The man who knew too much: Stephen Jay Gould's opus posthumous.
The Structure of Evolutionary Theory, by Stephen Jay Gould. The Belknap Press of Harvard University Press, 2002. 1,433 pages. $39.95.
Not long before publishing his first book, the young Bertrand Russell received some advice about literary technique. It came from his future brother-in-law, Logan Pearsall Smith, the aesthetic and slightly loopy brother of Alys Pearsall Smith, who became Russell's first wife. Seven years older than Bertrand, Logan had studied the classics at Balliol and hung with artists in Paris; he was an imposing if dubious source of postures and opinions. Years afterward, in an essay entitled "How I Write," Russell recollected that Alys's brother "was at that time exclusively interested in style as opposed to matter," and although Russell had opposite priorities, he was impressionable. Logan confidently offered various rules, of which Russell mentioned only a few: Place a comma after every four words; never use "and" except at the beginning of a sentence. "His most emphatic advice was that one must always rewrite," Russell remembered. "I consciously tried this, but found that my first draft was almost always better than my second. This discovery has saved me an immense amount of time." With experience, Russell found his own congenial literary methods, partly grounded in his devotion to mathematics and his early determination "to say everything in the smallest number of words in which it could be said clearly." Perfect clarity was the ultimate style. A sentence should be as lean as an equation. He would correct mistakes of substance, recasting entire passages, but never second-guess a first draft on grounds that were merely stylistic. In 1945, after half a century of steady literary output, he published A History of Western Philosophy, his romping survey of thinkers from Thales to himself, a book that's witty and terse at 836 pages. Five years later he won the Nobel Prize for literature. So much for a brother-in-law's advice.
Stephen Jay Gould, a very different kind of brilliant man, was also a very different kind of writer. But he shared one thing with Bertrand Russell: the disinclination to rewrite. Gould, who died on May 20, 2002, at the age of sixty, composed his essays and books on a typewriter. There was no delete key. He pulled pages from the carriage, ream after ream, and gave them to a secretary for typographic cleanup. Then off to the editor of the moment, who was not encouraged--so it seems--to edit.
The method worked well enough to make him a best-selling author of popular books such as Wonderful Life (about the fossil evidence of the Burgess Shale), a widely read magazine columnist for twenty-seven years, and a provocative contributor to the scientific literature of paleontology and evolutionary biology. All the while he taught at Harvard. At the time of his death, he had recently ended his gig as a columnist, but only "to move on to other scholarly and literary matters," and he showed no intention of slowing his pace before adenocarcinoma of the lung prematurely ended his life. His output was large, his breadth of knowledge vast, his interests eclectic, and his mode of composition (this is putting it politely) brisk. "I mean I don't write drafts, is what I'm saying," Gould told an interviewer last year. "I never write a second draft. I almost never shift a paragraph. I add something if something new comes up. But I'm a believer in the old-fashioned technique of outlining--that is, you don't sit down and write until you pretty much know how it goes, what the logical structure is." Writing, for Gould, was a straightforward process of saying what he knew and what he thought--and, on any chosen subject, he knew and thought quite a bit.
Gould's extraordinary strengths as a science historian and his extravagant weaknesses as a writer have never been more abundantly displayed than in The Structure of Evolutionary Theory, the elephantine opus that appeared two months before his death. It's a peculiar book, big as a toaster, heavy as a bag of bricks, and so determined to be clear that, with all its reiterations and epitomes and summaries, it's blurry. (Unlike Russell, Gould never thought to associate clarity or elegance with concision.) The text alone runs to 1,343 pages, not counting bibliography or index and excluding the glossary that it should have but doesn't. It also lacks even a single page of acknowledgments, evidently because the author, after twenty years' work on this tome, could remember no one in particular to thank. The book is dedicated to two colleagues with whom Gould co-authored notable papers, Niles Eldredge and Elisabeth Vrba, but in spirit the dedication celebrates not two people but three--"the Three Musketeers/Prevailing with panache," the third musketeer being Gould himself. What sort of person writes a gigantic book, filled with history and biology and cultural arcana, staking his personal claim to be the Second Coming of Charles Darwin, and then congratulates himself in the dedication? Well, there is no such "sort" of person. Stephen Jay Gould was like nobody else.
