The Structure of Evolutionary Theory

by Stephen Jay Gould

  De Vries then develops this concept, of species selection as a set of guide­lines for a general theory of macroevolution. He argues that sustained evolu­tionary trends must arise by species selection for two reasons: (1) Variation among species represents the only available “fuel” for an effective process of selection. (2) Trends are clearly adaptive and accumulative, but the mutational origin of elementary species is both nonadaptive and discretely sudden. Mutations, therefore, cannot produce trends by themselves; a “higher-order” selection upon discrete mutational phenotypes must occur:

  The differentiating characteristics of elementary species are only very small. How widely distant they are from the beautiful adaptive organizations of orchids, of insectivorous plants and of so many others! Here the difference lies in the accumulation of numerous elementary characters, which all contribute to the same end. Chance must have produced them, and this would seem absolutely improbable, even impossible, were it not [Page 449] for Darwin's ingenious theory. Chance there is, but no more than any­where else. It is not by mere chance that the variations move in the re­quired direction. They do go, according to Darwin's view, in all direc­tions, or at least in many. If these include the usual ones, and if this is repeated a number of times, cumulation is possible; if not, there is simply no progression, and the type remains stable through the ages. Natural se­lection is continually acting as a sieve, throwing out the useless changes and retaining the real improvements. Hence the accumulation in appar­ently predisposed directions, and hence the increasing adaptations to the more specialized conditions of life (1905, p. 572).

  De Vries also recognizes that species selection must include two compo­nents, corresponding to birth and death biases in conventional organismic se­lection (1909a, volume 1, p. 200, and volume 2, p. 660): Species selection will favor those lineages that (1) produce more elementary species by muta­tion (birth bias), and (2) generate phenotypes fortuitously adapted to chang­ing local conditions (persistence bias).

  The need for strong birth biases, combined with the central claim (see p. 435) that periods of mutability affect only a few species (and for a very short time relative to their geological longevity), led de Vries to embrace the importance and near universality of long-term stasis within species — an argu­ment strikingly isomorphic with the apparatus that we introduced much later in developing punctuated equilibrium (Eldredge and Gould, 1972; Gould and Eldredge, 1977, 1993 — and, again, much to my chagrin for not knowing about de Vries' earlier version. The two accounts invoke entirely different principles of saltational vs. allopatric speciation, but the two arguments still employ an isomorphic logic).

  De Vries cites several supports for the empirics of stasis: the ability of systematists to define most taxa unambiguously; the persistence of identical phenotypes for centuries in populations that have become widely isolated (1909a, volume 1, p. 206); geological persistence through such epochs of ex­tensive climatic change as ice ages (1905, p. 696); documented longevity of many species through several geological periods (1905, pp. 698-699). De Vries then chides Darwinians for asserting imperceptible transmutation in the face of manifest, documented constancy. Darwinians have been driven to this inconsistency, de Vries asserts, by the gradualistic implications of their theory, but a more accurate view of evolutionary mechanisms affirms stasis as an ex­pectation, not an embarrassment to be ignored, or explained away by appeals to an imperfect fossil record:

  Many facts plead in favor of the constancy of species. This principle has always been recognized by systematists. Temporarily the current form of the theory of natural selection has assumed species to be inconstant, ever changing and continuously improved and adapted to the requirements of the life conditions. The followers of the theory of descent believed that this conclusion was unavoidable and were induced to deny the manifest fact that species are constant entities. The mutation theory gives a clue to [Page 450] the final combination of the two contending ideas. Reducing the changeability of the species to distinct and probably short periods, it at once ex­plains how the stability of species perfectly agrees with the principle of descent through modification (1905, p. 694).

  In his earlier writings (1905, 1909a), de Vries strongly supported adapta­tion as the primary result of trends forged by species selection. But he altered this conviction in later articles, and thereby came to espouse the full range of internalist critiques against Darwin (in questioning both gradualism and functionalism). In 1922, de Vries contributed a short, but highly revealing chapter to J. C. Willis' famous critique of adaptation based on the correlation of geographic distribution and geological longevity: Age and Area (Willis, 1922).

