist assertion that selection upon individual organisms is the locus of evolutionary change. Critiques of adaptationism and gradualism call into doubt the traditional consequences of the argument for creativity, while a concept of hierarchy, with selection acting upon such higher-level “individuals” as demes and species, challenges the reductionist claim. An expanded hierarchical theory would not be Darwinism, as strictly defined, but it would capture, in abstract form, the fundamental feature of Darwin’s vision–direction of evolution by selection at each level.
So, the point, more or less as with Gould (1980), is to question what he views as the exclusivity of adaptationism, gradualism, and reductionism inherent in the Modern Synthesis and to make a claim that an expanded, hierarchical theoretical approach is required.
After a Gouldian (i.e., somewhat esoteric and not terribly relevant, though brief) introduction, the paper continues through eight different sections.
What is Darwinism?
The meaning and appropriateness of the term “Darwinism” remain the subject of debate. I side with Scott and Branch (2009), who argue that the term is misleading and easily misused, and should never be taken as synonymous with “modern evolutionary theory”. Others disagree, especially those who believe that not that much has changed since the Modern Synthesis — nor needs to. (Creationists are particularly enamored with this term because it allows them to portray evolution as an ideology based on the views of a single individual — another very good reason to avoid endorsing it if its use provides no countervailing benefit). If anything, the term should be used only in a specific context, namely in reference to Darwin’s own views. This is how many historians use the term. However, even this narrow usage is not without challenges. As Gould notes:
Although “Darwinism” has often been equated with evolution itself in popular literature, the term should be restricted to the body of thought allied with Darwin’s own theory of mechanism, his second goal. This decision does not provide an unambiguous definition, if only because Darwin himself was a pluralist who granted pride of place to natural selection, but also advocated an important role for Lamarckian and other nonselectionist factors.
Gould then defines “Darwinism”, which he will argue is incomplete and requires “expansion”, as follows:
If we agree, as our century generally has, that “Darwinism” should be restricted to the world view encompassed by the theory of natural selection itself, the problem of definition is still not easily resolved. Darwinism must be more than the bare bones of the mechanics: the principles of superfecundity and inherited variation, and the deduction of natural selection therefrom. It must, fundamentally, make a claim for wide scope and dominant frequency; natural selection must represent the primary directing force of evolutionary change.
Gould suggests that there are two central claims in this view: 1) Natural selection is a primarily creative process, and 2) selection occurs at the level of organisms within populations. He spends time discussing them in sequence.
1) The creativity of natural selection. Darwinians cannot simply claim that natural selection operates since everyone, including Paley and the natural theologians, advocated selection as a device for removing unfit individuals at both extremes and preserving, intact and forever, the created type. The essence of Darwinism lies in a claim that natural selection is the primary directing force of evolution, in that it creates fitter phenotypes by differentially preserving, generation by generation, the best adapted organisms from a pool of random variants that supply raw material only, not direction itself. Natural selection is a creator; it builds adaptation step by step.
Thus defined, Gould then points out what he sees as three assumptions made by Darwinism in terms of the nature of variation:
(i) It must be copious since selection makes nothing directly and requires a large pool of raw material. (ii) It must be small in scope. If new species characteristically arise all at once, then the fit are formed by the process of variation itself, and natural selection only plays the negative role of executioner for the unfit. True saltationist theories have always been considered anti-Darwinian on this basis. (iii) It must be undirected. If new environments can elicit heritable, adaptive variation, then creativity lies in the process of variation, and selection only eliminates the unfit. Lamarckism is an anti-Darwinian theory because it advocates directed variation; organisms perceive felt needs, adapt their bodies accordingly, and pass these modifications directly to offspring. [Emphasis added].
