The Structure of Evolutionary Theory

by Stephen Jay Gould

  Applying the same logic to Darwin's analog, the beneficiary of life's diversification through division of labor is not the individual, or even the species, but the larger polity — or life itself through the principle of maximization. Thus we can grasp the link between division of labor as a pervasive structural principle, and Darwin's goal in application — the summum bonum of maximi­zation of life, achieved through division of labor with the larger polity, or life itself, as the beneficiary.

  The logic in this chain of reasoning also illustrates why the coupling of maximization with division of labor cannot validate Darwin's “principle of divergence.” These arguments may indicate why maximization should occur, but do not explain how such a plenteous state of nature arises. In other words, this chain of reasoning does not propose a cause for maximization. Above all else, Darwin clearly understood that his distinctive style of evolu­tionary argument demanded an explanation for any higher level phenome­non as a consequence of struggle among individual organisms for reproduc­tive success. Maximization and division of labor represent phenomenological statements about the constitution of life and ecology not claims about the efficient causes of diversity. Such statements provide, in other words, a basic [Page 232] description of diversity, but emphatically not a “principle of divergence.” Darwin's insight in his carriage did not merely systematize the notions of maximization and division of labor; Darwin had known and used these con­cepts for years. The transforming insight — the argument worthy of being called a “principle of divergence” and becoming a keystone of his entire evo­lutionary theory — occurred when Darwin recognized how he could apply his distinctive style of argument, based on organismal selection, to the higher-level phenomenology of diversity. In other words, the “principle of diver­gence” embodies Darwin's argument for how and why ordinary natural selec­tion must, as a predictable consequence, yield divergence of character, leading to multiplication of successful taxa, extinction of others, ecologic plenitude, maximization of life, and the hierarchical structure of taxonomy.

  THE GENESIS OF DIVERGENCE

  This perspective on the jelling of Darwin's principle of divergence resolves a recent debate among historians about the timing and reason for Dar­win's formulation. Proposals for timing have ranged from the late 1840's to 1858, with cardinal inspirations from biogeography (Sulloway), systematics (Limoges, Ospovat), the “botanical arithmetic” of chapter 2 in the Origin, leading Darwin to defend the greater evolutionary potential for diversifi­cation in large genera (Browne), inspiration from the arguments of politi­cal economy (Schweber) and switch from allopatric to sympatric models of speciation (Kohn). All these influences surely played their parts, for the prin­ciple of divergence calls upon a wide and complex range of convictions, spanning many years and much turmoil in Darwin's mind. But Darwin's for­mulation and formalization of the “principle of divergence” records his con­viction, and his great pleasure, that he could encompass all these ideas as predictable consequences of natural selection working by struggle among organisms — that he could, in other words, bring all the higher-level phenom­enology of maximization, division of labor, and so forth, into his own distinc­tive explanatory framework.

  Schweber dates the first full formulation to 1856 and writes (1988, p. 135): “That Darwin had the 'keystone' of the argument by January 1855 is proba­bly correct, but I would also suggest that the argument was still not com­plete in an important way — at least insofar as an explicit presentation is con­cerned. All the arguments up to that point referred to levels of descriptions above individuals: varieties, species, and higher taxa. Natural selection oper­ated on individuals, and the linkage by which diversity is accomplished had to be explicitly stated.” Using this insight, Schweber regards the fol­lowing note of September 23, 1856, as the first explicit formulation.* Just [Page 233] as unfettered individual competition yields the best social order in Adam Smith's world, so too will natural selection among organisms lead to max­imization by division of labor: “The advantage in each group becoming as different as possible, may be compared to the fact that by division of labor most people can be supported in each country. — Not only do the individuals in each group strive against the others, but each group itself with all its members, some more numerous, some less, are struggling against all other groups, as indeed follows from each individual struggling” (Dar­win, September 23, 1856, cited in Schweber, 1988). (This consistent stress on the role of individuals as primary causal agents also characterizes the writings in political economy that so influenced Darwin. For example, I omitted by ellipsis an intermediary passage in the statement from Smith quoted above on p. 124. It reads: “Although we speak of communi­ties as sentient beings; although we ascribe to them happiness and misery, desires, interests, and passions; nothing really exists or feels but individu­als. The happiness of a people is made up of the happiness of single per­sons.”)

  This proper characterization of Darwin's argument also overturned the most sensational charge ever based on the principle of divergence, and made with such attention in the public arena by Brackman (1980) — the claim that Darwin received Wallace's paper from Ternate earlier than the “official” date (June 18,1858), and then proceeded to steal the principle of divergence from him, thus formulating his complete theory by ripping off Wallace and cover­ing up the evidence. This charge, which can only be supported by ignorance of detail (see the analysis of Kohn, 1981), falls apart once we recognize Dar­win's full principle of divergence as an explanation of maximization by natu­ral selection through division of labor. Darwin clearly formulated this com­plete argument in 1856, and sent a lucid epitome to Asa Gray in 1857. Thus, a possible receipt of Wallace's paper earlier in June, or even in late May of 1858, cannot affect this chronology.

