The Structure of Evolutionary Theory

by Stephen Jay Gould

  Darwin begins by stating his ad hoc assumption: “As m1 tends to inherit all the advantages of its parent M [see right side of Fig. 3-6], with the additional advantage of enduring somewhat more drought, it will have an advantage over it, and will probably first be a thriving local variety, which will spread and become extremely common and ultimately, supplant its own parent” (in Stauffer, ed., 1975, p. 239).

  But Darwin immediately senses a problem and recognizes that descendants might not retain the parental range, and that ancestors might survive the on­slaught of their phyletic children by living in a different station, thereby avoiding competition. “If m1-10 had been produced, capable of enduring more [Page 246] drought, but not at the same time enduring an equal amount of moisture with the parent M, both parent and modified offspring might co-exist: the parent (with perhaps a more restricted range) in the dryer stations, and m1-10 in the very driest stations” (ibid., p. 240).

  Darwin then rejects this reasonable but debilitating scenario with his ad hoc assumption, though he senses the weakness of his proposal and salvages his argument almost apologetically, especially at the end:

  In the imaginary case of the varieties m1-10 which are supposed to inherit all the characters of M, with the addition of enduring more drought; these varieties would inhabit stations, where M could not exist, but in the less dry stations m1-10 would have very little power of supplanting their parent M; nevertheless during unusually dry seasons m1-10 would have a great advantage over M and would spread; but in damper sea­sons M would not have a corresponding advantage over m1-10 for these latter varieties are supposed to inherit all the characters of their parent. So there would be a tendency in m1-10 to supplant M, but at an exces­sively slow rate. It would be easy to show that the same thing might oc­cur in the case of many other new characters thus acquired; but the sub­ject is far too doubtful and speculative to be worth pursuing (ibid., pp. 241-242).

  Species selection based on propensity for extinction

  Relentlessly probing as usual, Darwin now identifies another weak point in his argument. If divergence follows this predictable and necessary pattern, given the propensity of natural selection to favor extreme variants in all direc­tions, then what prevents this inexorable process from reaching absurdity in a final state of such precise and extended diversification that each species con­tains but a single individual? (This issue became important for Darwin when he moved from his earlier allopatric view of speciation to embrace a largely sympatric model with an intrinsic and predictable “motor” for the generation of species by selection of extreme variants): “But if the time has not yet ar­rived, may it not at some epoch come, when there will be almost as many spe­cific forms as individuals? I think we can clearly see that this would never be the case” (in Stauffer, ed., 1975, p. 247).

  Darwin proposes three reasons for nature's avoidance of such an absurd outcome, the first conventional, but the second and third invoking species se­lection on population properties of size and variability. For the first reason, Darwin cites ecological notions that have since become standard — limiting similarity and a restricted number of “addresses” in the economy of nature. Diversity does beget more diversity, and the physical environment sets no strict a priori upper bound. But limits imposed by “inorganic conditions” will eventually cause selection to rein in the intrinsic process of ever finer diversi­fication:

  Firstly, there would be no apparent benefit in a greater amount of modification than would adapt organic beings to different places in the polity [Page 247] of nature; for although the structure of each organism stands in the most direct and important relation to many other organic beings, and as these latter increase in number and diversity of organization, the conditions of the one will tend to become more and more complex, and its descen­dants might well profit by a further division of labor; yet all organ­isms are fundamentally related to the inorganic conditions of the world, which do not tend to become infinitely more varied (ibid., p. 247).

  But Darwin also recognizes that ecological limits may not be sufficient to restrain diversification, and he advances two other mechanisms, based on species selection against traits of small populations. First, Darwin argues that long before diversity reaches a physical limit (as quoted just above), species selection against finely divided taxa with consequently small populations will balance ordinary natural selection for further diversification: “If there exist in any country, a vast number of species (although a greater amount of life could be supported) the average number of individuals of each species must be somewhat less than if there were not so many species; and any species, repre­sented by but few individuals, during the fluctuation in number to which all species must be subject from fluctuations in seasons, number of enemies, etc., would be extremely liable to total extinction” (ibid., pp. 247-248).

  Finally, Darwin cites another reason for species selection against popu­lations of unusually small size. Such populations are not only more prone to extinction; they are also less subject to further speciation because such a re­stricted number of individuals per population implies an insufficiency of op­portunity for the origin of rare favorable variants: “Lastly we have seen . . . that the amount of variations, and consequently of variation in a right or ben­eficial direction for natural selection to seize on and preserve, will bear some relation within any given period, to the number of individuals living and lia­ble to variation during such period: consequently when the descendants from any one species have become modified into very many species, without all be­coming numerous in individuals, . . . there will be a check amongst the less common species to their further modification” (ibid., p. 248).

  In a lovely closing metaphor, Darwin provides a fine description of the cen­tral hierarchical concept of balance produced by negative feedback between levels. Number of species will equilibrate at a stable level of diversification when positive selection at the organismal level (Darwin's argument for the advantages of extreme variants) becomes balanced by negative selection at the species level (disadvantages of small population size): “the lesser number of the individuals,” Darwin writes, “serving as a regulator or fly-wheel to the increasing rate of further modification, or the production of new specific forms” (ibid., p. 248).

