Lecturers || Stephen Jay Gould Home || Stephen Jay Gould: Commentary
Four fifths of the way down John Catalano's WWW site, The World of Richard Dawkins is a hypertext link for "The G files". Sandwiched between the links for "Memes" and "Evolution," most readers of popular biology will immediately recognize the reference. The "G" stands for Gould -- as in Stephen Jay Gould. The fact that there is a link leading to Gould's writings on a WWW page is not in itself remarkable. After all, Gould is America's most famous evolutionary theorist. What is remarkable is that this file is contained on a site dedicated to Gould's principal opponent in debates on the mechanisms of evolution, Richard Dawkins.
Dawkins trained as an ethologist and currently holds the Charles Simonyi Professor of the Public Understanding of Science at Oxford University. Gould trained as a geologist and currently holds the Alexander Agassiz Professor of Zoology and is Curator for Invertebrates in the Museum of Comparative Zoology at Harvard University. More than the Atlantic Ocean separates these thinkers, however, for they stand at opposite ends of the orthodoxy of evolutionary science.
Gould's epithets provide a convenient point of departure for a detailed analysis of these two thinkers' ideas. Gould refers to himself as an "evolutionary pluralist" while he refers to Dawkins as a "hyper" Darwinian. These labels are intended to describe their approaches to evolutionary biology: Gould has spent most of his career supplementing Darwin's theory of natural selection with other evolutionary mechanisms; Dawkins has sought to establish natural selection at the level of the gene. These labels also describe the creative temperament and methods that these thinkers bring to their work: Gould has a taste for the complexity of large systems; Dawkins prefers cutting phenomenon down to its simplest components. Since these two thinkers have repeatedly engaged each other through high profile mediums, it not only is easy to analyze their theoretical and methodological differences, such an analysis provides an interesting glimpse into how deeply scientific thought and modern culture interpenetrate.
Clicking on the "G Files" summons the recent vestiges of this controversy. Here you will find links to Gould's two-article plea for the "Pleasures of Pluralism" (only tangentially directed at Dawkins) published last June in the New York Review of Books, and Gould and Dawkins' reviews of each other's most recent books. But don't let the late publication dates fool you, the debates are decades in the making.
Dawkins ensured his fame as an evolutionary theorist with the publication of The Selfish Gene in 1976. A popularization of ideas first put forth by G. C. Williams in 1966, Dawkins argued that the gene and not the population must be viewed as the privileged biological unit for natural selection.2 The radical implication of this argument, put forth in provocatively clear prose by Dawkins, is that organisms are only the "vehicles" or "survival machines" for genes.3
Although Dawkins' frequently evokes the gene and DNA to make his arguments, his gene is most useful if it remains a theoretical construct. For instance, Dawkins' formulation of the gene does not incorporate the complex picture of the eukoryotic genome molecular geneticists are now forming. Especially important is the recent understanding of the complexity of genetic regulatory regions, coding region repetition in intervening genetic sequences, and the fact that a single coding region can be spread over an astonishing distance on the chromosome. No, Dawkins' gene does the most analytical work for him if it remains a non-material entity that can shrink or expand to fit the theoretical circumstances: "It is a fair point that the gene is an abstraction. [...] What I've said in the Selfish Gene is that I agree that we're not talking about a particular unit. There's a continuum."4
DNA is important for Dawkins, however. Well, at least the idea of DNA. Dawkins isn't interested in actual DNA sequence or DNA folding or protein DNA interactions; he is interested in the idea of the DNA as linear binary code. This allows Dawkins to make the leap from gene as material entity to gene as information. In Dawkins' writings the DNA is the digital information "river" that programs the messy analogue world that we inhabit. "Success in evolution is building programs that don't crash [...] The best way to look at an individual animal is as a robot survival machine carrying around its own building program."5
Gould's hyphenated modifier, "hyper", refers to Dawkins' explanations of evolutionary theory using the smallest possible units of analysis and for his single minded reliance on the theory of natural selection.6 Although some of his later books layered rich interpretations over his "selfish gene" argument, Dawkins' established his modus operandi early: explanatory power through reduction. As even his friend and admirer the philosopher Daniel Dennett has noted "[s]ome people object to Dawkins as being what I now call a greedy reductionist--that is, they think he's vastly oversimplifying, trying to get the job done with too few levels of explanation. Even though some version of that objection may be true, it's not a big deal. [...] Dawkins is not wrong--he's just been too optimistic sometimes."7
As the debates documented in the "G Files" testify, Stephen Jay Gould's modus operandi is the scholarly antithesis to Dawkins: explanatory power through synthesis. True to the modifier "pluralist," he has embraced a number of intellectual traditions in his efforts to broaden our understanding of evolutionary mechanisms. For example, Gould relies heavily on multidisciplinary analysis, personal anecdote, and historical and literary references in his writing. Gould's use of history is especially sophisticated. He not only recognizes current historical methods and issues; he has contributed solid scholarship in the history of biology. Where Dawkins looks toward the computer sciences for inspiration, Gould turns toward literature and history.
