The Structure of Evolutionary Theory

by Stephen Jay Gould

  So many questions about historical influence find their best resolution in our understanding of time frames and time scales. Cuvier may have won at least a rhetorical victory in a scientific debate of great intellectual import but limited duration and public impact during two months in 1830. But this event in clock time then yielded to a literary tradition of retelling, orches­trated in large part by the sole surviving protagonist. The original technical issues evoked little interest or understanding among the chief literary retellers who, in Appel's words (1987, p. 175), “came to see Geoffroy as a heroic fig­ure, Cuvier as a paltry fact collector, and the debate as a major event in French intellectual history.”

  What version, then, should we embrace if we must address the largely [Page 312] meaningless, but endlessly fascinating, issue of “who won?” Something hap­pened in 1830; a factual basis of word and gesture once unfolded. But we cannot recover the original scene, and the actual debate would be subject to endless interpretation in any case. If later judgments and interpolations loom so large in the mythological versions learned and accepted by students for nearly two centuries since the actual events (and I use “mythological” in the primal and powerful, not the pejorative, sense) — Geoffroy's triumph among the literati, Cuvier's in most scientific accounts — then these constructions re­place the unattainable original and become an important reality in their own right.

  Finally (as I shall stress throughout this book), in the deepest sense, and by a plethora of disparate criteria, neither Geoffroy nor Cuvier could have “won” because neither man held tools of triumph in principle. Formalism and functionalism represent poles of a timeless dichotomy, each expressing a valid way of representing reality. Both poles can only be regarded as deeply right, and each needs the other because the full axis of the dichotomy oper­ates as a lance thrown through, and then anchoring, the empirical world. If one pole “wins” for contingent reasons of a transient historical moment, then the advantage can only be temporary and intellectually limited. Such an ephemeral victory did occur in the recent history of evolutionary theory — the exaltation of functionalism in the hardening adaptationism of the Mod­ern Synthesis, codified in the late 1950's and early 1960's, and marked by cel­ebrations of the Darwinian centennial in 1959 (see Chapter 7). In fact, this very historical context led me to emphasize structuralism and formalism in this book (because its insights have been neglected in modern evolutionary biology, not because formalist approaches could ever be labeled as “more true”), and to revivify the great formalist thinkers, from Geoffroy to Owen to Galton, Bateson, and Goldschmidt — not as an antiquarian indulgence, but for the current utility of their ideas. (Other ages have needed to rescue func­tionalism from equally limited formalist domination.)

  But I would not spend so much time on this endeavor for reasons of selfish and personal interest alone. Formalism resartus has been externally moti­vated by great advances in genetics and development (see Chapters 10 and 11) — a system of knowledge that requires a structure of explanation based as much on how organisms can be built, as on how they do work. We should give the last word to Goethe who, in choosing the debate between Geoffroy and Cuvier for his swansong to the world, eloquently defended the claims of both men. He wrote primarily of ideas and facts, but also of archetype and adaptation, arguing that “the more vitally these two functions of the mind are related, like inhaling and exhaling, the better will be the outlook for the sciences and their friends.”

  RICHARD OWEN AND ENGLISH FORMALISM: THE ARCHETYPE

  OF VERTEBRATES

  No formalism please, we're British

  I own a letter (see Fig. 4-12) written in October 1879 by Richard Owen and [Page 313] addressed simply: “J. Pearson Langshaw, Esq. (in or near) Lancaster.” The stamp cost only a penny, and Her Majesty's post managed to make the deliv­ery with such minimal information. Owen announced that he had a “crowlet to pluck” with his friend for not visiting on a recent excursion near his where­abouts (Owen suspected a fear of further humiliation at the checkerboard as the reason for Langshaw's avoidance). Owen then spoke of a visit to Ireland, describing a new locality for Megaceros (the “Irish Elk”), and noting that Britain had formed part of a mainland during Pleistocene low sea levels. He ended with a note of chauvinism at the height of British imperial and indus­trial expansion: “I very much enjoyed a fortnight with the Tory Member for the County of Wicklow; visited a new locality of Megaceros, confirmatory of its antiquity, and coevality with the Elephants and Rhinoceroses which roamed over the continent represented now by certain Islands that set the rest of the world to rights.”