Gould's Structure raises a host of issues about evolutionary processes and the intellectual lineage of Darwinism--huge issues concerning the history of life on Earth and the phenomenon of biological diversity, issues relevant to anyone who has ever marveled at the shape of an orchid, worried about the destruction of tropical forests, or argued God-and-Darwin with a smart, obdurate creationist. The book also presents an entirely different set of questions: concerning what was good and what wasn't good about Gould as a writer.
Gould loved to echo Darwin, as he did in the title of his long-running column for Natural History magazine, "This View of Life." The allusion was to the final sentence of The Origin of Species: "There is grandeur in this view of life, with its several powers, having been originally breathed into a few forms or into one ..." Another of Gould's favorite Darwinian phrases comes from the first sentence of The Origin's final chapter: "As this whole volume is one long argument, it may be convenient to the reader to have the leading facts and inferences briefly recapitulated." Gould's book too, he tells us, represents "one long argument," of which the fundamental points are simple and few.
His subject is the Darwinian theory--evolution by natural selection--and its history as an evolving abstraction, derived from observable facts. "For some reason still unclear to me," Gould admits, "I always found the theory of how evolution works more fascinating than the realized pageant of its paleontological results." Although amused by dinosaurs, although expert on fossil snails, he preferred the airy rustle of ideas. Having sedulously examined the theory as it was articulated by Darwin himself, in The Origin and certain other writings, Gould finds its essence in three cardinal assertions. He labels each with a one-word tag: agency, efficacy, scope. Never mind for the moment what the words subsume. He calls these the three branches of Darwinian logic, beside which all else is subsidiary and inessential. Each of the three assertions, variously modified and challenged, has played a major role in evolutionary theorizing and research from the time of Darwin's immediate successors to the present. Gould's project in The Structure of Evolutionary Theory is to illuminate the three assertions within their original Darwinian context, to trace their history throughout later thinking, and then to argue that they have been crucially revised--but not negated--by important new theoretical insights (based both on fresh data and on fresh understandings of old data) during the last third of the twentieth century. The essence of Darwinian theory, he claims, has been embraced and transmogrified by these recent insights, like a lump of coal squashed into diamond. Preeminent among the new theorists he places himself.
The three essential Darwinian assertions are easy to grasp. Agency is Gould's label for the concept that evolution by natural selection occurs at the level of organisms and not at any other level, such as that of species, or populations, or genes. Organisms are the agents that do the deeds of which natural selection keeps score. In other words, individual animals and plants and other creatures--not whole species of animals and plants and other creatures, not populations of them, and not selfish genes--compete against one another to leave a greater share of offspring and thereby achieve higher representation in subsequent generations. "Darwin's brave and single-minded insistence on the exclusivity of the organismic level," Gould says, "although rarely appreciated by his contemporaries, ranks as the most radical and most distinctive feature of his theory."
The term efficacy signifies that natural selection is the effective mechanism by which evolution occurs. This was the dangerous idea that broke open Victorian thinking when Darwin announced it in 1859. Evolution itself had been a familiar, vague notion at the time, a dim sense of lineal connectedness based on empirical observations of the natural world but unsupported by any cogent hypothesis about what caused it. No one until Darwin (and a young man named Alfred Russel Wallace, a freelance specimen collector out in the Malay Archipelago who, without Darwin's scientific and social connections, arrived at the same idea independently) had hit upon natural selection to explain how evolution works.
The third in Gould's triad of terms, scope, embodies the idea that natural selection has produced all the major patterns of life on Earth, as well as the minor ones. In scientific parlance: natural selection accounts for macroevolution as well as microevolution. The differences between this species of beetle and that one? Attributable to natural selection. The differences between a beetle and a grasshopper? Attributable to natural selection. Between a giraffe and a trilobite? The same. Natural selection reaches up and down the scale of time, side to side across the breadth of taxonomic diversity, explaining why species, living and extinct, have displayed such a wondrous range of variousness in shape, habit, complexity, and fate. Its scope is sufficient, Darwin argued, to generate the entire divergence of life-forms by incremental accretion of tiny contrasts.