He also points out two additional points that he considers part of classical Darwinian evolution, though these “are not absolute prerequisites or necessary deductive consequences”:
(i) Gradualism. If creativity resides in a step-by-step process of selection from a pool of random variants, then evolutionary change must be dominantly continuous and descendants must be linked to ancestors by a long chain of smoothly intermediate phenotypes. … (ii) The adaptationist program. If selection becomes creative by superintending, generation by generation, the continuous incorporation of favorable variation into altered forms, then evolutionary change must be fundamentally adaptive. If evolution were saltational, or driven by internally generated biases in the direction of variation, adaptation would not be a necessary attribute of evolutionary change.
So, here again Gould mentions saltationism, specifically as a challenge to adaptationism (and, obviously, gradualism). There is no doubt that he was interested in this as a possible mechanism, though in 1980 this was largely in reference to the origin of basic forms of new innovations — for example, the origin of jaws from gill arches through developmental mutations. He does not go into details at this point in the 1982 paper, but there is no mention that saltationism is part of punctuated equilibria here.
As in 1980, though much more briefly, he points out that proponents of narrow-sense Darwinism do not rule out mechanisms other than gradual adaptation but that they relegate these to exceptions of minor importance.
Then on to the second central claim:
2) Selection operates through the differential reproductive success of individual organisms (the “struggle for existence” in Darwin’s terminology). Selection is an interaction among individuals; there are no higher-order laws in nature, no statements about the “good” of species or ecosystems. If species survive longer, or if ecosystems appear to display harmony and balance, these features arise as a by-product of selection among individuals for reproductive success.
As a primary consequence, this focus upon individual organisms leads to reductionism, not to ultimate atoms and molecules of course, but of higher-order, or macroevolutionary, processes to the accumulated struggles of individuals. Extrapolationism is the other side of the same coin–the claim that natural selection within local populations is the source of all important evolutionary change.
To summarize, Gould defines a specific intepretation of evolution based the primacy of natural selection, and identifies what he considers its main claims and assumptions. These include the creative role of selection; the occurrence of ubiquitous, undirected, and incremental variation; gradual transformation over time, the main outcome of which is adaptation; and the emphasis on organisms in populations as the only important level of selection.
Darwinism and the Modern Synthesis
This section recounts Gould’s view that the Modern Synthesis of the 1930s and 1940s (which united Mendelian inheritance and Darwinian natural selection) began as relatively pluralistic but later hardened around an adaptationist core (see Gould 1983 for a full discussion). In this paper, as in 1980, he quotes a summary of the synthesis by Mayr (this time Mayr 1980):
The term “evolutionary synthesis” was introduced by Julian Huxley … to designate the general acceptance of two conclusions: gradual evolution can be explained in terms of small genetic changes (“mutations”) and recombination, and the ordering of this genetic variation by natural selection; and the observed evolutionary phenomena, particularly macroevolutionary processes and speciation, can be explained in a manner that is consistent with the known genetic mechanisms. [Mayr 1980, p.1].
I might interject another (in my opinion, remarkable) quote of Mayr’s in the preface to the 1998 reprint of the same book:
In a way, the synthesis was nothing but a confirmation of Darwin’s original theory, even though Darwin had published prior to the development of genetics and cytology, and had been forced to treat the origin of variation as a black box. His basic theory, that evolutionary change is due to the combination of variation and selection, was, however, completely sound and is daily confirmed by every evolutionist. This basic principle of Darwinism is as solid at the end of the twentieth century as it was in 1850, untouched by all the advances made by evolutionary biology and other branches of biology in the intervening years. [Mayr 1998, p.xiii].
I say “remarkable” because it is indeed astounding that Darwin’s basic idea of natural selection has survived so well in the face of 150 years of research. With that I certainly agree. However, it would also be remarkable if, by this paragraph, Mayr meant to claim that not much else has been added to evolutionary theory beyond natural selection. (I notice that several years ago I wrote in the margin beside this quote, simply, “Wow”).
Back to Gould. After quoting Mayr’s assessment, he argues that he is not erecting a straw man and that it actually does represent the main thrust of the synthesis. Also of note, he argues against attempts to re-brand the synthesis to incorporate new findings such that they are not seen as challenges:
The modern synthesis has sometimes been so broadly construed, usually by defenders who wish to see it as fully adequate to meet and encompass current critiques, that it loses all meaning by including everything.
