Revolutions in evolution: Stephen Jay Gould in perspective

Issue: 100

Paul McGarr

‘Perhaps the most important advance in evolutionary theory since Darwin wrote The Origin of Species in 1859’.1 That assessment was given by the eminent, and left wing, biologist Steven Rose speaking at the Marxism 2003 event in London. The work he was referring to was Stephen Jay Gould’s The Structure of Evolutionary Theory, a massive volume of some 1,400 pages published in 2002.2

Gould knew that he was dying as he worked on the book, and died of cancer last year at the age of 60. His legacy is immense. He is best known as a brilliant populariser of science, mainly through his collections of essays. These stemmed from a column in the US magazine Natural History. He started the series in 1974 and for more than 25 years kept up an unbroken monthly sequence, totalling over 300 essays in all.

In them he ranged over issues in evolution, natural history and often way beyond. Sometimes he countered opponents like creationists, sometimes he explained important scientific ideas and debates. Always the essays were beautifully written and accessible while never oversimplifying the science.

Gould drew heavily on history, art, architecture and much else, including baseball which he had a passion for, all of which he wove into his arguments to brilliant effect. Not many people can pull off explaining one of the most important theoretical arguments in evolution by treating readers to a long discussion about the statistics of home runs in US baseball seasons!

This ability to combine serious science with popular imagery was one of the qualities which made Gould the best known scientist in the US in recent years, a figure recognisable enough to appear as a caricature in an episode of The Simpsons.

There is another, and sometimes well deserved, caricature of scientists as ignorant of history and art—and a parallel, even more well founded reverse image too. Gould powerfully challenged that divide in everything he wrote. His essays were collected and published in books from The Panda’s Thumb and Ever Since Darwin to the more recent Bully for Brontosaurus, Dinosaur in a Haystack and I Have Landed. Anyone who has read these collections will, I am sure, agree they are among the best of science writing.3

Stephen Jay Gould was also a politically engaged scientist and made little secret of his generally leftish views. On occasion he was even attacked for being a Marxist. Though that was not a label he usually applied to himself, he is reputed to have had a picture of Lenin above his desk! Gould certainly waged constant war against right wing ideas. Throughout his working life he battled against the powerful forces in the US who push creationist views (that the Biblical account in Genesis was the truth and that evolution is false).4

Another front in Gould’s wars was with the IQ industry—the idea that there is a single measurable thing called intelligence, with the usual corollary that this ‘intelligence’ shows important genetic differences across races and classes. In 1981 Gould wrote the book The Mismeasure of Man which utterly destroyed, with wit and precise detail, the nonsense around IQ testing and the associated arguments about intelligence, race and class. He reissued an updated version in the 1990s to counter the publication of The Bell Curve, a pernicious and racist retreading of right wing and unscientific ideas on ‘intelligence’. Gould’s original book deservedly won one of the most prestigious book awards in the US.

Gould was, though, first and foremost an active and working scientist, a palaeontologist studying the fossil record of life on earth. His special field was as one of the world’s experts on the evolution of land snails in the West Indies. It was his scientific work that saw him become a distinguished professor at Harvard University. He also become the president of the American Society for the Advancement of Science.

As well as countless specialist papers he wrote several full-length scientific books. Some, like Wonderful Life and Life’s Grandeur, were written with the more popular reader in mind. Two others were written more for professional colleagues, the first being Ontogeny and Phylogeny—which analysed the important relationship between the way any individual organism develops throughout its life cycle (ontogeny) and the historical evolution of the species of which the organism is a member (phylogeny).

Gould’s second, and most important, major scientific work is The Structure of Evolutionary Theory. It is long, perhaps too long, and in parts technical as it addresses key debates with professional colleagues. But much of it demands no technical knowledge at all and is a joy to read. Many of the more technical parts are also such that a keen non-specialist can, with some effort, follow his arguments.5

In the course of expounding his argument Gould addresses key issues in what have been dubbed ‘The Darwin Wars’. These are the very public rows between people like Gould and the US Marxist biologist Richard Lewontin on the one hand, and those who have been labelled the Darwinian Fundamentalists, or Ultra-Darwinians, on the other. This group is usually taken to include the British biologists Richard Dawkins, most famous for books like The Selfish Gene and The Blind Watchmaker, and John Maynard Smith. The fundamentalists also usually include US language theorist Stephen Pinker and US philosopher Daniel Dennett, though in reality all these people have very different ideas on many key questions, political as well as biological.6

Making Gould’s arguments accessible

In this article I want to try and sketch out Gould’s key arguments in The Structure of Evolutionary Theory and try to make them accessible, as I believe they deserve to be, to a wider layer of political activists.

I am not a biologist and do not have the expertise to judge whether Gould is right on many issues. As Gould would have been the very first to insist, evidence from the proper scientific study of the real world will ultimately decide the matter. But Gould also rightly insisted that debates in science should not be confined to scientists, and should be open to full and properly informed public debate. In that spirit, and at least having read people putting all the key arguments, I venture my own conclusion. This is that on most key issues Gould will prove to be nearer the truth than his critics and The Structure of Evolutionary Theory will come to be regarded as a landmark in scientific history.

Why should Marxists, socialists and people interested in politics care about these debates in science? First, I can’t imagine anyone interested in the world we live in not wanting to know how that world came into being and has developed. The subject is fascinating and enriching for any rounded understanding of the world.

For Marxists in particular evolutionary theory has always been important, a fact testified to by the excited reaction of Marx himself to the publication of Darwin’s original work. Marxism is a worldview best summed up as historical materialism. If Marxism proper focuses on history and political change, it nevertheless needs to be grounded in a wider set of theories about how the material world, including the natural and biological world, has developed and changed. Marxism is not a substitute for such a theory—being a Marxist will never on its own help you understand evolution or biology. But Marxists have a keen and legitimate interest in all theories of the history and development of the material world.

Second, debates around evolution have always had a sharp political resonance. This was true in Darwin’s day, and has remained so ever since. There were the right wing Social Darwinists of the late 19th century. In the 1930s the key figure in population genetics, R A Fisher, unashamedly pushed for eugenics, or selective breeding of humans. Today we suffer the nonsense of pseudo-sciences such as evolutionary psychology, purporting to explain why human nature is inherently selfish, why women are different (for which read inferior) to men, and so on.

One final point before turning to Gould’s arguments. Like any prolific writer, especially one engaged in often polemical arguments, Gould changes his mind, or puts different emphases at different times. I am basing this article on his final work, which he intended to be a summation of his views. I have no doubt that these views can be contradicted by quotes from earlier works. What that proves, other than that those engaged in real debate modify their views as a result, I do not know.

Darwin’s theory

Gould’s starting point is to lay out the key elements of Charles Darwin’s theory of evolution by natural selection, first published in The Origin of Species in 1859.7 This was, and remains, a revolutionary theory. In its core arguments it remains essentially valid, and central to any scientific worldview today. As the key points of the theory have so often been misunderstood let me sketch them:

In a given environment, all organisms produce more offspring than can survive to then reproduce themselves.

All organisms naturally vary.

Offspring inherit characteristics from their parents and tend to be more like their parents than others.

At least some of the variations in an organism lead to a greater number of its offspring surviving and reproducing relative to the offspring of others.