Although Gould recognizes the primary importance of natural selection in evolution, he has lobbied freely and eloquently for complicating simple accounts of natural selection. He has supplemented the customary argument that populations gradually gain in adaptive fitness through competition with bold complications and detailed exceptions. The complications appear under three main themes: offering exceptions to the account of evolution as gradual cumulative change (evolution as progress), pointing to the importance of historical contingency (thus leaving room for catastrophic climatic changes, such as the impact of a large comet), and establishing the importance of non-adaptive side consequences of variation (biological structures that constrain function even though they were not specifically selected for).
For Gould evolution is not progressive. Period. He has used metaphorical, historical, and geological analyses to make this case. Gould has often recognized the power of metaphor for articulating and communicating concepts. A favorite target for Gould is the metaphor of evolution as a ladder. Ladders suggest a pinnacle and a single route for reaching it. Natural selection, however, only creates adaptation to "local circumstances", "not any global scheme of progress."8 Gould has dedicated much of his historical work to competing views of time underlying geological theories,9 how recapitulation doctrine needs to be considered an ordering principle of late nineteenth century thought (not just natural selection),10 and how notions of progress should be considered as part of the "context of discovery" of late Victorian science.11
Gould's geology has also argued against progressive gradual change. In 1972 Gould and fellow geologist Niles Eldredge published an article that changed the way that many evolutionary theorists interpreted the fossil record.12 During a period of renewed interest in methods among professional geologists,13 Gould and Eldredge suggested that according to the dominant model of speciation, Ernst Mayr's theory of allopatric speciation, evolutionary theorists should not expect to find a geological record that reflected gradual change. If speciation did occur through the geographic isolation that Mayr postulated, an investigator should find jumps in the fossil record at a specific geological site. Consider a small group of individuals that become geographically separated from a larger population. If this small population goes through gradual selection to meet the demands of the new environment, eventually this population may even diverge enough to be considered a separate species. Consequently, if the newly diverged population is ever reintroduced back into the environment of the original population and either exterminates the original population or drives it into a different location, a geologist looking at the fossil record would not find gradual change. Instead, one would find relative stasis and then a large jump when the newly diverged population took over the environment.14
Lately Gould has used his work on metaphors as well as his theorizing on fossils to argue for a new conception of evolutionary change. Specifically Gould would like for us to change our image of the evolutionary tree of life from a cone shaped conifer to a "bottom heavy" bush.15 He suggests that species have undergone "maximal initial proliferation" during specific periods in the earth's history.16 In this model there is a rush of speciation from a common ancestor followed by a large scale "decimation" of most lineages. The remaining species then go through gradual proliferation and evolutionary change.17
One of Gould's most enduring papers, "The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme," provides a good example of Gould's appreciation for the dynamics of complex systems. As one of Gould's favorite writings, and because of the striking way that it incorporates interdisciplinary examples, it is also a useful paper for understanding how Gould likes to work.
Although it criticizes the unwise application of adaptationist thinking in evolutionary biology, the article opens with an architectural example: the spandrels of Saint Mark's Cathedral in Venice. A spandrel, according to Gould and co-author Richard Lewontin, is an architectural structure required by the construction of fan vaulted ceilings.18 For instance, if one mounts a dome on four arches that meet at right angles one creates four new triangular shaped surfaces. These surfaces are called spandrels. In the cathedral of San Marcos in Venice artisans decorated these surfaces with detailed mosaics portraying the four biblical rivers: the Tigris, the Euphrates, the Indus, and the Nile. Although beautifully ornamented, the architectural space for these mosaics arose as a side-consequence of an engineering requirement.
For Gould and Lewontin the dome mounted on the arches is an analogue for adaptation. It is a physical structure that meets engineering requirements just as certain biological structures might be said to be a result of the demands of a local environment. The spandrels, however, can not be explained as a direct consequence of engineering demands; they are "side-consequences of the adaptive decision to mount the dome."19 In Gould and Lewontin's view many important biological structures should be considered non-adaptive side consequences. In their paper they list two sets of constraints that could give rise to non-adaptive consequences: constraints imposed by phylogenic relationships and constraints imposed by development.20
The spandrel analogy is a powerful reminder of the importance of high level complexity in biological structures. "...[F]or if selection can break any correlation and optimize parts separately, then an organism's integration counts for little. Too often, the adaptationist programme gave us an evolutionary biology of parts and genes, but not of organisms."21
This brings me back to the G files.
Perhaps it is fitting that the most modern yet ephemeral of technologies separates these two men's ideas. For despite the appeals to Darwin, the ideas of both of these men could never have been articulated outside of the scientific, moral, and cultural context of the late-twentieth century.
Gould is the most obvious. With his refusal to acknowledge progress, his assertion that there is no privileged perspective to evolutionary change, and his open embrace for the "messiness" of large systems, Gould shares many qualities with his colleagues in the humanities. As we have seen, Gould fruitfully utilizes his love for interdisciplinary collaboration and thinking to formulate scientific concepts and strengthen explanations. Gould represents the synthetic humanist tradition of the modern academy.