  Brackman, of course, does not deny these facts. He must therefore claim that Darwin had been spooked by Wallace for years, that he pinched the ini­tial idea of diversification from Wallace's 1855 paper, and that he then moved faster (and stealthily) when the firmer statement arrived in 1858. But if we turn to Wallace's 1855 paper, we note that this article contains nothing relevant to a principle of divergence properly defined as a set of complex arguments for linking natural selection on organisms with the phenomenol­ogy of higher levels of biological organization. At most, Wallace's 1855 paper includes a passing description of the simple property of divergence itself — a fact well recognized by Darwin, who had been noting the centrality of this theme since the transmutation notebooks of the late 1830's (see quote on p. 229). (We shall also see, at the close of this section (p. 248), that even Wallace's 1858 paper contains only a cursory statement about divergence with no hint of the central feature of bridging levels.) Brackman has con­fused the noting of a fact with the development of an explanation. He has also failed to recognize Darwin's long awareness of the fact and its impor­tance. [Page 234]

  DIVERGENCE AS A CONSEQUENCE OF NATURAL SELECTION

  In Chapter 2, I noted the radical character and intellectual power of Darwin's primary argument as embedded in the Malthusian insight about natural se­lection. In an irony that overturned the entire tradition of natural theology, Darwin held that all the higher order “harmonies” of good design and eco­logical balance arose as side consequences of a process — struggle among or­ganisms for personal reproductive success — that would demand an opposite interpretation if we sought moral messages in nature. Now, in the mid 1850's, Darwin attempted the same philosophical coup to accomplish for diversity exactly what he had done for adaptation in the initial formulation of natural selection — that is, to render a higher level “good,” the maximization of life through division of labor, as a side consequence of organismic struggle. In January 1855 (in the note that Schweber views as the genesis of the principle of divergence), Darwin takes this fateful repeat step into the philosophical radicalism of rendering higher harmonies by individual struggle (quoted in Schweber, 1988): “On Theory of Descent, a divergence is implied and I think divers
ity of structure supporting more life is thus implied ... I have been led to this by looking at heath thickly clothed by heather and a fertile meadow both crowded, yet one cannot doubt more life supported in second than in the first; and hence (in part) more animals are supported. This is the final cause but mere result from struggle (I must think out last proposition).”* [Page 235]

  So Darwin recognized in early 1855 that maximization would have to be explained by natural selection (a “mere result from struggle”); he also stated that development of such an argument would be complex and difficult (“I must think out last proposition”). The “principle of divergence of character,” or more succinctly the “principle of divergence,” emerged as the result of this intellectual labor. How, then, did Darwin finally render maximization of life as a consequence of struggle, or ordinary natural selection?

  Darwin's solution, embedded as Kohn notes (1985) in his increasing will­ingness to accept sympatric speciation, holds that natural selection will gener­ally favor the most extreme, the most different, the most divergent forms in a spectrum of variation emanating from any common parental stock. Thus, each vigorous and successful stock produces a cone of varying forms about its own modal design (see Fig. 3-5 on p. 242). If natural selection generally fa­vors extreme variants in such arrays — the core claim of the “principle of di­vergence of character” — then vigorous ancestors will generate two or more descendant taxa fanning out towards maximally different form and adapta­tion. Two sequelae now complete the argument by drawing both ecological plenitude and taxonomic structure from the principle of divergence: First, the process of divergence must continue (see Fig. 3-5), impelling each vigorous descendant to produce still more advantageous extremes — thereby entraining phyletic trends of constantly increasing specialization. (The full extension ele­vates subspecies to species, species to genera, etc. — as extreme variants prolif­erate and diversify. The taxonomic tree of life emerges as an ultimate result.) Second, descendants will, in general, be competitively superior to parents, and must therefore tend to exterminate them in competition — for the number of species cannot increase indefinitely, and some ecological mechanism for re­placement of ancestors must exist.

  In the Origin, Darwin begins, in his characteristic fashion, by analogy to artificial selection. Breeders, he argues, tend to favor extreme variants when trying to improve a stock; nature must follow suit (1859, p. 112). For the breeder's conscious aim, Darwin substitutes the natural advantages of ex­treme variants: “The more diversified the descendants from any one species become in structure, constitution, and habits, by so much will they be better enabled to seize on many and widely diversified places in the polity of nature, and so be enabled to increase in numbers” (1859, p. 112). With a botanical example, Darwin then strongly argues that divergence occurs because natural selection tends to favor extreme variants:

  We well know that each species and each variety of grass is annually sowing almost countless seeds; and thus, as it may be said, is striving its utmost to increase its numbers. Consequently, I cannot doubt that in the course of many thousands of generations, the most distinct varieties of any one species of grass would always have the best chance of succeeding and of increasing in numbers, and thus of supplanting the less distinct varieties; and varieties, when rendered very distinct from each other, take the rank of species (1859, pp. 113-114). [Page 236]

  Parental forms will then tend to succumb because natural selection favors their extreme and divergent descendants in competition: “As in each fully stocked country natural selection necessarily acts by the selected form having some advantage in the struggle for life over other forms, there will be a con­stant tendency in the improved descendants of any one species to supplant and exterminate in each stage of descent their predecessors and their original parent” (1859, p. 121).