This ‘natural selection’ means that these particular variations will become more prevalent in the population (as a consequence of the differential survival and laws of inheritance) as they are passed on to future generations.

These are the bones of Darwin’s argument. There are also, though, important conditions for the theory to work, and consequences from it, which Darwin himself made abundantly clear.

Adaptation: The most important consequence is that the variations which are naturally selected are those that give a better chance for surviving and reproducing against the background of a particular environment. This is what Darwin called adaptation, and it is central to evolution. Natural selection sees organisms becoming adapted to the environment they must survive in. If the environment changes then organisms will be ‘naturally selected’ whose variations better adapt them to that changed environment.

Gradualism: Intrinsically bound up with Darwin’s theory is the notion of gradualism in evolution at all scales. This is the view that the major explanation for all evolutionary change in the history of life on earth was the slow, piecemeal and gradual result of generation by generation and barely noticeable adaptations of organisms.

Organisms: Another crucial point for Darwin concerns the agency of evolution. It is in most cases, he insisted, organisms—individual animals or plants—which interact with their environments and so are subject to natural selection, and not generally species or groups.

Nature of variation: A requirement for this view of slow, gradual adaptation of organisms to be true, and one which Darwin himself insisted on again and again, concerns the character of the variation to which organisms are naturally subject. For Darwin this variation must be copious (lots of it), small scale, and without any preferred direction. Natural selection requires this kind of variation as the raw material on which it works.

That the variation in organisms does not have any preferred direction is very important. If there are preferred directions, to do for example with some internal biological principle, then this would be powerful enough to overwhelm the slow, piecemeal and weak, but over long periods powerful, action of natural selection.

There have been challenges to Darwin’s full theory ever since he outlined it. Gould describes them, and explains why all were beaten off—though he also points out that some raised important issues which still demand serious discussion.

He explains how it was not really until around the 1930s that Darwinism became generally accepted by the majority of biologists.8 Darwin had no idea of the mechanism of inheritance—how organisms passed on their characteristics to offspring. Gould charts how it was the rediscovery at the beginning of the 20th century of Gregor Mendel’s work on laws of inheritance and genes that gave a huge boost to Darwin’s theory.

From the 1930s a series of brilliant scientists—R A Fisher, J B S Haldane and Sewall Wright—then developed powerful mathematical models giving expression to Darwin’s theory in terms of the spread of genes linked to favourable variation through populations. This work laid the basis for a process from the 1940s to the 1960s of the full development of what came to be labelled the Modern Synthesis. Key figures in developing this theory included Ernst Mayr and Theodosius Dobzhansky. This orthodoxy has undergone further development (without radically changing in its basic concepts and arguments) in recent years as our understanding of the molecular basis of genes has rapidly developed.

This ‘orthodox’ Darwinian theory emphasised that evolution was about nothing other than the gradual adaptation, by means of the accumulation of small changes through natural selection, of organisms to their environment. All evolutionary change could be explained by this single process operating at that single level. That ‘all’ is important. It includes not just how change within particular species or populations takes place. It also includes what is called ‘transpecific’ evolution, the change or transition from one species to another, and also the larger or broader pattern of species coming into existence and becoming extinct throughout the history of life on earth.

Mayr, who became the central figure in the orthodoxy, put it explicitly:

The proponents of the synthetic theory maintain that all evolution is due to the accumulation of small genetic changes, guided by natural selection, and that transpecific evolution is nothing but an extrapolation and magnification of the events that take place within populations and species.9

It is against this background that Gould develops his arguments. Gould is a Darwinian. He accepts and defends the core of Darwin’s theory as the basis for any serious understanding of evolution. But Gould and others have developed a series of arguments that revise and expand key aspects of Darwin’s theory into what they argue is a richer and more accurate understanding.

He makes clear his overall project by quoting one definition of ‘the neo-Darwinian paradigm’ which ‘asserts that this history of life at all levels—including and even beyond the level of speciation and species extinction events, embracing all macroevolutionary phenomena—is fully accounted for by the processes that operate within populations and species’. Gould declares, ‘I dedicate my book to refuting this traditional claim’.10

Let me sketch out some of the key issues involved in that project.

Gene selection

This is the attempt to argue that the key unit of natural selection is the gene. It is a view most explicitly stated by Richard Dawkins in his book The Selfish Gene, but one essentially shared by Dennett, Pinker and many others.

Dawkins explicitly says, ‘The fundamental unit of selection…is the gene, the unit of heredity.’ And ‘selection occurs at only one lowest level—the gene’.11 In typically colourful prose Dawkins adds, ‘We are survival machines—robot vehicles blindly programmed to preserve the selfish molecules known as genes’.12 Gould argues, drawing on arguments developed by the science philosopher David Hull, that this view is not Darwinian, and that it is factually, scientifically and logically wrong.

He counters that, in general, it is the organism that interacts with the environment and the organism that is therefore the unit of selection. The key point in Gould’s argument is that natural selection is a theory of causes, not simply correlations. It is not good enough, he insists, to say that adaptive evolution of organisms is linked to and correlated with the changing frequencies of particular genes in a population. The theory argues, in its classical Darwinian formulation, that the adaptive evolution of organisms is the causal result of interaction between organisms and the environment. Genes, Gould insists, are generally not the thing that interacts directly with the environment, they are therefore not the causal agent, and so not the unit of natural selection.

Genes certainly influence the characteristics of an organism in crucial ways. And by the change in their relative frequencies in a population of organisms, genes record the results of natural selection. But Gould argues that by focusing on these frequencies (which they do because they are amenable to sophisticated mathematical treatment) the gene selectionists make a fundamental error:

The misidentification of replicators [genes] as causal agents of selection—the foundation of the gene-centred approach—rests upon a logical error best characterised as a confusion of bookkeeping with causality.13

Let me quote Gould at some length, as the argument is at the heart of what the press have labelled ‘The Darwin Wars’ and is crucial:

When we consider the character of natural selection as a causal process, we can only wonder why so many people confused a need for measuring the results of natural selection by counting the differential increase of some hereditary attribute (bookkeeping) with the mechanism that produces relative reproductive success (causality)… Units of selection must be actors within the guts of the mechanism, not items in a calculus of results…14

In sexual organisms, and in other higher level individuals, genes do not usually interact directly with the environment. Rather they operate via the organisms that function as the agents in the ‘struggle for existence’. Organisms live, die, compete and reproduce; as a result, genes move differentially to the next generations… Of course genes influence organisms; one might even say, metaphorically to be sure, that genes act as blueprints to build organisms. But such statements do not substantiate the critically necessary claim that, therefore, genes interact directly with the environment when organisms struggle for existence. The issue before us—the venerable problem of ‘emergence’—is largely philosophical and logical, and only partly empirical. Genes would interact directly only if organisms developed no emergent properties—that is, if genes built organisms in an entirely additive fashion, with no non-linear interaction among genes at all [but] organisms are stuffed full of emergent properties… Dawkins has reversed nature’s causality: organisms are active units of selection; genes, while lending a helping hand as architects, remain stuck within these genuine units.15

The reference to non-linear interaction underlines that Dawkins’ argument is an extreme example of what is called ‘reductionism’. Non-linearity is a mathematical concept which means that you can’t understand a total picture by simply adding up the individual contributions of the component parts.