With his emphasis on the importance of genes and his placement of organisms on a theoretical level subordinate to the digital textual apparatus that animates them, Dawkins represents the analytic tradition in the modern academy. At first glance then, it seems ironic that Dawkins shares some theoretical positions with his more radically modern colleagues in the humanities who claim that "there is nothing beyond the text."22 For instance, both Dawkins and prominent literary thinkers turn away from theories that promote individual causality (or agency) to find more enduring truths in the information of the "archives".23 In the hands of both, it is "the word" that animates "the man."
Less theoretically, the very metaphors that Dawkins' uses to explain the permanence of the genetic language--the distinction between digital and analogue--gain their explanatory power because of the prevalence of digital information in late twentieth century. As Dawkins observed recently, "[c]omputers are by far the best metaphor for lots of things, because they are so immensely complicated."24 Perhaps this quotation would be in even greater accord with Dawkins' intentions had it stated that "computers are useful because they make highly complicated circumstances look simple."
Each thinker also inherits challenges that result from their modus operandi. Dawkins must continue to cultivate the richness of explanation one needs for understanding higher level systems-much as he demonstrated in his wonderful book The Blind Watchmaker (1986). Gould must continue to find ways to articulate complexity and richness in a compelling and readable fashion without "dumbing-down" his ideas to the level popular writing too often promotes.
What possible lesson can we learn from our hypertext trip across the "G files"? First of all, the very public nature of the debate is informative. If the public can see such high profile evolutionary theorists engage each other on questions of fundamental importance as the mechanism of evolution, then perhaps they can gain a greater insight into the workings of science itself. The reason that Gould and Dawkins don't agree is that they often concentrate on different levels of analysis and different scales of biological complexity. Dawkins, in his "hyper" postmodern zeal, has shed his skin and uncovered the bytes of information competing in a digital information stream; Gould looks from the skin out and recognizes the importance of pluralism and tolerance in maintaining stable macro-structures. The question remains, however, how do we easily chart a means for jumping from levels of explanation without loosing too much in the process?
Although computer science and the application of information theory to living systems has been incredibly fruitful, we must not forget the importance of interactions that are lost when the explanation drops to the lowest common denominator. This is the beauty of San Marco's spandrel--it demonstrates what is lost when one jumps too quickly to explanations on a lower level of organizational complexity. If one is interested only in the mosaics one fails to glimpse the importance of the architectural form; if one is only interested in the architectural structure one misses the beautiful surface. This is a professional as well as a theoretical issue. Studying the surface and studying the structure of spandrels require different levels of analysis supported by different professional traditions, as does the study of interaction among genes versus the interactions among individuals and populations. Consequently, a higher level synthesis will need to incorporate a tolerance for other professional traditions. As the philosopher of biology Elliot Sober has reminded us "There is no a priori reason to prefer lower-level selection hypothesis over higher-ones. This preference is not a direct consequence of 'logic' or 'the scientific method'..."25
Unfortunately, Dawkins' postulations too often assume a methodological homogeneity for all explanatory levels. Granted he has contributed to the important application of information theory to evolutionary theory, but his distinction between digital and analogue information and his avowed preference for explanations relying on digital information systems often threaten to turn his ideas into merely provocative simplifications. As we have already seen, the gene contains important hierarchies of information above its linear text. What is needed is an approach that helps us identify the information embedded in these levels of explanation and its importance in the evolution of organic forms. Most likely this approach will embrace the gene in its complex material manifestations and not just as a theoretical postulate. Perhaps unfairly, I refer Dawkins to Alan Sokal's admonition to his colleagues in the humanities "[i]f all is discourse and 'text,' then knowledge of the real world is superfluous..."26
One thing we can be sure of--we should never expect nor want Gould and Dawkins to agree. For a successful scientific discipline should not be measured by the amount of agreement among its members, but the total number of questions that the discipline allows to be asked. And this is the final and perhaps the most intriguing lesson from the "G files." As an open electronic file holding the ideas of each thinker in their own words, it has brought together two disparate views of evolution under a single associative link without reducing one to the other. It is a platform for promoting cross talk and, hopefully, a contributor to a social structure where some higher level synthesis may eventually take place. Perhaps the most appropriate way to think of this hypertext link is as an informational spandrel--where a full appreciation of its scientific and social utility is not gained by reducing its form to its content, but by appreciating the side-consequences that eventually might arise from the structured intellectual hierarchy. Thus although this link is a product of computer science and is found on the page of an admirer of Richard Dawkins', it functions closer to the multi-disciplinary pluralism exemplified in the best writings of Stephen Jay Gould.
I would like to take this opportunity to thank a number of "Stanford voices" important for this paper: Tim Lenoir for inspiration, ideas, and showing me what it means to be a good scholar; Henry Lowood, for having the right reference; Peter Godfrey-Smith for forcing me to thinking about the philosophical issues behind evolutionary debates; and Hans Ulrich Gumbrecht for exposing the "monsters in my closet."
Phillip Stevens Thurtle is completing a Ph.D. dissertation titled Inheritance Incorporated: Information Exchange and the Growth of Genetic Reasoning in the History Department at Stanford University. He is currently Lecturer in Communications and the Comparative History of Ideas at the University of Washington.
Top of Page || Home || Stanford University Libraries || Stanford University