Obviously, defining what the Modern Synthesis actually is will be important in current discussions about whether it needs expanding. Specifically, it is important to determine whether the “Modern Synthesis” has already been transforming gradually to incorporate all new discoveries — and if so, whether or at what point it no longer warrants description by the same name. Even anagenetic change can, at some point, be said to have led to a novel species.
What is Happening to Darwinism
This is a brief section that lays out the crux of the claim being made. Here it is in full.
Current critics of Darwinism and the modern synthesis are proposing a good deal more than a comfortable extension of the theory, but much less than a revolution. In my partisan view, neither of Darwinism’s two central themes will survive in their strict formulation; in that sense, “the modern synthesis, as an exclusive proposition, has broken down on both of its fundamental claims”. However, I believe that a restructured evolutionary theory will embody the essence of the Darwinian argument in a more abstract, and hierarchically extended form. The modern synthesis is incomplete, not incorrect. [Emphasis added].
Critique of Creativity: Gradualism
In this section, Gould clearly identifies the target of his criticism — it is not natural selection, but exclusive assumptions regarding gradualism and adaptationism:
At issue is not the general idea that natural selection can act as a creative force; the basic argument, in principle, is a sound one. Primary doubts center on the subsidiary claims–gradualism and the adaptationist program. If most evolutionary changes, particularly large-scale trends, include major nonadaptive components as primary directing or channeling features, and if they proceed more in an episodic than a smoothly continuous fashion, then we inhabit a different world from the one Darwin envisaged.
Critiques of gradualist thought proceed on different levels and have different import, but none are fundamentally opposed to natural selection. They are therefore not directed against the heart of Darwinian theory, but against a fundamental subsidiary aspect of Darwin’s own world view–one that he consistently conflated with natural selection, as in the following famous passage: “If it could be demonstrated that any complex organ existed, which could not possibly have been formed by numerous, successive, slight modifications, my theory would absolutely break down”.
He then brings up saltationism for the second time:
At the levels of microevolution and speciation, the extreme saltationist claim that new species arise all at once, fully formed, by a fortunate macromutation would be anti-Darwinian, but no serious thinker now advances such a view, and neither did Richard Goldschmidt, the last major scholar to whom such an opinion is often attributed. Legitimate claims range from the saltational origin of key features by developmental shifts of dissociable segments of ontogeny to the origin of reproductive isolation (speciation) by major and rapidly incorporated genetic changes that precede the acquisition of adaptive, phenotypic differences. [Emphasis added].
Once again, Gould does not claim radical macromutations cause speciation (let alone that they are the mechanism of punctuated equilibria). As in 1980, he cites “saltationist” processes in two particular contexts: 1) developmental mutations of large impact that are involved in the emergence of new innovations, 2) chromosomal mutations causing reproductive isolation.
He goes on:
Are such styles of evolution anti-Darwinian? What can one say except “yes and no.” They do not deny a creative role to natural selection, but neither do they embody the constant superintending of each event, or the step-by-step construction of each major feature, that traditional views about natural selection have advocated. If new Bauplane often arise in an adaptive cascade following the saltational origin of a key feature, then part of the process is sequential and adaptive, and therefore Darwinian; but the initial step is not, since selection does not play a creative role in building the key feature. If reproductive isolation often precedes adaptation, then a major aspect of speciation is Darwinian (for the new species will not prosper unless it builds distinctive adaptations in the sequential mode), but its initiation, including the defining feature of reproductive isolation, is not.
The main point here is that “saltationism” (in the two contexts he invokes it) undermines standard interpretations because it allows non-adaptive changes to play a major role before adaptive processes take over.