These three essential principles of Darwinism also provide the organizational plan for Gould's sprawling book. He takes them one by one--over and over, until they ring more familiarly than doorbell chimes--explaining how agency, efficacy, and scope represent the three major concepts necessary for comprehending the progress of evolutionary theory from Darwin to Gould.
In his first chapter, a throat-clearing introduction that goes on for eighty-nine pages, he tells in outline what he will tell at great length in the chapters that lie ahead, and why his various points should be recognized as original and important. "Long books, like large bureaucracies, can easily get bogged down in a baroque layering of summary within summary," he admits unwisely, before proceeding to demonstrate how true that is. His second chapter is an exegesis of The Origin of Species, illuminating each of the three cardinal assertions (bong bing bong: agency, efficacy, scope) within Darwin's founding text. The rest of Gould's book falls into two halves, roughly equal in size, very different in character, interconnected like a pair of aircraft carriers bound together with old rope. The first half is an historical survey and critique of evolutionary theory from Jean-Baptiste Lamarck (the most influential pre-Darwinian evolutionist) through such famous post-Darwinians as August Weismann, Francis Galton, and Hugo de Vries, as well as such little-known but interesting nineteenth-century figures as the American paleontologist Alpheus Hyatt (like Gould, an expert on extinct snails), to the principal authors and books of the so-called Modern Synthesis of the 1930s and '40s. That synthesis, in which classic Darwinian ideas were conjoined with more recent insights from genetics, paleontology, and systematics by a handful of men (including R. A. Fisher, J.B.S. Haldane, Theodosius Dobzhansky, George Gaylord Simpson, and Ernst Mayr) yielding a broader, more robust, and more up-to-date neo-Darwinian theory, is a particular target of Gould's disapproval. After its valuable early phase, he argues, the Modern Synthesis suffered a "hardening" in later revisions and reinterpretations of the primary texts, during the 1950s and early '60s, which left it too rigid and dogmatic for anyone's good. In particular, it embraced an overzealous program of finding "adaptationist" explanations for virtually every attribute of every living creature. The alternative to such adaptationism is Gould's own firmly held view: that chance and history and some deep internal factors, not just adaptation through natural selection, play formative roles.
The book's first half features characters as well as theory. It's a portrait gallery of intellectual also-rans, showing through close but fair-minded studies of select evolutionists and their works that each of the three cardinal Darwinian assertions has been tested, over time, against non-Darwinian ideas that might have supplanted it but didn't. The second half of Structure is more abstract and prescriptive. Here Gould makes his case that, beginning With the start of his own career, each of the three assertions has been dramatically modified by auxiliary ideas, similar enough to be considered loosely Darwinian, that have left it intact but stronger. The major theoretical revision that he describes and advocates won't be "pure sweetness and light" for orthodox Darwinism, Gould warns. "Much that has been enormously comfortable must be sacrificed to accept this enlarged theory with a retained Darwinian core--particularly the neat and clean, the simple and unifocal, notion that natural selection on organisms represents the cause of evolutionary change, and (by extrapolation) the only important agent of macroevolutionary pattern." That's the takeaway message, rendered down from 1,343 pages to a few lines. Agency isn't just a matter of organism versus organism; efficacy isn't just a matter of natural selection; scope isn't just a matter of the vast accumulation of tiny, organism-level differentiations.
Creationists should take no solace (though some have tried) from Gould's critique of those Darwinian orthodoxies, since the revised theoretical edifice he offers is every bit as materialistic as Darwin's own. And it is built, he insists, on Darwin's foundation. He elaborates that point with an extended architectural metaphor, a characteristic touch of Gouldian cross-disciplinary flavoring: the Cathedral of Milan, a multi-century construction project that, as tastes and fancies changed, successively incorporated gothic, baroque, and retrogothic styles. Darwin's three cardinal assertions correspond to the cathedral's eastern end, its original core, embodying tall windows with "glorious flamboyant tracery" in the high gothic manner. Piled onto that, like Milan's later baroque lintels and pediments, are the discordant addenda of post-Darwinian speculation. Upon those (but replacing them, not just topping them--here his metaphor gets shaky), Gould erects some retro-gothic spires of his own. He ranges across two hundred years of evolutionary and preevolutionary thought, spanning the fields of paleontology, organismal biology, taxonomy, macroevolution, developmental biology, and genetics, to highlight what has been discarded, what has been saved, and what has been added. It all amounts to an encyclopedic compendium of historical facts and scientific arguments, and a very great deal of typing.