It is, I think, beyond argument that biological organisms are almost always non-linear in this sense. While not denying the vital importance of genes, you simply can’t reduce how an organism works down to the level of individual genes. Gould rightly argues, ‘Any non-linearity precludes the causal decomposition of a body into genes considered individually—for bodies then become, in the old adage, “more than the sum of their parts”.’16

In passing, the case against genetic reductionism was underscored in powerful manner with the announcement of the very small number of genes in the human genome last year. This makes any idea of a simple reduction between features of the whole organism (in this case a human being) and this or that gene impossible.

It seems to me that Gould’s case is very powerful, and I, for one, find it convincing. A close reading of Dawkins’ later works, such as The Extended Phenotype, suggests that he knows this too. Dawkins ends up half-abandoning his arguments in The Selfish Gene. While he prefers to focus on genes, others may legitimately focus on organisms instead. He says, ‘I am not saying that the selfish organism view is wrong.’ In fact ‘the point is that neither of the two perceptions’ is ‘the correct or “true” one. They are equally correct.’ And he adds, ‘I doubt that there is any experiment that could be done to prove my claim… Readers must decide for themselves whether they like the new view [his] better than the old.’ Whatever else this rather strange ‘take your pick’ view of science is, it is certainly a long way from the confident, if misguided, rhetoric of The Selfish Gene.17

Relative frequency

An important aside is that very few of those involved in these debates dispute that the things that each points to actually happen. They acknowledge that the natural world is full of different and complex causes and phenomena. The argument is one about relative frequency—which of the causes or phenomena is the dominant one in a statistical sense. This, whether we like it or not, is the nature of truth in this branch of science.

Darwin himself, for example, always insisted that natural selection was not the only mechanism of evolution, but rather that is was the main or dominant mechanism. In a famous passage which an exasperated Darwin added to the sixth edition of The Origin of Species in 1872, he wrote:

As my conclusions have lately been much misrepresented, and it has been stated that I attribute the modification of species exclusively to natural selection, I may be permitted to remark that in the first edition of this work, and subsequently, I placed in a most conspicuous position—namely at the close of the Introduction—the following words: ‘I am convinced that natural selection has been the main but not the exclusive means of modification.’ This has been to no avail. Great is the power of steady misrepresentation.18

With that warning in mind let me turn to perhaps the most important, pillar of Gould’s argument.

Punctuated equilibrium

The theory of punctuated equilibrium was first put forward in 1971 by Gould and his colleague Niles Eldredge. Much has been written about it, and much of it has badly, sometimes wilfully, misunderstood what the theory actually says. Some of those labelled Ultra-Darwinians, or Darwinian Fundamentalists, claim now that the theory doesn’t really say anything they didn’t already acknowledge. This simply won’t do. They rejected the theory sharply for many years. Only accumulating evidence has pushed them to say that, well we knew it all along and it is nothing new.

What is the argument? For Darwin and the defenders of the Modern Synthesis all evolutionary change is gradual and piecemeal, a succession of small steps. One obvious objection to this Darwin himself pointed to: that’s not how it looks from the fossil record. There you get a different pattern. Species seem to appear suddenly in the fossil record, then persist relatively unchanged throughout the existence of that species, and then become extinct.

Darwin’s answer, and the orthodox response throughout the last 150 years, was to say that the fossil record is full of gaps and is imperfect. They argued that the reality was that evolution proceeded by the slow, steady accumulation of step by step changes. This gradualism should mean that between one species and another there will have been all sorts of intermediary forms. If we don’t see them in the fossil record, and generally we don’t, this is a result of the imperfections in the record.

Gould and Eldredge argued that while the fossil record is imperfect, it wasn’t that bad and its evidence should be taken more seriously. They put forward their theory to explain what they said was a reality.

There is a crucial issue here, and the failure to grasp it is at the root of more misunderstanding of arguments about punctuated equilibrium—and evolution in general—than any other. This is geological time. When geologists, or evolutionists looking at the pattern of life in geological history, talk about ‘sudden’ they are usually talking in geological time. Sudden in this context means between one sedimentary layer and the next in rock formations—rocks created by the gradual laying down of sediments carried in water. That can mean anything from tens, more likely hundreds, of thousands of years to even several million years!

The theory of punctuated equilibrium has been attacked for being about ‘evolution by jerks’ or ‘sudden jumps’. At least some of those making such attacks are wilfully misrepresenting the argument. The theory is not about sudden jumps at all in any everyday sense of time. In fact it is based on the orthodox theory of natural selection. Its whole argument was originally based on theories about how new species develop taken directly from Ernst Mayr, one of the architects of the Modern Synthesis.

Gould stresses:

The theory of punctuated equilibrium attempts to explain the macroevolutionary role of species and speciation as expressed in geological time. Its statements about rapidity and stability describe the history of individual species; and its claims about rates and styles of change treat the mapping of these individual histories into the unfamiliar domain of ‘deep’ or geological time—where the span of a human life passes beneath all possible notice, and the entire history of human civilisation stands to the duration of primate phylogeny as an eyeblink to a human lifetime.19

Gould argues that punctuated equilibrium is what you would expect to see when you shift from a fairly orthodox understanding of how evolution and speciation takes place at a micro level to the way this would then appear in the very different time-frame of geological history: ‘Punctuated equilibrium emerges as the anticipated expression, by proper scaling, of microevolutionary theories about speciation into the radically different domain of “deep” or geological time’.20 And, ‘Our theory holds, as a defining statement, that ordinary allopatric speciation, unfolding gradually at microevolutionary scales, translates to punctuation in geological times’.21

The really important and novel point about punctuated equilibrium is in fact the precise opposite of what is sometimes focused on. It is the ‘equilibrium’, or what Gould calls stasis. This is the fact that when you look at the fossil record species have a stability and coherence, and don’t undergo much systematic evolution throughout most of their geological existence. This stasis is what was long denied by Gould’s opponents. For them there should be constant gradual change within lineages wrought by natural selection.

Gould argues instead that the stasis of individual species is the reality, and the prevailing pattern over geological time. When you look at the fossil record you do see variation within species as time passes. But this is essentially confined within certain limits, a kind of oscillation around a mid-point, not generally a process of gradual transformation into new species as the Modern Synthesis would suggest.

Stasis does not mean ‘rock stability’ or utter invariance of average values for all traits through time… Of course mean values will fluctuate through time [but]…punctuated equilibrium makes the strong claim that, in most cases…a species at its last appearance [in the fossil record] before extinction, does not systematically differ from the anatomy of its initial entry into the fossil record, usually several million years before.22

Gould argues that species in fact have well developed mechanisms for maintaining stability over time. Speciation—the evolution of new species—occurs in very definite conditions, often in small populations isolated from the main group of an existing species—though there is still real and unresolved debate about the way speciation happens, as Gould acknowledges. But the theory of punctuated equilibrium does not depend on a particular model of how speciation takes place. The crucial point is the pattern of long-term stability of species and then speciation being the crucial factor in evolutionary change on a geological time-scale.