  Yet, for all his political and institutional allegiance to his native land, Brit­ain's greatest vertebrate anatomist cast his intellectual lot with the continent that lay abreast of those “certain islands” and championed the strongest ver­sion of formalism — the theory of single generating archetypes, at least for all vertebrates — in the land so well adapted for the functionalism of Paley and the Bridgewater Treatises.* Owen sensed his incongruity and recognized that his formalist message would be better heard in France or Germany than in his own country. On the very first page of his greatest formalist monograph, On the Nature of Limbs (1849), Owen wrote: “I became fully conscious how for­eign to our English philosophy were those ideas or trains of thought con­cerned in the discovery of the anatomical truths, one of which I propose to explain on the present occasion in reference to the limbs or locomotive extremities. [Page 314] A German anatomist, addressing an audience of his countrymen, would feel none of the difficulty which I experienced” (1849, p. 1). And when Owen attempted to secure a German translation of this work, he wrote to Rudolf Wagner of Gottingen (quoted in Desmond, 1982, p. 48): “The subject is better adapted for the character of mind and thought of a German audience than for our matter of fact English.”

  As Ospovat (1981) and other scholars have shown, Owen was not the only British scientist who caught (and rode) the wave of formalist excitement emanating

  4-12. For one penny and a very approximate address, the British postal service managed to deliver Richard Owen's letter to his friend Langshaw. (Author's collection.) Note his playful, but quite chauvinistic comments about England versus the continent on the last page of the letter.

  [Page 315]

  from the continent; in such a strongly social profession as science, even the most profoundly idiosyncratic thinker lies embedded in his contemporary world (see Chapter 11 on the leading 20th century British formalist, D'Arcy Thompson). Yet, as I have emphasized throughout this chapter — for the theme must play a central role in any proper understanding of Darwin and the essence of Darwinian theory — formalism remained a minority position in Britain, poorly adapted and fundamentally alien to a culture freighted with several centuries of functionalist preference.

  As a fitting illustration of this functionalist milieu, consider Britain's pre­eminent philosopher of science in the generation of Owen and Darwin: Wil­liam Whewell. (As author of a Bridgewater Treatise, Whewell cannot be considered a neutral commentator.) The first edition (1837) of his most im­portant and comprehensive work, History of the Inductive Sciences, presents a functionalist perspective, almost as rigid and exclusive as Paley's. By the third edition (1869), however, Whewell claimed a change of heart, stating that Owen had provoked the alteration. Yet Whewell's “revised” attitude re­mains quintessential functionalist, if anything even more so because he now recognizes and understands the formalist alternative, but relegates this “new­comer” to marginality. Whewell begins by providing a fair and concise con­trast of the two schools as portrayed in the debate between Geoffroy and Cuvier — though he scarcely hides his preferences in judging functionalism as “truths which are irresistibly apparent and which may therefore be safely taken as the bases of our reasonings”:

  According to this theory [Geoffroy's], the structure and functions of animals are to be studied by the guidance of their analogy only [homology in modern parlance]; our attention is to b
e turned, not to the fitness of the organization for any end of life or action, but to its resemblance of the other organizations by which it is gradually derived from the original type ... On the other hand, the plan of the animal, the purpose of its or­ganization in the support of its life, the necessity of the functions to its existence, are truths which are irresistibly apparent, and which may therefore be safely taken as the bases of our reasonings. This view has been put forward as the doctrine of the conditions of existence: it may also be described as the principle of a purpose in organization; the struc­ture being considered as having the function for its end (1869, p. 483).