The most insistent claims that Gould makes for his revised version of Darwinian theory involve concepts that are familiar to biologists but may sound arcane to the general reader: punctuated equilibrium, hierarchical selection, exaptation, deep homology, and developmental channeling, among others. If any of those has reached your ken, it's probably punctuated equilibrium, an idea that rode into wider public view on the coattails of Gould's celebrity as a science expositor.
The gist of punctuated equilibrium is that most evolutionary change takes place in relatively brief spurts, when new species diverge from old ones, and that species otherwise tend to remain unchanged throughout long stretches of time. The spurts of change punctuate the equilibrium condition of stasis. Eldredge and Gould proposed the idea and coined the name (originally they called it punctuated equilibria, but in the course of later publications it became singular) in a 1972 paper published in a paleontological journal. For a while it engaged mainly paleontologists, who were hungry for a better explanation of gaps in the fossil record where one species seems to give way too suddenly to another. Biologists who study living creatures ("neontologists," as Gould calls them) were generally less impressed. Some considered the theory wrong; or, if not wrong, trivial; or, if not trivial, derivative. Hadn't the patriarchal Ernst Mayr suggested essentially the same thing back in 1954, describing a phenomenon he termed "genetic revolution" among small, isolated populations? Maybe, but Mayr never made the idea as conspicuous or controversial as Gould and Eldredge have done in their various elaborations and defenses of it.
Nor did Ernst Mayr claim, as Gould does in his big book, that this phenomenon, however labeled, was the key to a radically revised understanding of macroevolution. If new species arise by punctuated equilibrium, Gould asserts, then the big patterns of biological diversity must result from the differential survival of species matched against one another, not the differential survival of competing individuals within species. Gould calls it "the grand analogy," this claim that species selection accounts for macroevolutionary trends in roughly the way that natural selection accounts for small evolutionary trends. He places it at the center of his own "one long argument," with a 150-page discussion of why species should be considered individual agents of selection and then a 280-page treatise on how punctuated equilibrium supports a new view of hierarchical selection and macroevolution. The logic is intricate, the test cases are intriguing, the data seem persuasive, the presentation is authoritative though unabashedly biased, and a spare summary here could never do it justice. I might as well try to show you the Congo River in a soup ladle.
And those are just two chapters among twelve. Although he was not a poet in any sense, Gould was Whitmanesque: "I am large, I contain multitudes." One of the merits of Structure is that its author was a science historian of exceptional capacities and appetites, not just a scientist and a science writer. Evolutionary biology is more an historical science than an experimental one. It proceeds mainly from found data--that is, measurements and observations of living creatures and dead specimens, close scrutiny of fossil remains--toward hypotheses about what has happened and how life works. The historical dimension is often overlooked (by narrow academic programs, for instance, that allow someone to get a Ph.D. in evolutionary biology without ever having read The Origin of Species) but not by Stephen Jay Gould. Near the start of his book, he calls himself"a paleontologist and part-time historian of science." Gould has studied Darwin's works, his life and times, with the thoroughness of a biographer. He has read Darwin's precursors, contemporaries, and successors, all the important evolutionary documents and many of the marginal ones, in their original languages, including German, French, Italian, and Latin. "I am a historian at heart," he attests near the book's end. His Structure is an epic chronicle as well as a critique of evolutionary theory, delivering high scholarship and an appreciative sense of human narrative without compromising the scientific concepts. Many full-time historians may be more knowledgeable about their immediate subjects, Gould concedes, but none except him enjoys "enough intimacy with the world of science (knowing its norms in their bones, and its quirks and foibles in their daily experience) to link this expertise to contemporary debates about causes of evolution." It's a brag, but it's true.