  In summary, I have documented both Darwin's discomfort with his forced attempt to explain the primary species-level phenomenon of diversity by nat­ural selection of extreme organismal variants, and his inability to complete the argument without an explicit invocation of species selection against taxa with small populations. His attempt to render his “principle of divergence” [Page 248] by organismal selection alone founders on three bases, the first two negative in logical barriers forced by the premise of organismal exclusivity, and the third positive in a potential “rescue” sought by acknowledging a necessary role for species selection: (1) He promotes a calculus of ultimate organismal success in terms of number of descendant taxa, but cannot extend his argu­ment for diversification by selection of extreme variants to achieve the re­quired perfect transfer to the species level. (2) In order to explain trends, he backs himself into a contradictory and ad hoc explanation for the elimination of ancestral forms by natural selection in competition with descendants. (3) Faced with the logical dilemma of runaway diversification under natural se­lection alone, he advocates negative species selection based upon small popu­lation sizes to bring the process of divergence into equilibrium.

  POSTSCRIPT: SOLUTION TO THE PROBLEM OF THE

  “DELICATE ARRANGEMENT”

  The power of a new framework often becomes most apparent in its capa­city to solve small and persistent puzzles. I therefore end this section with a solution to an old conundrum, and with another refutation for Brakeman’s (1980) claim that Darwin pinched the principle of divergence from Wallace, and then lied to cover up the theft. What did Wallace say about divergence? Did he really develop the principle in useful ways that Darwin had not an­ticipated, and might have coveted as hi
s own? When we turn to Wallace's celebrated Ternate paper (sent to Darwin in 1858), we find only a cursory statement about divergence. Wallace only discusses anagenetic trends of de­scendants away from ancestors. He does not even consider the production of multiple taxa from single sources: “But this new, improved, and populous race by itself, in course of time, gives rise to new varieties, exhibiting several diverging modifications of form, any of which, tending to increase the facili­ties for preserving existence, must, by the same general law, in their turn become predominant. Here, then, we have progression and continued diver­gence deduced from the general laws which regulate the existence of animals in a state of Nature, and from the undisputed fact that varieties do frequently occur” (Wallace, 1858, in Barrett et al., 1987).

  We must conclude that Wallace regarded a principle of divergence as “no big deal.” He grasped the idea in outline and apparently found no problem therein. His short statement could not possibly have taught Darwin anything useful, for Darwin had already carried the argument far beyond this basic comment. Why then did Wallace fail to share Darwin's puzzlement, excitement and sense of complexity about the principle of divergence? I can imagine two explanations. Either Wallace simply didn't think the issue through to all the difficulties and implications that Darwin recognized. (After all, malarial fits on Ternate are less conducive to deep thought than years of protracted strolling around the sandwalk at Down.) Or he did think the issue through and, finding nothing problematical, therefore devoted little attention to the subject. If the second alternative is correct, then my framework for [Page 249] considering the principle of divergence as a problem in levels of selection can resolve Wallace's puzzling lack of appreciation.

  As Kottler (1985) has shown, Wallace and Darwin were not identical peas from the pod of natural selection. They battled long and hard on several cru­cial issues, mostly involving Wallace's panselectionism vs. Darwin's more sub­tle view of adaptation (Gould, 1980d). One key area of disagreement cen­tered upon the target of natural selection. Darwin labored to work out a consistent theory that virtually restricted selection to struggles among organ­isms (see Chapter 2 for his interesting reasons). Wallace, as Kottler shows, never grasped the centrality, or even the importance, of the issue of levels and agency for a theory of natural selection. He moved from level to level as the situation seemed to imply, choosing whatever target of selection would best support his panselectionist leanings. (For example, he ascribed hybrid steril­ity to species selection in order to preserve his conviction that features of such importance must originate as active adaptations; whereas Darwin, committed to a consistent theory of organismal selection, regarded hybrid sterility as an incidental side consequence of accumulated differences arising by ordinary selection in two initially isolated lineages — see p. 131.) Thus, if Wallace ever pondered the principle of divergence to the point of recognizing an issue in levels of selection, he would not have responded, as Darwin did, with such a sustained, almost impassioned, quest for resolution. Wallace would not have identified any problem at all, for he never grasped the thorny issue of a need to specify levels in the first place. A simple statement about divergence would have sufficed — as Wallace indeed provided. Darwin, in his over reliance upon organismal selection, may never have reached the finish line in explicating the principle of divergence; but Wallace scarcely got off the blocks.

  Coda

  No one would argue that persistence in history makes anything right or even worthy — lest cruelty, murder and mayhem win our imprimatur by a misplaced criterion of longevity. Still, in the world of ideas, long pedigrees through disparate systems, and recurrence in the face of attempted avoid­ance; usually signify something about the power of an idea, or its necessary place in the logic of a larger enterprise.