  Of course, de Vries had always opposed adaptationism for the origin of species (1922, p. 224) because selection cannot craft good design if species arise in single steps: “Specific characters have evolved without any relation to their possible significance in the struggle for life. The facts are contrary to the main principle of the selection theory of Darwin” (1922, p. 226). De Vries in­terpreted Willis' argument as a final proof for this linchpin of his theory — and he expressed delight: “The general belief in adaptation as one of the chief causes of the evolution of specific characters is best directly contradicted by the statistical studies of Willis . . . This result must be considered as the one great proof, which the mutation theory still wanted for its acceptance in the field of systematic zoology and botany” (1922, p. 227).

  But differentia of higher taxa arises by cumulation during species selection, and may therefore be adaptive. “It is a curious fact that most of the striking instances of beautiful adaptation to special forms of life are characters of gen­era and sub-genera, or even of whole families, but not of single species. Climbing plants and tendrils, insectivorous plants, desert types — submerged water plants, and numerous other instances could be adduced” (1922, p. 22). (De Vries regarded the distinction between nonadaptation in most defining traits of species, and adaptation for the differentia of higher taxa, as virtual proof for the nonselectionist origin of species by saltation and the functional origin of higher taxa by species selection.)

  Still later, however, and further inspired by Willis, de Vries reasserted even this restricted role for adaptation — as he recognized that an a posteriori func­tional correlation of form and environment (especially for broad characters of higher taxa) need not indicate adaptive fashioning under current circum­stances. Using an argument of exaptation (Gould and Vrba, 1982, and Chap­ter 11), de Vries recognized and embraced the alternative view that such char­acters arise for an immediate reason (often nonadaptive), then radiate out randomly by Willis' argument, and finally survive in environments that, by good fortune for the species, favor a set of characters originally evolved for reasons unrelated to current function:

  Everywhere in nature, in geological periods as well as at present, the morphological characters of newly originated types have no special significance [Page 451] in the struggle for life. They are not known to aid them in their initial dispersal. They may afterwards prove to be useful or useless, but this has no influence upon their evolution. Obvious instances of useful­ness occur, as a rule, only at much later periods during the wandering of the new forms, when unexpectedly they arrive in environments specially fitted for them. The usual phrase, that species are adapted to their envi­ronment, should therefore be read inversely, stating that most species are now found to live under conditions fit for them. The adaptation is not on the side of the species, but on that of the environment. In a popular way we could say that in the long run species choose their best environment (1922, pp. 226-227).

  I ran into Ernst Mayr as I was completing this chapter and asked him if he had ever met de Vries. “No,” he said, “botanists and zoologists didn't talk to each other very much in those days, and, anyway, I was a Lamarckian then.” The reasons for their failure to meet may have been non-ideological (largely generational and disciplinary), but I treasure, nonetheless, the image of the world's greatest D
arwinian at the close of the 20th century, then about the same age as de Vries at this most non-Darwinian endpoint of his career, speaking about their non-interaction. De Vries did come to inhabit a different world. Whatever his love and fealty for Darwin, de Vries expunged the guid­ing concept of natural selection from his hero's own realm of the origin of species. But de Vries then reinserted selection into the higher domain of macroevolution — at least until he eventually dropped the functional theme from this world as well. De Vries developed cogent critiques, though his al­ternative mechanism can no longer be defended. His bannings and separa­tions must now be judged as too stark. Instead, we need to expand and modify Darwin's world to a hierarchical view of selection operating differently and simultaneously at several levels of nature's individuality — and not segre­gate natural selection to exclusive operation in a single domain, whether organismal (for Darwin) or speciational (for de Vries).