In the original version of punctuated equilibrium Gould and Eldredge argued that the stasis within species was so powerful that most meaningful evolutionary change in a geological time-frame only occurred with speciation, as that is certainly how it looks from the fossil record. Speciation, in a sense, they argued, was evolution. They have modified their theory in the light of an argument put forward by Douglas Futuyma. He argues that evolutionary change may go on inside a species outside speciation events. But he insists that such change will be ephemeral, and leave no serious trace in the geological record, unless it is first ‘fixed’ by being incorporated into a new species which then persists for some time as a stable entity, exactly the kind of stasis Gould points to.

Futuyma argues:

Although speciation does not accelerate evolution within populations, it provides morphological changes with enough permanence to be registered in the fossil record. Thus it is plausible to expect many evolutionary changes in the fossil record to be associated with speciation.23

Gould agrees that, ‘Speciation does not necessarily promote evolutionary change; rather speciation “gathers in” and guards evolutionary change by locking and stabilisation for sufficient geological time’.24

A huge section of Gould’s book, among the most technical, consists in putting forward evidence and technical studies to back up his argument on punctuated equilibrium. In principle the theory is eminently testable by quantitative studies of its predictions about the fossil record against the predictions of alternatives. Gould argues that the accumulating evidence generally comes down on his side. To my mind the evidence he presents on this is powerful.

Some of those dubbed Ultra-Darwinians have argued that Gould’s species ‘stasis’ is simply another way of talking about what they would call perfect adaptation of organisms to their environment. More precisely the Modern Synthesis would argue that organisms become ever better adapted to their environment by natural selection until they reach a peak of ‘adaptive fitness’, and that then there is no reason for further evolution until the environment radically changes.

This simply won’t do, and misses the point about how powerful the evidence for long-lasting stasis over geological time is in the fossil record. As Gould notes, ‘If stasis merely reflects excellent adaptation to environment, then why do we frequently observe such profound stasis during major climatic shifts like ice age cycles, or through the largest environmental change in a major interval of time?’25 We have to account for why species show powerful stability even in the face a of a changing environment. We also, and crucially, have to account for the changing pattern of species through geological time. It is in this area that the theory of punctuated equilibrium makes its most radical claims.

Some have also charged that the theory is not really saying anything new or important. Dawkins says it is ‘an interesting but minor wrinkle’ in evolutionary theory. Dennett says it is ‘a false alarm revolution that was largely if not entirely in the eyes of the beholder’.26 Such judgements are, I think, wrong.

In my view Dawkins and Dennett have a particular, narrow and wholly unjustified view of what evolution is about. They reduce the whole of evolution to the single focus of adaptive design of individual ‘body’ parts in organisms. Anything else, such as the entire pattern of evolution through the history of life on earth, the coming and going of species and whole classes of life, and much more, simply doesn’t matter. Gould argues, ‘They [Dawkins and Dennett] regard punctuated equilibrium as trivial because our theory doesn’t speak to the restricted subset of evolutionary questions that, for them, defines an exclusive domain of interest for the entire subject’.27 Gould is surely correct to insist that evolution is about much more than Dawkins and Dennett allow.

Species selection

At the core of the theory of punctuated equilibrium is not just the idea that species show stability, or stasis. It is also that species in many ways behave as ‘individuals’, that they have a collective existence and collective or ‘emergent’ properties.

Gould argues that the pattern of life’s history on earth can only be understood fully when you see that these ‘species individuals’ and not just individual organisms also undergo a form of natural selection.

He argues that things such as the range of variability within a species (obviously a property of the species, not of an individual organism), geographic range, or the rate at which a species itself produces new species (again things which cannot be reduced to the property of an individual organism) can and do affect a species’ chances of surviving through geological time. He goes on to argue that the pattern of life’s history in his model of punctuated equilibrium is a direct result of this kind of species selection.

There has been much controversy in the history of evolution over concepts of ‘group’ or ‘species’ selection, notably with a famous 1966 book by George Williams which powerfully argued against it.28 Williams’ book, it should be noted, was also the key manifesto in the drive to reduce all of evolution to gene selection. Williams himself has now largely abandoned his stance and in essence accepts a hierarchical view in which selection operates at a number of different levels, a position not too different from that advocated by Gould.

The argument is not about whether species selection can or does happen. It is about, as ever, relative frequency: how often it occurs and how powerful it is. All key figures in the history of evolutionary theory have accepted that species selection could or did occur. Darwin himself argued for a limited role for species selection, while maintaining his main focus on selection at the level of individual organisms. R A Fisher, the key founder of the Modern Synthesis, insisted that species selection could occur, but argued that it would be rare. Gould argues that the reasons some of these people gave for the supposed rarity of species-level or group selection are not valid in the light of modern knowledge.

To avoid a long and technical discussion I will give just one example and note also that Gould’s most determined opponents accept he is right on this.

Gould points to an example accepted as crucial by all sides in the argument—sex ratios in animals. The particular example is about demes, locally isolated interbreeding populations, rather than entire species. But as the argument is about any form of ‘group’ selection this does not really matter:

Some well-documented patterns in nature seem hard to explain without a strong component of inter-demic selection. Female-biased sex ratios…provide the classic example because two adjacent levels [of the focus of selection] make opposite and easily tested predictions: conventional organismic selection should favour a 1:1 [sex] ratio by Fisher’s famous argument; while inter-demic selection should promote a strongly female-biased ratio… Williams accepted this framework, which he proposed as a kind of acid test for the existence of group selection. He allowed that female-biased ratio would point to group selection, but denied that any had in fact been documented…but empirical examples of female-biased ratios were soon discovered aplenty… I think all major participants in the discussion now admit a strong component of inter-demic selection—and reported cases now number in the hundreds so this phenomenon cannot be dismissed as an odd anomaly in a tiny corner of nature. Williams now accepts this interpretation, writing that ‘selection in female-biased Mendelian populations favours males, and that it is only the selection among such groups that can favour the female-bias’.29

Richard Dawkins also accepts Gould’s argument on this, but says, ‘Species-level selection can’t explain the evolution of adaptation: eyes, ears, knee joints, spider webs, behaviour patterns, everything, in short, that many of us want a theory of evolution to explain.’ Gould responds, and with justice, ‘Everything? Does nothing else but adaptive organismal design excite Dawkins’ fancy in the entire and maximally various realm of evolutionary biology and the history of life?’30

Dawkins acknowledges that species selection is bound up with the core of what is important about the theory of punctuated equilibrium. He concedes that punctuated equilibrium:

…does, in a sense, move outside the neo-Darwinian syntheses narrowly interpreted. This is about whether a form of natural selection operates at the level of entire lineages, as well as at the level of individual reproduction… To conclude the discussion of species selection, it could account for the pattern of species existing in the world at any particular time. It follows that it could also account for changing patterns of species as geological ages give way to later ages, that is, for changing patterns in the fossil record.31

Nevertheless Dawkins then argues that species selection ‘is not a significant force’ and that ‘species selection may occur but it doesn’t seem to do anything much’. With entirely justified exasperation Gould notes, ‘Doesn’t “the pattern of species existing in the world at any particular time” and “changing patterns in the fossil record” represent something of evolutionary importance?’32

Dennett also finally concedes:

The right level at which to look for evolutionary trends, he [Gould] could then claim, is not the level of the gene, or the organism, but the whole species or clade. Instead of looking at the loss of particular genes from gene pools, or the differential death of particular genotypes within a population, look at the differential extinction rate of whole species and the differential ‘birth’ rate of species—the rate at which a lineage can speciate into daughter species… It may be true that the best way of seeing the long term macroevolutionary pattern is to look for differences in ‘lineage fecundity’ instead of looking at the transformation in the individual lineages. This is a powerful proposal worth taking seriously.33