The best example of this strength is Gould's treatment of Etienne Geoffroy Saint-Hilaire, a French naturalist of the early nineteenth century, whose specialty was vertebrate morphology and whose chief contribution to the thinking of his day was a transcendently formalist vision of animal anatomy, suggesting that mammals, birds, reptiles, and fish were all variants of an archetypal form. Geoffroy never quite embraced a theory of evolution, but his intimations pointed the way. In his view, a parsimonious Creator had used a single underlying pattern for all the vertebrate animals. Gould outlines Geoffroy's career, including his long expedition to Egypt among Napoleon's entourage, and the development of his ideas, which came to relief in his rivalry with Georges Cuvier, a young friend who had built his own brilliant career as a morphologist while Geoffroy was gone. Besides being a better scientific politician, Cuvier was a functionalist, with a guiding vision directly opposed to Geoffroy's. Cuvier's functionalism was no more evolutionary than Geoffroy's formalism, but it differed in seeing a Creator who, unconstrained by some archetypal form, shaped animal species to their adaptational needs. The Geoffroy-Cuvier rivalry climaxed in an 1830 debate at the Academie des Sciences, which caused enough stir throughout European intellectual circles that Goethe, a serious naturalist as well as a poet, followed the reports reaching Weimar. Goethe even wrote two articles on the controversy, his sympathies tilting toward Geoffroy. Conventional wisdom has held that Cuvier won the dispute, but Gould calls it a draw, one partly settled by the fact that Cuvier died soon afterward, whereas Geoffroy lived long enough to enlist not just Goethe to his side but other influential literati. If you've ever wondered why Balzac dedicated Pere Goriot to "the great and illustrious Geoffroy Saint-Hilaire," in Gould's book you can find the answer.
The reason this old intellectual scuffle is worth reviewing, Gould explains, is that the formalism-functionalism dichotomy has never gone away. It has echoed throughout the history of evolutionary theory. The hardened adaptationist dogma of the later Modern Synthesis represented an "exaltation of functionalism," but in the rhythms of time even exalted ideas shall be humbled. Formalism, Gould hints, has recently regained credibility through unexpected discoveries in genetics and evolutionary developmental biology. Seven hundred pages later he invokes Etienne Geoffroy Saint-Hilaire, in the modern context of deep homology and developmental channeling, to make the connection between old, rejected ideas and plausible new ones. Nobody can match Gould at such gambits of historically informed scientific elucidation.
But unfortunately the book's merits are less than enough to redeem it; they're buried amid so much else. All of Gould's faults, tics, self-indulgences, ineptitudes, vanities, and infelicitous excesses as a writer are allowed to multiply. The prose alternates between chatty informality and jargon-bestrewn abstrusity, but even when informal it lacks economy. Nothing is implied, everything is stated. Cliches abound, and his habit of putting them in quotation marks, as though they were someone else's bits of bad writing, doesn't help. He has a penchant for coining cute new terms and catch-phrases as candidates for general adoption, some of which (exaptation, spandrels--for two concepts involving the interplay of accident and adaptation) are useful, but many of which are strained if not silly--such as Goldilockean solution, Cordelia's dilemma, the paradox of the visibly irrelevant, quirky functional shift, miltons, and franklins. The writing is often imprecise and clumsy, and there's far too much of it, not just because the scale of the subject is large but because the author's chief principle of organization seems to be inclusiveness. Why omit any piece of tenuously related trivia that comes to mind? Why let a sentence flow if it can be interrupted with an erudite parenthetical aside? A case in point is Gould's quotation of Ariel from The Tempest on the matter of"a sea-change," then his noting that these lines of Shakespeare's "also appear on the tombstone of the great poet Percy Bysshe Shelley (also the author of the preface to his wife's novella, Frankenstein, which cites Erasmus Darwin"--that is, Charles's famous grandfather--"in its first line of text)." And why say in few words what can be prefaced, then stated, then recapitulated in many? Why edit oneself, if oneself is Stephen Jay Gould and nobody else insists on doing it?
Once, in the dewy past, Steve Gould was a Night young scientist taking his first licks as a popular writer. His debut column for Natural History, published in January 1974, was a modest essay about animal morphology titled "Size and Shape." Although decorated with an analogy from medieval church architecture, and with a reference to Wagner's Das Rheingold, and another to The Bride of Frankenstein, it stated its premise quickly in uncluttered prose: "Animals are physical objects. They are shaped to their advantage by natural selection. Consequently, they must assume forms best adapted to their size." That piece and thirty-two others reappeared in his first collection, Ever Since Darwin, published in 1977, and I am not the only reader who remembers the book as a revelatory delight, combining jovial humanity and a rich mix of cultural interests with fascinating little excursions into evolutionary biology. The average length of an essay in Ever Since Darwin was about seven pages. Then something changed.