  RICHARD GOLDSCHMIDT'S APPROPRIATE ROLE AS A FORMALIST EMBODIMENT OF ALL THAT PURE DARWINISM MUST OPPOSE

  “It was nearly eleven hundred, and in the Records Department, where Winston worked, they were dragging the chairs out of the cubicles and group­ing them in the center of the hall, opposite the big telescreen, in preparation for the Two Minutes Hate ... A hideous, grinding screech . . . burst from the telescreen . . . The Hate had started. As usual, the face of Emmanuel Gold­stein, the Enemy of the People, had flashed onto the screen” (George Orwell, 1984). In my own factual version of this fictional archetype, we snickered over a deluded man rather than screaming at a potentially dangerous en­emy — but the expectation of group reaction, based on little more than our ig­norance combined with the prompting of our leaders, still evokes an eerie and uncomfortable feeling of similarity in my memory. [Page 452]

  I began my graduate work at Columbia University in 1963 (after undergraduate study at small, iconoclastic Antioch College), just a few years af­ter the codification of the hardest versions of the Modern Synthesis in and around the Darwin centennial celebrations of 1959 (see Chapter 7). I had never heard of Richard Goldschmidt. Yet his name surfaced in almost every course — never with any explication of his views, but only in a fleeting and de­risive reference to something called a “hopeful monster.” Students then re­sponded with a derisive sign of recognition — as our professors seemed to ex­pect as a badge of membership in some inner circle. I found the oft-repeated exercise — one might almost have called it a ritual — offensive and demeaning, both to Goldschmidt and to any notion of my potentially independent intelli­gence.

  My memories cannot be deemed either exaggerated or idiosyncratic. Frazzetta (1975, p. 85) recalled similar experiences from another university: “No one stopped to consider whether in all of Goldschmidt's assailable proposi­tions, there existed anything worth thinking about. There was no time for such consideration as long as there was so much merry mayhem to be carried out. In my university classes, the name 'Goldschmidt' was always introduced as a kind of biological 'in joke,' and all we students laughed and snickered dutifully to prove that we were not guilty of either ignorance or heresy.” Guy Bush (1982) corroborates our memories: “When his name did come up it was inevitably in the context of 'hopeful monsters' and to the accompani­ment of subdued snickers and knowing nods. It didn't take long to learn that Richard B. Goldschmidt was not to be taken seriously as an evolutionary bi­ologist.”

  A few years later, when I unearthed Goldschmidt's Material Basis of Evolution from our library (and found much of value amidst some admitted non­sense), a senior colleague and former professor decided to check his own copy to see if he had formerly dismissed the book too harshly. He could not find the volume on his shelf, and only then remembered that he had discarded the book several years earlier as containing nothing of value!

  Every orthodoxy needs a whipping boy, but why Goldschmidt? A question of personality, perhaps? Students and colleagues tend to remember Gold­schmidt (1878-1958) as kind and even courtly, but also as arrogant and im­perious, thus fulfilling anyone's stereotypical image of Herr Doktor, the Ger­man Professor. He did indeed hold such an official and topmost status, as first director of genetics at the Kaiser Wilhelm Institute for Biology in Berlin — un­til his Jewish ancestry forced relocation to Berkeley, and the start of a sec­ond career, in the late 1930's. Viktor Hamburger told me that fellow students called Goldschmidt “the Pope,” in reference (not deference) to his imperious-ness. (This apparently anomalous title may not be so peculiar for a promi­nent, established German-speaking Jew, a group that often surpassed the av­erage Prussian in loyalty and patriotism. I can still hear the acid words of my Yiddish-speaking Hungarian grandmother, recalling the snubs of well-bred Viennese girls, after her father sent her to a Jewish school in the capital of the [Page 453] Austro-Hungarian Empire.) Goldschmidt certainly practiced his penchant for pious proclamations ex cathedra: “I am certain that in the end I shall turn out to have been right” (1960, p. 307).

  These factors of personality may have heightened his candidacy, and exac­erbated the depth of collegial reaction, but Goldschmidt surely became a whipping boy primarily, and properly, for ideological reasons. Goldschmidt's 1940 book, The Material Basis of Evolution, based on the Silliman Lectures given at Yale in 1939, became the standard text for his apostasy. We may specify several rationales, based on the major claims of this volume, for Goldschmidt's anathematized status among the synthesists.