Finally, Ernst Mayr himself, the doyen of the Modern Synthesis, accepts that punctuated equilibrium and species selection are a linked and crucial advance in theory:

It was generally recognised that regular variational evolution in the Darwinian sense takes place at the level of the individual and population, but that a similar variational evolution occurs at the level of species was generally ignored. Transformational evolution of species (phyletic gradualism) is not nearly as important in evolution as the production of a rich diversity of species and the establishment of evolutionary advance by selection among these species. In other words, speciational evolution is Darwinian evolution at a higher hierarchical level. The importance of this insight can hardly be exaggerated.34

As hinted at in that quote, species selection working alongside ‘normal’ organism-level selection, and their scaling into geological time as the pattern of punctuated equilibrium, are part of a wider hierarchical view of evolutionary theory. Gould argues that a fully developed theory of evolution requires such a hierarchical view in which natural selection operates in different ways at a series of different levels, from the gene, the cell and the organism to the deme and the species.

I mentioned gene selection. Gould does argue for some forms of selection acting at the gene level, as part of an integrated hierarchical explanation. The point, he insists, is not to reduce selection and the processes by which evolution takes place to one exclusive level. It is rather to examine at each level if there is a causal interaction with the environment which results in selection.

So he argues that there are some processes in which genes, or more precisely sections of DNA (though often ones which are not functioning genes), interact in this way and are selected for. Gould also points out, though, that most genetic transformations at a molecular level that are incorporated into evolution now appear to be ‘selectively neutral’ rather than being selected for. This is the implication of the extremely important theory of ‘neutral theory of molecular evolution’ developed by the Japanese geneticist Motoo Kimura—and there is strong evidence to suggest that, at least in part, this theory is correct.35

Gould is emphatically not arguing in some vague, hand-waving way for a hierarchical model. Rather he begins to spell out how the process of selection works at each level and how it then relates to other levels:

For Darwin’s near-exclusivity of organismic selection, we now propose a hierarchical theory with selection acting simultaneously on a rising set of levels, each characterised by distinctive, but equally well defined, Darwinian individuals within a genealogical hierarchy of gene, cell lineage, organism, deme, species and clade. The results of evolution then emerge from complex, but eminently knowable, interactions among these potent levels, and do not simply flow out and up from a unique causal locus of organismal selections.36

I don’t know if he is correct, and certainly, as he acknowledges, there is still much work to be done. In particular the relative importance and the mechanisms underlying forms of selection at different levels need much more investigation. But to me Gould’s basic arguments seem strong.

Spandrels and exaptation

To illustrate how Gould begins to flesh out his arguments in his hierarchical model of how evolution works, let me look at what he calls spandrels. In typical Gould fashion this is a term which he and Richard Lewontin borrowed from architecture—to be more precise from the architecture of the San Marco cathedral in Venice, though spandrels exist in many buildings.

If you have, as you do in San Marco’s, four round arches standing on the sides of a square and you put a dome on top the result is to create a series of roughly triangular spaces between the dome and the lines of the arches. These spaces are an example of what architects call spandrels.

They are a necessary structural by-product of the main thrust of design—putting a dome on top of the arches. But the spandrels themselves were not the purpose of or the key to the design. It may happen though that an architect can then use these spandrels for some other purpose, as they do with brilliant mosaics in San Marco. The spandrels were not originally designed for and are not an ‘adaptation’ for the mosaics. They have instead been what Gould calls ‘exapted’ for that purpose later. In other words, current use and historical origin are two different things.

Gould carries the argument over into evolution. Contrary to the Modern Synthesis, Gould and Lewontin argue that not every feature in an organism is an adaptation. Rather there are things produced in organisms as necessary by-products of other adaptations, or for other reasons.

These then exist, and they are either selectively neutral or any selective disadvantage they may have is massively outweighed by the selective advantage of the feature to which they are linked. But once they exist, then later they can then become ‘exapted’, actively selected for in other circumstances, such as if the environment changes.

Alongside this is also the phenomenon that something may be originally an adaptation for one purpose, but in new circumstances it can then be ‘exapted’ and used for an entirely different purpose. These are all actually well known and generally accepted phenomena in evolution. Darwin, with his usual perceptiveness, pointed to them using the label ‘correlations of growth’ and, to take someone who has attacked Gould, John Maynard Smith does too.37 Feathers, for example, are an exaptation. They were evolved in the reptilian ancestors of birds where they served for temperature regulation. Only later were they then ‘exapted’ in a different context for flight.

Gould’s originality here, other than coining nice words and images, lies more in his argument that these phenomena—spandrels and exaptation—are crucial to his hierarchical model of selection at different levels. Much of the discussion is quite technical, but essentially he puts two arguments.

One is that referred to above, and is about a particular level of selection, such as the organism. Once spandrels exist in an organism for ‘non-adaptive’ reasons, such as a necessary structural linkage to some other adaptive feature, they can then in later, changed, circumstances be the basis for exaptation by selection.

More subtly, though, Gould also argues that changes, even adaptive changes, at one level of selection will also have associated spandrels in their expression or effect at a higher level. These spandrels then supply much of the variation which then becomes the basis for later selection at that higher level:

To epitomise the central argument: under a hierarchical theory of selection, any novelty introduced for any reason (usually adaptational) at any level, must propagate a series of effects to biological individuals at other levels of the hierarchy. Duplication of genetic elements by direct selection at the gene level, for example, propagates redundancies to the organismic level; any organismic adaptation at Darwin’s level propagates changes to the encompassing species-individual, as expressed in such species traits as population size, geographic range, and coherence among subparts (organisms). These propagated effects must be defined and treated as spandrels.38

And these spandrels then form part of what Gould calls the ‘exaptive pool’, elements which are then available for selection at the higher level.39

To sum up, Gould argues for an integrated theory in which selection at a hierarchy of different levels, from gene through organism to species, is needed to explain the pattern of evolution. He argues that this is an expansion and addition to the orthodox Darwinian focus on natural selection at the organism level, not a challenge to that. And he argues that the levels of selection are connected by specific mechanisms, in which his concepts of spandrels and exaptation rather than only adaptation play a role. Finally he argues that the result is to create the pattern of life’s history summed up under the label of punctuated equilibrium.

Two other key elements need to be addressed in Gould’s argument.


There is a long history in evolutionary theory of attempts to argue that internal, structural factors in organisms play a key role in directing or channelling evolution in addition to, or even instead of, natural selection. The specific details of such explanations have mostly been proven wrong, though many of the most eminent biologists in history have been attracted by them for very good reasons. Gould examines this history in fascinating detail. He then argues that the whole debate needs to be revisited in the light of modern scientific advance. In particular he argues that recent developments in the area of evolutionary development (looking at how genes govern development of organisms) have shown that there are structural features which do indeed channel evolution.

He seems to be on very strong ground here, and I think even his critics would accept this. Let me give just two examples.