Gould became, as a writer, successful, overconfident, and blowsy. He found his voice, and it was garrulous. He paraded his erudition more insistently. The length of his essays increased, not so much because the topics were more complicated now but because he indulged himself in rambling, expatiating, and digressing. He knew many things about many subjects, and he seemed unwilling to leave any thought unexpressed, any subsidiary point unnoted. He could draw a quotation or a fact from diverse sources--Alexander Pope and Muhammad Ali, Izaak Walton and Walt Disney, George Eliot and Karl Marx, baseball and architecture, Gilbert and Sullivan. He seemed to have read everything, but maybe he read too quickly and too easily to appreciate those two sacred virtues of good prose: economy and grace. By 1980, as displayed in Gould's second collection of essays, The Panda's Thumb, the main features of his mature style had emerged. You could love them or loathe them. He lost some readers and gained many others. Several years later, having survived his first bout of cancer, he began writing The Structure of Evolutionary Theory. He had a fresh chance at life. This would be his big book. If he couldn't say everything, he could at least make doubly sure he said most of it. He started typing and didn't stop for two decades.
Who was the book written for? Not for you and me, but never mind. That makes it no less interesting as a literary phenomenon. When a figure with such broad public reach publishes a mammoth book--even if it's to some degree a technical work, aimed at his scientific colleagues--we mere readers are entitled to take notice. Besides, much of it is intellectual history, not argumentation about current issues in macroevolution, and therefore no more daunting to a general reader than to an evolutionary biologist who knows convergence from homology and a clade from a deme but not William Bateson from William Paley.
Then again, is The Structure of Evolutionary Theory really aimed at his scientific colleagues? He never says so flatly, but he suggests it repeatedly--voicing hope, at one point, that his combination of historiographic and scientific skills will "make my efforts useful, in a distinctive way, to my colleagues." Scientists reviewing it in their professional journals (Douglas Futuyma in Science, David Jablonski in American Scientist, David B. Wake in Nature) have tended to accept that premise, placing Structure in a category with Gould's only other technical volume, Ontogeny and Phylogeny, published twenty-five years earlier. But the big book is more stubbornly personal than Ontogeny and Phylogeny. It contains almost no mathematics, a virtue to general readers but not to population biologists and geneticists, who nowadays make their most intricate arguments with equations. By Gould's own testimony (in that late interview), his manuscript wasn't as thoroughly edited or peer-reviewed as scientific books generally are. And how many evolutionary biologists, busy with their teaching, their research, not to mention their own writing, and accustomed to the compressed prose of professional publications, have either the time or the inclination to read five hundred pages of science history as a prelude to seven hundred pages of loquacious reiteration of ideas that Gould has already made familiar in his journal articles?
I have an alternate theory, which is that this book became Gould's Finnegans Wake, an impossible thing destined to be more honored than read. Perhaps in its original conception it was, yes, a technical volume that might exert immediate influence in the chat rooms of paleontology and evolutionary biology. Or perhaps it was to be written for two audiences, a trickier challenge, educating and impressing the amateurs while making its pointed claims to the professionals. Sometime during the long decades of composition, I suspect, the author's purposes and expectations changed. The book grew, as did Joyce's last novel (a mere seventeen-year effort), in all dimensions of brilliance, erudition, ambition. It outgrew the very notion of readership. It outgrew bookish limits and gravity. Humphrey Chimpden Earwicker morphed into Here Comes Everybody. Only one reader was really capable of embracing its variousness, its magnitude, and that was the author himself. It became a prose edifice, a monument--the pyramid willed into being by the dead pharaoh. At the end, Stephen Jay Gould must have known that he was writing mainly for himself. I lived and wondered and studied and thought, he was saying. Here's the evidence.
David Quammen' s books include The Song of the Dodo, about evolution and extinction on islands, and the forthcoming Monster of God, a meditation on man-eating predators.