  1. Above all, and in his characteristically uncompromising manner, Goldschmidt held that new species arose saltationally, by a mode of genetic change different in kind from the alterations that yield adaptive modification within species. (The controversial nature of this difference in “kind” identifies the key issue for a proper assessment of Goldschmidt, as we shall see.)

  2. In Goldschmidt's view, Darwin had correctly described change within species as gradual, adaptive, and diversifying — but this mode of evolution leads only to the establishment of a Rassenkreis (a polytypic species), never to the formation of a new species. True species must be separated by “bridgeless gaps.” Goldschmidt organized The Material Basis of Evolution in two se­quential sets of chapters, entitled Microevolution and Macroevolution. In a scheme of argument that could not have been “better” designed to rouse ascendant neo-Darwinians to anger, the first part extols Darwinian processes in their strictly limited domain, while the second emphasizes their impo­tence in producing new species (while proposing workable alternatives in the saltationist mode). Goldschmidt links the two sections with the following paragraph — an anti-Darwinian clarion call that he printed entirely in italics: “Subspecies are actually, therefore, neither incipient species nor models for the origin of species. They are more or less diversified blind alleys within the species. The decisive step in evolution, the first step toward macroevolution, the step from one species to another, requires another evolutionary method than that of sheer accumulation of micromutations” (1940, p. 183).

  3. Apostates may generate maximal anger, but not every opponent can gain such an anathematized status. A fool by nature, or a scholar who displays ignorance in the field of his chosen iconoclasm, cannot qualify, and will attract more pity than rage. Apostates must be smart, skilled, potentially effective (and therefore feared) — and also former adherents to the orthodoxy they now reject. Goldschmidt could not be dismissed as an ignorant “lab man,” unacquainted with the source of strongest Darwinian arguments — field data of natural history. He had undertaken one of the most thorough studies ever attempted on the empirics of geographic variation in a single species, the gypsy moth Lymantria dispar. He states that he had expected to affirm the Darwinian apparatus at all scales: “As a convinced Darwinian I believed geo­graphic races to be incipient species. I hoped to prove by such an analysis the correctness of this idea. I was completely acquainted with what twenty years [Page 454] later was rediscovered as 'the new systematics,' and my convictions, as ex­pressed in 1920 and 1923, were practically the same as those of present-day Neo-Darwini
ans” (1960, p. 318).

  If Goldschmidt had simply rejected Darwinism outright, he would have angered the synthesists quite sufficiently. But Goldschmidt proceeded further to an argument almost guaranteed to arouse far deeper frustration. He pro­claimed the selectionist mechanism as completely sufficient to account for all differentiation within species — and then announced that this basic style of microevolution bore no causal relevance to the production of new species. In other words, he denied the cardinal extrapolationist premise that evolution in the small — the only mode routinely subject to direct observation — could, by extension in time, produce the entire panoply of life's history. To many syn­thesists, Goldschmidt's ideas ranked as an ultimate council of despair. How can science proceed at all without such a uniformitarian and operational premise?

  Goldschmidt, as an enfant terrible, clearly enjoyed the fuss that he had engendered: “I certainly had struck a hornet's nest. The Neo-Darwinians re­acted savagely. This time I was not only crazy but almost a criminal” (1960, p. 324). He also provoked a vigorous and extended reaction. Most evolution­ists know, for example, that Ernst Mayr wrote one of the great classics of the Synthesis, Systematics and the Origin of Species (1942), as a direct response to Goldschmidt's Material Basis. Mayr recalled (1980, p. 420): “Even though personally I got along very well with Goldschmidt, I was thoroughly furious at his book, and much of the first draft of Systematics and the Origin of Spe­cies was written in angry reaction to Goldschmidt's total neglect of such over­whelming and convincing evidence.”