One is the stunning discovery that there is a common set of genes called Hox genes underlying crucial phases in the development process and basic body plans of both arthropods (segmented animals, such as insects), and mammals—even though they have been separated as lineages by some 500 million years of evolutionary history.40

These are what are technically called ‘homologous genes’. What matters is that this was not supposed to happen on the orthodox model. Any such ‘homologies’ should have been erased by the years of independent, constant and gradual adaptive change. No less an authority than Ernst Mayr in his canonical summation of the Modern Synthesis declared that even looking for such homologies was a waste of time:

In the early days of Mendelism there was much search for homologous genes… Much that has been learned about gene physiology makes it evident that the search for homologous genes is quite futile except in very close relatives.41

That view doesn’t stand up in the light of modern discoveries. A second example is worth giving here to underline how revolutionary these findings are. This is the stunning discovery about one of the key examples used to underline the power of adaptation through natural selection—the independent evolution of the eye in many widely separated lineages.

The eye has been independently evolved, it seems, in at least six entirely different animal ‘phyla’—large groupings or divisions into which the biological world is divided. For example, vertebrae (animals with backbones, including mammals) and the very different cephalopods, animals such as octopuses, have both independently evolved single-lens eyes from very different starting points. In vertebrates the eye has evolved from an outgrowth of the brain. In octopuses and squids it has evolved from the skin.

This example of ‘convergent evolution’ has, rightly, been used to show how natural selection can, from radically different starting points, evolve similarly superb adaptations again and again. Yet now it has been discovered that a set of genes, called Pax6, play a similar role in both, as well as playing a similar role in the quite different multiple-lens insect eye.42

What all this strongly suggests is that there are very important constraints and channels underlying evolutionary processes in a way that does not fit easily with the views of the Ultra-Darwinians. This whole field of evolutionary development is rapidly advancing and is likely to throw up more important and unexpected results in the coming years. Again though, like Gould’s earlier arguments, these findings are an addition to and enrichment of Darwin’s basic theory, not a contradiction of it.

Direction and contingency

The final argument to turn to is Gould’s particular take on patterns in the history of life, and on issues of direction, chance and necessity in that history.

He argues persistently, and rightly, against the idea that there is a ‘ladder of progress’ in evolution. So he attacks the idea of the ‘ascent of mankind’—often seen on pictures showing a short and stooping ape gradually turning through a smooth sequence of intermediaries into an upright (and usually, white middle class male) modern human.

This is not just some old fashioned view. For instance, I was recently in the otherwise excellent and modern archaeological museum in Strasbourg and was confronted by precisely such an image at the start of the section on human evolution. Gould is right to polemicise against what is simply a misreading of the evidence and misunderstanding of how evolution works.

He argues, correctly I think, that the process of natural selection is in itself not directional. There is nothing in it which necessarily pushes evolution towards, for example, more complex, more sophisticated, or more intelligent organisms. Evolution and adaptation in the face of changing environmental conditions, he argues, can equally be towards (roughly speaking) simpler organisms. There is no preferred direction or vector of change.

Gould also makes the point that the dominant, in the sense of most typical, form of life has not changed throughout evolutionary history—bacteria dominate evolutionary history and life. Again, this is an incontrovertible fact, however much it may make some humans feel uncomfortable.

In some of his earlier writings Gould pushed his argument to the extreme, and denied any tendency or pattern towards more complex organisms at all. In The Structure of Evolutionary Theory, and at greater length in his earlier Life’s Grandeur, he advances a more interesting and defensible position. He maintains his points about lack of direction underlying the evolutionary process and of bacteria as the unchanging most dominant form of life. But now he concedes that nevertheless there does seem to be a pattern in life’s history, one towards the evolution of more complex organisms over the time since life began on earth. The most recent and most obvious example to emerge is us, self conscious organisms.

Gould now says, ‘The history of life includes some manifestly directional properties—and we have never been satisfied with evolutionary theories that do not take this feature of life into account’.43 He also proposes an explanation to resolve this seeming paradox, an explanation which I think stands up. He argues that while at any time evolution towards (roughly speaking) more simple or more complex organisms is equally likely, there is one important difference between the two.

This is that there is a limit below which any evolution cannot go in one direction. Put simply, below a certain level it ceases to be life at all. On the other hand there is no such limit, so far as we know, at the other end of the scale, in the direction of more complexity. There is a wall of minimal complexity at one end of the distribution of life’s forms but no such limit of maximal complexity at the other. This will, he argues, produce a trend.

Even though the underlying process is inherently directionless, the different ‘boundary conditions’ at each end of the distribution of life’s forms (on a scale of some measure of complexity) will result in a longer and longer tail in that distribution in the direction of more complexity as time goes on. In short, over time there is a tendency for the most complex organisms (always a small minority of all organisms) to become more complex.

In this argument Gould famously uses baseball averages to explain his reasoning. That popular image helps get a subtle mathematical argument across more easily. And that mathematical argument is, in my view, sound. I think Gould gives a good explanation of how you can both have an underlying process which in itself has no direction yet which nevertheless produces a pattern, a trend in a certain direction.44

Gould concludes, ‘We understand the stability and continued domination of bacteria as the outstanding feature of life’s history, with the much vaunted progress of complexity towards mammalian elegance reinterpreted as a limited drift of a minor component of diversity into the only open space of complexity’s theoretical distribution’.45 Despite the somewhat grudging language this shouldn’t disguise a genuine shift on Gould’s part, towards what I think is a defensible position.

Gould has one further argument worth examining, on what he calls contingency. He points out that if you look at the pattern of life’s history on earth it is punctuated by a series of great extinctions, in which much—and in one case almost all—of life was wiped out.

The most famous is the Permian extinction around 245 million years ago, in which 95 percent of all species were wiped out. Another more recent such event was the Cretaceous/Tertiary or KT extinction around 65 million years ago, which saw the demise of the dinosaurs. I say ‘event’ in the sense that these were ‘sudden’ in a geological time-frame—which means the ‘extinctions’ could have been processes which lasted an awfully long time by any ‘human scale’ measure of time. We don’t yet know the full reasons for any of these extinctions, though there is very good evidence that the KT extinction was linked to the impact of a giant meteor.

Gould’s argument is that chance or contingency played a key role in determining which species died out and which survived in these events, and which therefore were then the subject of further evolution. He argues also that these catastrophic mass extinctions are different to the normal patterns of evolution, including those under his theory of punctuated equilibrium.

There are some problems with Gould’s formulations, and when pushed I think he is claiming rather less than his rhetoric suggests. He is certainly right to stress that there are major events in the history of life which are not on the same scale as those operating at other times. It is not every (geological) day, so to speak, that you get the demise of entire groups like the dinosaurs, and certainly not every day that 95 percent of all known species disappear.

He is also right to say that viewed from the perspective of ‘normal’ environmental changes events such as, say, a meteor hitting earth are in a sense ‘contingent’ or chance-like. But that depends crucially on your frame of reference. From the point of view of biological processes operating on earth 65 million years ago a meteor strike may be contingent. But it is entirely predictable and subject to physical laws (eg Newton’s) from a wider perspective on nature. Also, most theories of the Permian extinction (by far the biggest and most significant such event) draw on environmental changes driven by physical processes within the earth itself, and do not depend on such an ‘external’ event as a meteor strike.

The processes which drove the Permian extinction may seem ‘contingent’ or unpredictable if your focus is simply the biological world on earth immediately beforehand. But widen your perspective to include in the picture geophysical processes under way inside the earth, and in principle open to full explanation, and things look rather different.