Punctuated equilibria makes its first appearance in the next section. As with the 1980 article, it is mentioned in the context of discussions about macroevolutionary trends produced by species selection. Here is the full section:
At the macroevolutionary level of trends, the theory of punctuated equilibrium proposes that established species generally do not change substantially in phenotype over a lifetime that may encompass many million years (stasis), and that most evolutionary change is concentrated in geologically instantaneous events of branching speciation. These geological instants, resolvable in favorable stratigraphic circumstances (so that the theory can be tested for its proposed punctuations as well as for its evident periods of stasis), represent amounts of microevolutionary time fully consistent with orthodox views about speciation. Indeed, Eldredge and I originally proposed punctuated equilibrium as the expected geological consequence of Mayr’s theory of peripatric speciation. The non-Darwinian implications of punctuated equilibrium lie in its suggestions for the explanation of evolutionary trends (see below), not in the tempo of individual speciation events. Although punctuated equilibrium is a theory for a higher level of evolutionary change, and must therefore be agnostic with respect to the role of natural selection in speciation, the world that it proposes is quite different from that traditionally viewed by paleontologists (and by Darwin himself) as the proper geological extension of Darwinism.
Gould concludes this section by pointing out that the debate between gradualism and punctuationism is not about natural selection, but about the larger patterns of change. Specifically, in admittedly “ridiculously oversimplifed” terms, whether change is “constant” or “difficult” as a phenomenon.
Critique of Creativity: Adaptation
The message of this part of the essay will be familiar to Gould’s readers: not everything is an adaptation. (If I may, I would argue that “Not everything is an adaptation” should be one of the most commonly emphasized lines in evolution courses, up there with “Organisms do not evolve, populations do”). Gould cites neutral molecular evolution and developmental constraints as counterpoints to adaptationism. He also notes that current function may not provide much information regarding historical origin, for example because of co-option. He recognizes that these can be interpreted in adaptationist terms, for example if past adaptations determine constraints and if co-option merely involves a shift from one adaptative function to another. However, as Gould and Vrba (1982) subsequently pointed out (he cites that paper here as in press), non-adaptive features can also be exapted. As usual, it is a question of relative frequency. As he says:
Evolutionists admit, of course, that all selection yields by-products and incidental consequences, but we tend to think of these nonadaptations as a sort of evolutionary frill, a set of small and incidental modifications with no major consequences. I dispute this assessment and claim that the pool of nonadaptations must be far greater in extent than the direct adaptations that engender them.
Gould closes this section by repeating that he is not challenging natural selection, only strict adaptationism:
I do not claim that a new force of evolutionary change has been discovered. Selection may supply all immediate direction, but if highly constraining channels are built of nonadaptations, and if evolutionary versatility resides primarily in the nature and extent of nonadaptive pools, then “internal” factors of organic design are an equal partner with selection. We say that mutation is the ultimate source of variation, yet we grant a fundamental role to recombination and the evolution of sexuality–often as a prerequisite to multicellularity, the Cambrian explosion and, ultimately, us. Likewise, selection may be the ultimate source of evolutionary change, but most actual events may owe more of their shape to its nonadaptive sequelae.
Is Evolution a Product of Selection Among Individuals?
As noted in the abstract, one of Gould’s major arguments is that evolutionary theory should undergo an expansion to include a hierarchical outlook. This includes natural selection as a primary mechanism, but one that takes place at several levels:
I believe that the traditional Darwinian focus on individual bodies, and the attendant reductionist account of macroevolution, will be supplanted by a hierarchical approach recognizing legitimate Darwinian individuals at several levels of a structural hierarchy, including genes, bodies, demes, species, and clades.
He briefly mentions the idea of “selfish DNA”, which provides a convincing example of selection within genomes. I agree (Gregory 2004, 2005). Species selection is more complex, but one of the basic requirements is that it must be possible to view species as “individuals” with births, lifespans, differential reproduction, and deaths. It is here that punctuated equilibria comes in, because it provides species with a comparatively short “gestation” and birth, a potentially long lifespan of relative stasis, differential production of daughter species by cladogenesis, and finally extinction. As Gould states:
If new species usually arose by the smooth transformation of an entire ancestral species, and then changed continuously toward a descendant form, they would lack the stability and coherence required for defining evolutionary individuals. The theory of punctuated equilibrium allows us to individuate species in both time and space; this property (rather than the debate about evolutionary tempo) may emerge as its primary contribution to evolutionary theory.