  All evolutionists know that the Origin of Species contains only a single fig­ure. This statement has been endlessly repeated in textbooks and lectures, but the true significance of this figure remains obscure, because we nearly always misinterpret the diagram (Fig. 3-5). We read this sole figure as Darwin's basic illustration of evolution as a branching process. But Darwin did not construct his diagram for such a general purpose. Rather, he devised this unique figure to provide a surgically precise description of the principle of divergence, ac­companied by several pages of explanatory text (pp. 116-126). Note how only two species of the original array (A—L) ultimately leave descendants — the left extreme A and the near right extreme I. Note how each diversifying species first generates an upward fan of variants about its modal form, and how only the peripheral populations of the fan survive to diversify further. Note that the total morphospace (horizontal axis) expands by divergence, even though only two of the original species leave descendants. Darwin writes (1859, p. 121): “In each genus, the species, which are already ex­tremely different in character, will generally tend to produce the greatest number of modified descendants; for these will have the best chance of filling new and widely different places in the polity of nature: hence in the diagram I have chosen the extreme species (A) and the nearly extreme species (I), as those which have largely varied, and have given rise to new varieties and spe­cies.” Darwin also states that the success of extremes records the action of natural selection in its usual mode of organismic struggle: “And here the im­portance of the principle of benefit being derived from divergence of charac­ter comes in; for this will generally lead to the most different or divergent variations (represented by the outer dotted lines) being preserved and accu­mulated by natural selection” (1859, p. 117).

  THE FAILURE OF DARWIN'S ARGUMENT AND THE NEED FOR

  SPECIES SELECTION

  I apologize to readers for this laborious (though, I trust, not uninterest­ing) exposition of Darwin on divergence, but I have now reached the crux both of this argument and, in one sense, of this entire book as well. I am ad­vocating both the necessity and importance of a hierarchical expansion of the theory of natural selection, defending this position by combining the stan­dard techniques of validation in science and scholarship: empirical example, the logic of argument, and historical illustration. Why, then, should so much space be accorded to Darwin's views on divergence of character — especially since I have just documented Darwin's attempt to render this second keystone [Page 237] of his theory as a consequence of ordinary natural selection at the organismic level?

  I have proceeded in this way for a simple reason: Darwin's argument doesn't work, and he came to recognize this failure in the face of his brave at­tempt. Darwin struggled mightily to render this second keystone of his full theory by natural selection alone, but he could not carry the logic to comple­tion. He failed because his full argument demands a major contribution from species level selection (or, at the very least, strong attention to explicit sorting at the species level). I don't know that Darwin ever grasped this need in a fully explicit way, committed as he was to the exclusivity of selection on organ­isms. But he recognized the crucial difficulty at several places in his exposi­tion; and, with his usual honesty, he made his distress palpable again and again.

  I perceive his discomfort in the labored description of divergence given in the Origin — 15 pages for a few points repeated many times, first in one way, then in another, all in a book so compressed that Darwin wanted to include the word “abstract” in the title. (John Murray, his publisher, demurred on ob­vious practical grounds). I sense Darwin's malaise in the fact that for this con­cept alone (among all the complex ideas developed in the Origin) he supplied both a figure and a meticulous “caption” (as it were), running for nearly ten pages. I note his dissatisfaction in the frequent shifting of attribution within his text — from consequences of organismic struggle (his usual and distinctive argument) on the one hand, to advantages for higher level units (usually spe­cies) on the other. In my reading, these shifts cannot be interpreted as com­fortable transitions rooted in a confident reduction of higher level phen
omen­ology to lower level causality (as he had achieved in explaining adaptation by natural selection), but must instead be regarded as genuine gropings and confusions. Finally, Darwin recorded his distress in explicit exclamations of doubt — from his “I must think out [this] last proposition” of the 1855 note, to his description of divergence as “this rather perplexing subject” in the Ori­gin (1859, p. 116).

  If the founder of the non-hierarchical organismal view, this doggedly persistent, fiercely honest and brilliant thinker, tried so hard to make the canoni­cal argument of natural selection work for the central higher-level phenome­non of species diversity — and could not bring the logic of his argument to a satisfactory completion — then perhaps his failure tells us something about the necessity of hierarchical selection. Darwin, by his own formulation, faced two great issues — adaptation and diversity. He tried to render both by natural selection based on struggle among organisms — adaptation by the Malthusian insight of 1838, diversity by the principle of divergence formulated in the mid 1850's. His explanation worked well, or at least sprung no logical holes, for adaptation. But he could not carry the same argument through, despite exten­sive and valiant attempts, for diversity — the primary domain of species selec­tion, as all modern advocates hold (see, for example, Gould and Eldredge, 1988, replying to Maynard Smith's 1988 misconception).