This is not an argument for some kind of strict deterministic view, in which if only we include all the necessary information we could predict everything in the smallest detail. Within science that kind of view, which traces back to Laplace, has been undermined forever by developments in chaos theory and its associated areas. I am here merely questioning Gould’s notion of ‘radical contingency’.46

There is also, surely, no argument that in all the cases of mass extinction the immediate, or what is called the ‘efficient’ cause was a severe and sustained change in the global environment. The mass extinction was thus the consequence of natural selection (operating in Gould’s sense at all its levels, from gene through organism and species to whole phyla) in this radically changed environment. This may be geologically sudden, and from some points of view at least unexpected or contingent. It is also true that radical changes in circumstances can cut across in dramatic fashion the patterns of evolution of normal times. But these are also processes that remain open to full scientific explanation. They are not simply chance or random events in the way that Gould’s rhetoric about ‘radical contingency’ all too easily suggests—even if he would have argued that this was not exactly what he meant.

Gould makes much of his argument that maybe you can explain what has happened after such dramatic changes as mass extinctions, with hindsight, but you cannot in any meaningful sense say the event of the pattern of extinction was predictable in advance. In fact I am not sure that our understanding of evolution at any level and even in ‘normal’ times has yet reached the stage where we can predict the precise impact of any real environmental change at all. We still seek to explain all these impacts and changes after they have happened.

Is Gould saying merely that faced with a truly dramatic and sustained environmental transformation (such as serious raising of average global temperatures) all we can say is that the evolutionary effects would be huge? Is he saying that we are unable to say in detail what these effects would be in advance, even if any surviving humans could perhaps explain them after the event?

If he is saying this, then he is right. But that is a long way from what his talk of ‘radical contingency’ suggests. To say that we cannot predict in detail doesn’t mean that we can say nothing. There are always some constraints, some channels down which historical events are more likely to go than others from a given starting point.

If the environment changed so that there was no fresh surface water on earth then I cannot predict exactly how evolution would pan out. I am fairly sure though that the future would be bleak for freshwater fish. Or, if the earth’s atmosphere was to revert to the oxygen-free composition it had for much of earth’s history, again, I cannot spell out the precise evolutionary impact. I am pretty certain however that the prospects would be rather bleak for quite a lot of species, including our own.

Gould’s rhetoric constantly seems to want to push the argument to a position where his concept of ‘radical contingency’ completely rules the history of life, and history in general. But, despite forming the core argument of long works such as his Wonderful Life, the concept is, to my mind, never clearly spelled out. His most famous analogy in attempting to explain it is his ‘tape of life’, first put forward in his discussion of the famous Burgess Shale fossils:

I call this experiment ‘replaying life’s tape’. You press the rewind button and, making sure you thoroughly erase everything that actually happened, go back to any time and place in the past—say, to the seas of the Burgess Shale. Then let the tape run again and see if the repetition looks at all like the original. If each replay strongly resembles life’s actual pathway, then we must conclude that what really happened pretty much had to happen. But suppose that the experimental versions all yield sensible results strikingly different from the actual history of life? What could we then say about the predictability of self conscious intelligence? Or of mammals? Or of life on land? Or simply of multicellular persistence for 600 million difficult years?47

Gould has a point, but an examination of his argument ends up, I think, with a different conclusion to his ‘radical contingency’. I suspect Gould was caught because he had not fully taken on board some of the scientific debates around determinacy and chance which emerged over the last decade or so. He doesn’t want to accept, and rightly so, a strictly deterministic picture. So he reaches for an alternative view, the one he dubs ‘radical contingency’.

It is certainly relevant to these discussions that we know today, in the light of chaos theory, that small, and, yes, even chance or contingent events, such as tiny changes in initial conditions, can radically alter final outcomes in unpredictable ways even in far more determined systems than biology. That certainly does means that in Gould’s ‘tape of life experiment’ the outcomes will almost certainly be different on each rerun, even radically different in important ways.

But there is more to it than that and history is not then simply contingency. The possible paths of development in any such system are constrained in important ways—which is why the very label of ‘chaos’ theory for the mathematics underlying some of this understanding is profoundly misleading. In biology, for example, Gould himself in The Structure of Evolutionary Theory does much to spell out some of the very real constraints on biology. Recent work by biologists like Stuart Kauffman has also cast light on the very real structural constraints and organisational possibilities underlying all life.48

And it is also true that the old and important arguments about ‘convergent evolution’ under the force of natural selection retain their force. Put crudely, natural selection will come up with ‘good tricks’ again and again. In a physical reality where light plays such a central role then eyes, for example, are likely to be evolved in some lineages sooner or later, however Gould’s tape plays out in detail.

Putting all this together suggests to me that what is needed is not some silly opposition between rigid determinism on the one hand and some vague notion of radical contingency on the other. It is instead an integrated and scientific understanding of how all these elements combine to produce the complex and rich reality of the world around us.

Gould extends his arguments on contigency from biology to a wider view of history, including social history. I will close, as he does his book, by looking at these.

He argues, for example, that the entire course of American history was changed by the actions of a certain Joshua Lawrence Chamberlain, who led a decisive bayonet charge in the battle of Gettysburg in the 19th century American Civil War. Gould argues that without this small event the South may have won the civil war.

Leave aside whether that is a correct judgement on the actual history. It is, of course, possible that events could have turned out so that the South won the war (though some important factors were stacked against them doing so). This would undoubtedly have had important effects on the particular way history developed afterwards.

It may have meant that the US would not have risen to the position it has today as the most powerful centre of global capitalism. But such a change would have been most unlikely to have halted a process already well under way, namely the development of capitalism as a global system with its wars and imperialism. It would have changed the way these processes were played out in very important ways, and changed the final architecture of that global system. But it would not have halted an underlying process rooted in the logic of capitalism. History is certainly not predetermined, but that does not mean ‘radical contingency’, or that anything goes. Some developments, some channels or patterns of events, are more likely than others, and there are a limited number of alternative possible developments at any stage in history, whether social or biological.

Gould himself acknowledges this general argument on the very last pages of his book, in a way that comes close to an abandonment in all but rhetoric of his talk of radical contingency:

I will grant one point to scientific colleagues and freely allow that if Charles Darwin had never been born, a well-prepared and waiting scientific world, abetted by a cultural context more than ready for such a reconstruction of nature, would still have promulgated and won general acceptance for evolution in the mid 19th century. At some point, the mechanism of natural selection would also have been formulated and eventually validated.

He adds that equally, ‘The Renaissance would have unfolded…if Michelangelo had never been born’.49

Gould argues that if either Darwin or Michelangelo had not been born then the precise, and richly interesting, patterns of how evolutionary theory was established or how the Renaissance unfolded would have been different.

Yes. But some such processes, as he acknowledges, would have unfolded all the same, driven by other deeper forces at work in society. As this argument is put on the last page of Gould’s final and most important work we could take it as his final and considered view. If it was, then, perhaps, he was guilty of little more than bending the stick too far against rigid determinists in his rhetoric about ‘the controlling power of contingency’. But bend it too far he did nonetheless.