He then points out that having the properties of “individuals” does not guarantee species selection. Indeed, there was substantial debate among macroevolutionary theorists about what would count as species selection and what could be reduced to organism-level selection (e.g., Vrba 1983, 1989; Lloyd and Gould 1993). Nevertheless, he argues:
We need not, however, confine ourselves to the simple fact of individuation as an argument against Darwinian reductionism. For the strong claim that higher-level individuals act as units of selection in their own right can often be made. Many evolutionary trends, for example, are driven by differential frequency of speciation (the analog of birth) rather than by differential extinction (the more usual style of selection by death). Features that enhance the frequency of speciation are often properties of populations, not of individual organisms, for example, dependence of dispersal (and resultant possibilities for isolation and speciation) on size and density of populations.
There has been quite a bit written about population- and species-level properties, including their heritability, since this paper (e.g., Jablonski 1987; Hunt et al. 2005). Still, the main point is clear enough. There are also shades of the distinction between “sorting” (a pattern of differential survival and reproduction) and “selection” (one cause of sorting), which was introduced more explicitly later (Gould and Vrba 1986). Finally, Gould responds to some common criticisms of species selection as an important phenomenon.
Evolutionary Pattern by Interaction Between Levels
Gould notes that a hierarchical approach would not imply that all levels are disconnected from one another. Rather, interactions among levels might be responsible for some of the patterns that are seen:
The hierarchical model, with its assertion that selection works simultaneously and differently upon individuals at a variety of levels, suggests a revised interpretation for many phenomena that have puzzled people where they implicitly assumed causation by selection upon organisms. In particular, it suggests that negative interaction between levels might be an important principle in maintaining stability or holding rates of change within reasonable bounds.
He once again mentions “selfish DNA”, which shows interaction between intragenomic and organism-level selection. He also brings up “overspecialization”, in which adaptation that results from organism-level selection can lead to shorter species longevity. (For a recent example, see Van Valkenburgh et al. 2004). In addition to antagonism between levels, Gould proposes that sometimes selection on more than one level can lead to, and enhance, the same result. Finally, he suggests that sometimes a feature can be subject to selection at a higher level that is not impeded at the lower level, thereby making it widespread among extant lineages (such as sexual reproduction, which delays extinction and accelerates speciation).
We live in a world with reductionist traditions, and do not react comfortably to notions of hierarchy. Hierarchical theories permit us to retain the value of traditional ideas, while adding substantially to them. They traffic in accretion, not substitution. If we abandoned the “either-or” mentality that has characterized arguments about units of selection, we would not only reduce fruitless and often acrimonious debate, but we would also gain a deeper understanding of nature’s complexity through the concept of hierarchy.
A Higher Darwinism?
The final section of the essay highlights Gould’s notion of a hierarchical theory, which he says is neither a minor extension of standard theory nor a revolution. This is because it is not a simple addition to selection at the organism level, but neither does it invoke any new general mechanisms. It is a matter expanding the application of well-understood processes:
What would a fully elaborated, hierarchically based evolutionary theory be called? It would neither be Darwinism, as usually understood, nor a smoothly continuous extension of Darwinism, for it violates directly the fundamental reductionist tradition embodied in Darwin’s focus on organisms as units of selection.
… the hierarchically based theory would not be Darwinism as traditionally conceived; it would be both a richer and a different theory. But it would embody, in abstract form, the essence of Darwin’s argument expanded to work at each level. Each level generates variation among its individuals; evolution occurs at each level by a sorting out among individuals, with differential success of some and their progeny. The hierarchical theory would therefore represent a kind of “higher Darwinism,” with the substance of a claim for reduction to organisms lost, but the domain of the abstract “selectionist” style of argument extended.