  Causal — and not merely descriptive — accounts of hierarchy have infused evolutionary biology in this way from the beginning of our subject. Lamarck initiated this conceptual rubric with a version in the invalid mode of causal differences based on opposition between levels. Darwin knew what he didn't like about this style, and his theory — preserved unchanged to our own ortho­dox commitments today — sinks a strong foundation in an active rejection of hierarchy in this inherently combative mode. Weismann and Darwin him­self — the two greatest evolutionists and deepest thinkers with an explicit commitment to the single-level theory of natural selection — tried to extend the logic of this idea to encompass every important issue in evolution. Both [Page 250] became stymied, and eventually surrendered, with pleasure in Weismann's case at least, to the need for hierarchy in devising any complete system for the logic of evolutionary explanation in selectionist terms — Weismann on subcellular selection for explaining trends, Darwin on species selection for encompassing diversity. Both men accepted causal hierarchy in the modern and valid sense of similar forces working in distinct ways at different levels.

  Very few evolutionists know anything about this history, and may there­fore doubt the importance of the subject. But downgrading on this criterion would represent a great mistake rooted in the conservative premise that any­thing vital must be easily visible in all contexts. Anyone tempted to accept such a basis for dismissal should consider the conventional tales of rulers and conquerors — virtually the only subject matter of so many secondary school and undergraduate history courses — and recognize what modern scholars have taught us by probing the hitherto invisible pathways of daily life among ordinary people. Compare the overt and conventional history of diplomacy with the often more potent, but academically invisible, history of technology.

  Evolutionary thought began with hierarchy, wrongly conceived. Our canonical theory of natural selection arose as an attempted rebuttal. The most brilliant practitioners of that theory could not bring the argument to comple­tion; so both Weismann and Darwin brought hierarchy back, in a valid style this time, to render the theory of selection both coherent and comprehensive. I previously offered a choice of proverbs: you may view hierarchy as a bad penny or a pearl of great price. But hierarchy, like the poor, has always been with us (and, perhaps, shall inherit the earth as well!). This situation can only recall James Boswell's famous statement about one of Dr. Johnson's col­leagues who lamented that he had tried to be a philosopher, but had failed be­cause cheerfulness always broke through. Too many of us have tried to be good Darwinian evolutionists, and have felt discouraged because hierarchy always breaks in. I suggest that we rejoice — with good cheer — and welcome an under appreciated and truly indispensable old friend.

  [Page 251]

  CHAPTER FOUR

  Internalism and Laws of Form:

  Pre-Darwinian Alternatives

  to Functionalism

  Prologue: Darwin's Fateful Decision

  Thinking in dichotomies may be the most venerable (and ineluctable) of all human mental habits. In his Lives and Opinions of Eminent Philosophers (circa A.D. 200), Diogenes Laertius wrote: “Protagoras asserted that there were two sides to every question, exactly opposite to each other.”

  Darwin follows this tradition of dichotomy in a passage that he earmarked for special impact as the concluding paragraph of his crucial Chapter 6, “Difficulties on Theory.” I regard this passage as among the most important and portentous in the entire Origin, for these words embody Darwin's ulti­mate decision to construct a functionalist theory based on adaptation as pri­mary, and to relegate the effects of constraint (a subject that also commanded his considerable interest — see Section IV of this chapter) to a periphery of low relative frequency and subsidiary importance. Yet this passage, which should be emblazoned into the consciousness of all evolutionary biologists, has rarely been acknowledged or quoted. Darwin begins (1859, p. 206), ex­pressing his alternatives in upper case (and using the categories of the great debate between Cuvier and Geoffroy — see Section III of this chapter): “It is generally acknowledged that all organic beings have been formed on two great laws — Unity of Type, and the Conditions of Existence.”


  Conditions of Existence, of course, express the principle of adaptation — final cause or teleology to pre-evolutionists. Organisms are well designed for their immediate modes of life — and intricate adaptation implies an agent of design, either an intelligent creator who made organisms by fiat as an expres­sion of his wisdom and benevolence, or a natural principle of evolution that yields such adjustment between organism and environment as a primary re­sult of its operation. (Both Darwinian natural selection and Lamarckian re­sponse to perceived needs, for example, build adaptation as the most general consequence of their basic mode of action.)

  Darwin then continues by defining the other side of the classical dichot­omy: Unity of Type (1859, p. 206): “By unity of type is meant that fundamental [Page 252] agreement in structure, which we see in organic beings of the same class, and which is quite independent of their habits of life.” In another critically placed passage, introducing the subject of “Morphology” in Chapter 13, Darwin waxes almost poetic about unity of type (p. 434): “This is the most interesting department of natural history, and may be said to be its very soul. What can be more curious than that the hand of a man, formed for grasping, that of a mole for digging, the leg of the horse, the paddle of the porpoise, and the wing of the bat, should all be constructed on the same pattern, and should include the same bones in the same relative proportions.”