In conclusion, I think Gould’s The Structure of Evolutionary Theory will prove to be a landmark in the history of science. It will, I suspect, be proved right in many areas. On many things it will certainly be shown to be wrong too, but even then fruitful and stimulating.

Gould himself was always very fond of quoting the conclusion to The Origin of Species. Charles Darwin ended that book, which he called ‘one long argument’, by saying, ‘There is grandeur in this view of life’.50 The same can most certainly be said of the one long argument in Stephen Jay Gould’s last and greatest work, as it can of his rich legacy of other writings too.


Throughout I have in most cases quoted from authors other than Gould by referring to where he quotes them in his book, as this is where the quotes appear in the context of the arguments Gould is having. Anyone wanting to check the originals, and so make up their own mind if Gould has taken quotes out of context, can use the references in Gould.

  1. Steven Rose was speaking at the Marxism 2003 event, London, July 2003. A tape of his speech is available from Bookmarks.
  2. S J Gould, The Structure of Evolutionary Theory (Harvard University Press, 2002). Below referred to as TSOET.
  3. Most of Gould’s essays are available in such collections, in Penguin or other reasonably priced paperback editions.
  4. In my opinion the toll of constantly arguing against these forces in the US lay behind what I consider by far Gould’s weakest book, Rocks of Ages (Vintage, 1999). In it he argues essentially for a truce between science and religion— that each has its own legitimate sphere which the other should not trespass on. They are what he calls ‘NOMA’ or ‘non-overlapping magisteria’. Gould is not stupid, and as usual argues persuasively. It should also be stressed that his argument is not entirely without foundation, in that there are realms of human understanding which are distinctive and not reducible to each other—art and science, for example. But he pushes the argument much too far in the case of religion.
  5. One small criticism is that Gould’s editors seem to have overindulged him a little. The book is too long in parts, in the proper sense that repetition of points, while useful for emphasis, can if overdone cloud understanding. There are also some parts which are overwritten and a little too florid and dense with clever rhetoric and literary and historical references. Gould, as he constantly reminds us in his book, was not a modest man. Winston Churchill once described the 1940s Labour Party leader in Britain, Clement Attlee, as ‘a modest man with much to be modest about’. Gould was, by contrast, ‘an arrogant man, with much to be arrogant about’. He even described himself as ‘the most arrogant of literati’ and famously refused to allow people to interfere with his prose. I cannot help feeling though that a judicious and skilled editor firm enough to demand streamlining and paring down of Gould’s prose where needed would only have improved the book, without losing any of the delight of his unique, and to my mind, generally entertaining style.
  6. While on almost all issues of real substance I think Gould and his allies are right and those labelled ‘fundamentalists’ wrong this does not mean I think the latter have nothing interesting to say. In fact I think all deserve to be widely read, and that we can learn from all of them. The differences between them are important too. Put crudely I would say Maynard Smith is, by far, the most interesting and the best of these figures, and Pinker by far the worst. I should stress that I think for example that Dawkins’ The Selfish Gene (Oxford Paperbacks, 1989) is, despite its title, an interesting book, which has much that is good in it even though it also contains many arguments and conclusions which I think are profoundly mistaken. The same is true of books such as Dennett’s Darwin’s Dangerous Idea (Penguin, 1995). John Maynard Smith’s The Theory of Evolution (Canto, 1993) is, I think, still the best serious introduction to the whole subject. His more recent books such as Major Transition in Evolution OUP, 1997) and The Origins of Life (Oxford Paperbacks, 2000) are also important works. These debates are also emphatically not a simple question either of ‘left’ against ‘right’. Richard Dawkins is certainly not a right winger politically, as his very public arguments against the religious right and his principled opposition to the war on Iraq testify.
  7. It is worth stressing that The Origin of Species is not a difficult book. It is clearly and well written for a non-specialist audience and remains today perhaps the clearest exposition of the theory of evolution by natural selection.
  8. By Darwinism I mean the theory of natural selection, not the fact of evolution, which was much more widely accepted.
  9. Quoted in TSOET, pp1003-1004.
  10. TSOET, p55.
  11. Quoted in TSOE, p638.
  12. Quoted in TSOE, pp617-619.
  13. TSOET, p614.
  14. TSOET, p619.
  15. TSOET, p620.
  16. TSOET, p627.
  17. Quoted in TSOET, pp639-640.
  18. Quoted in TSOET, p254.
  19. TSOET, p766
  20. TSOET, p781.
  21. TSOET, p1009.
  22. TSOET, p767.
  23. Quoted in TSOE, p802.
  24. TSOET, p801.
  25. TSOET, p878.
  26. Both quoted in TSOET, p1019.
  27. TSOET, p1020.
  28. G C Williams, Adaptation and Natural Selection (OUP, 1966).
  29. TSOET, p648.
  30. TSOET, p1020.
  31. Quoted in TSOET, p1020.
  32. TSOET, p1020.
  33. Quoted in TSOET, p1021.
  34. Quoted in TSOET, p894.
  35. See TSOET, p685. Also see J Maynard Smith, The Theory of Evolution (Canto, 1993), p103ff.
  36. TSOET, p32.
  37. See J Maynard Smith, as above, p304. Darwin pointed out for example that, ‘The sutures in the skulls of young mammals have been advanced as a beautiful adaptation for aiding parturition [birth], and no doubt they facilitate, or may be indispensable for this act; but as sutures occur in the skulls of young birds and reptiles, which have only to escape from a broken egg, we may infer that this structure has arisen from the laws of growth, and has been taken advantage of in the parturition of the higher animals.’ Darwin names this category ‘correlations of growth’ and offers a definition: ‘I mean by this expression that the whole organisation is so tied together during its growth and development, that when slight variations in any one part occur, and are accumulated through natural selection, other parts become modified. This is a very important subject, most imperfectly understood.’ Darwin also allowed that important, and not just trivial, characters could be shaped by such correlations of growth: ‘Hence we see that modifications of structure, viewed by systematists as of high value, may be wholly due to unknown laws of correlated growth, and without being, as far as we can see, of the slightest service to the species.’ All quoted in TSOET, pp332-336.
  38. TSOET, p1267.
  39. In fact Gould goes on to develop a wider range of theoretical concepts with names such as Miltons and Franklins in his account of this ‘exaptive pool’. I have some real doubts as to how valid his whole argument is. But I think it is a useful starting point for discussion. Even if his full treatment proves wrong, I think the notion of spandrels, the exaptive pool, and that they have a role in the linkage of different levels of selection is likely to prove useful.
  40. See TSOET, p1061ff.
  41. Quoted in TSOET, p1066.
  42. See TSOET, p1123ff.
  43. TSOET, p468.
  44. This argument is developed in full in Gould’s Life’s Grandeur. It is largely taken as read in TSOE.
  45. TSOET, p78.
  46. There is not space to discuss this here. For a wider discussion see my ‘Order out of Chaos’, in International Socialism 48 (Autum 1990).
  47. S J Gould, Wonderful Life (Penguin, 1991), pp48-50.
  48. See TSOET, p1208ff, and S Kauffman, At Home in the Universe (Penguin, 1995). Work by biologists such as Brian Goodwin also suggests similar strong structural constraints. See for example B Goodwin, How the Leopard Changed Its Spots (Phoenix, 1994).
  49. TSOET, p1342.
  50. C Darwin, The Origin of Species (Penguin, 1985), p459.