Moreover, selection will work differently on the objects of diverse levels. The phenomena of one level have analogs on others, but not identical operation. For example, we usually deny the effectiveness of mutation pressure at the level of organisms. Populations contain so many individuals that small biases in mutation rate can rarely establish a feature if it is under selection at all. But the analog of mutation pressure at the species level, directed speciation (directional bias toward certain phenotypes in derived species), may be a powerful agent of evolutionary trends (as a macroevolutionary alternative to species selection). Directed speciation can be effective (where mutation pressure is not) for two reasons: first, because its effects are not so easily swamped (given the restricted number of species within a clade) by differential extinction; second, because such phenomena as ontogenetic channeling in phyletic size increase suggest that biases in the production of species may be more prevalent than biases in the genesis of mutations.
Each level must be approached on its own, and appreciated for the special emphasis it places upon common phenomena, but the selectionist style of argument regulates all levels and the Darwinian vision is extended and generalized, not defeated, even though Darwinism, strictly constructed, may be superseded.
He closes by quoting and praising Darwin.
So, what can we take away from the essay? In my reading, Gould is arguing for a revision to what he considers a narrow view of Darwinian processes. There is no call for revolution, though his hierarchical view is more than a simple tweaking of the synthesis.
We are in a similar position today, I think. There are many mechanisms that extend beyond the borders of the theory developed 70 years ago. However, the basics of the synthesis are solid.
The question is whether these new phenomena can be included in something that would still reasonably be called the “Modern Synthesis”, or whether we should recognize another major transition (and probably not the final one) in the evolution of evolutionary theory. Like Gould, I believe most who favour the idea of an “Extended Synthesis” view it as a major step — but not a giant leap.
Scott, E.C. and G. Branch (2009). Don’t call it “Darwinism”. Evolution: Education and Outreach 2: 90-94.
Gould, S.J. (1980). Is a new and general theory of evolution emerging? Paleobiology 6: 119-130.
Gould, S.J. (1982). Darwinism and the expansion of evolutionary theory. Science 216: 380-387.
Gould, S. J. (1983). The hardening of the modern synthesis. In Dimensions of Darwinism , ed. by M. Grene, Cambridge University Press, Cambridge, pp. 71-93.
Gould, S.J. and E. Vrba (1982). Exaptation–a missing term in the science of form. Paleobiology 8: 4-15.
Gregory, T.R. (2004). Macroevolution, hierarchy theory, and the C-value enigma. Paleobiology 30: 179-202.
Gregory, T.R. (2005). Macroevolution and the genome. In The Evolution of the Genome ed. by T.R. Gregory, Elsevier, San Diego, pp. 679-729.
Hunt, G., K. Roy, and D. Jablonski (2005). Heritability of geographic range sizes revisited. American Naturalist 166: 129-135.
Jablonski, D. (1987). Heritability at the species level: analysis of geographic ranges of Cretaceous mollusks. Science 238: 360-363.
Lloyd, E.A. and Gould, S.J. (1993). Species selection on variability. Proceedings of the National Academy of Sciences USA 90: 595-599.
Mayr, E. (1980). Prologue: Some thoughts on the history of the evolutionary synthesis. In The Evolutionary Synthesis, ed. by E. Mayr and W.B. Provine. Harvard University Press, Cambridge, pp. 1-48.
Mayr, E. (1998). Preface, 1998. In The Evolutionary Synthesis, ed. by E. Mayr and W.B. Provine. Harvard University Press, Cambridge, pp. ix-xiii.
Van Valkenburgh, B., X. Wang, and J. Damuth (2004). Cope’s Rule, hypercarnivory, and extinction in North American canids. Science 306: 101-104.
Vrba, E.S. (1983). Macroevolutionary trends: new perspectives on the roles of adaptation and incidental effect. Science 221: 387-389.
Vrba, E.S. (1989). Levels of selection and sorting with special reference to the species level. Oxford Surveys in Evolutionary Biology 6: 111-168.
Vrba, E.S. and Gould, S.J. (1986). The hierarchical expansion of sorting and selection: sorting and selection cannot be equated. Paleobiology 12: 217-228.