  Whewell then states the chief claim for functionalist primacy: body parts exist primarily “for” their useful action: “That the parts of the body of ani­mals are made in order to discharge their respective offices, is a conviction which we cannot believe to be otherwise than an irremovable principle of the philosophy of organization, when we see the manner in which it has con­stantly forced itself upon the minds of zoologists and anatomists in all ages” (1869, p. 489). “In the organized world,” Whewell adds (1869, p. 491), “we may and must adopt the belief, that organization exists for its purpose, and that the apprehension of the purpose may guide us in seeing the meaning of the organization.” [Page 316]

  Whewell ends with a striking musical analogy, arguing that formalism conveys a certain pleasure and appreciation, but that full delight and instruction require apprehension in terms of purpose:

  To us this doctrine [of final causes or functionalism] appears like the natural cadence of the tones to which we have so long been listening: and without such a final strain our ears would have been left craving and unsatisfied. We have been lingering long amid the harmonies of law and symmetry, constancy and development; and these notes, though their music was sweet and deep, must too often have sounded to the ear of our moral nature, as vague and unmeaning melodies, floating in the air around us, but conveying no definite thought, molded into no intelligible announcement (1869, p. 495).

  The vertebrate archetype: constraint and nonadaptation

  All biologists know that Richard Owen defined the terms analogy and homology in their modern sense, and that he made a tripartite division of the second category into general, special and serial (thus demonstrating the gen­erative and developmental, rather than the evolutionary, basis of his underly­ing concept — see Chapter 10, pp. 1070–1076, for an extensive analysis of Owen's categories in the light of modern developmental biology). With this framework, constructed specifically in the light of the formalist-functionalist debate, Owen could engage the problem that Darwin would later designate as paramount in morphology — the special homology of similar parts with di­vergent functions. “What can be more curious,” Darwin would write (1859, p. 434 and p. 112 of this book), “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.”

  Surely, Owen reasoned, these underlying structural similarities could not be explained by common utility — the category that he had designated as “analogy.” And thus, the “British Cuvier” explicitly contradicted the central belief of his eponym by denying a functional explanation for homology: “The attempt to explain, by the Cuvierian principles, the facts of special homology on the hypothesis of the subserviency of the parts so determined to similar ends in different animals — to say that the same or answerable bones occur in them because they have to perform similar functions — involve [sic] many dif­ficulties, and are opposed by numerous phenomena” (Owen, 1848, p. 73).

  Owen clearly accepts the common conceptual taxonomy of his generation, for he argues that functionalism and formalism represent the only intelligible interpretations of morphology. We rightly reject functionalism for special homology, but if we deny formalism as well, then we retain, for explanation, nothing but a stochastic “slough of despond”: “With regard to the structural correspondences manifested in the locomotive members; if the principle of special adaptation fails to explain them, and we reject the idea of these corre­spondences as manifestations of some archetypal exemplar on which it has [Page 317] pleased the Creator to frame certain of his living creatures, there remains only the alternative that the organic atoms have concurred fortuitously to produce such harmony. But from this Epicurean slough of despond every healthy mind naturally recoils” (1849, p. 40).

  Owen chooses the formalist exit from Bunyan's swamp, and calls upon the guidance of Plato to pull him out (an even classier assist than Dante's employ­ment of Virgil). Special homology can only be resolved by recognizing the common generating pattern for all specific manifestations — the Platonic ar­chetype (or general homology) behind the variety of worldly incarnations. The archetype does not denote an object or an ancestor, but an abstract generating formula, a blueprint, and a formal cause. Owen engraved his ver­sion of the vertebrate archetype upon a seal and wrote to his sister Maria in 1852, trying to explain this arcane concept in layperson's terms: “It repre­sents the archetype, or primal pattern — what Plato would have called the 'di­vine idea' on which the osseous frame of all vertebrate animals — i.e. all ani­mals that have bones — has been constructed. The motto is 'the one in the manifold,' expressive of the unity of plan which may be traced through all the modifications of the pattern, by which it is adapted to the very habits and modes of life of fishes, reptiles, birds, beasts, and human kind” (in Owen, 1894, vol. 1, p. 388).

  In 1849, Owen published his treatise, On the Nature of Limbs, originally delivered as a lecture on February 9 at the Royal Institution. I regard this book as the best expression of Owen's archetypal theory, the most interesting document ever written in English to defend this strongest version of formalist theory in biology.

  Despite the title (aptly chosen and cleverly constructed as we shall see), Owen's treatise attempts to reduce the entire vertebrate skeleton, in all its manifold variety, to a single archetypal element, multiply repeated and spe­cialized. Owen writes: “General anatomical science reveals the unity which pervades the diversity, and demonstrates the whole skeleton of man to be the harmonized sum of a series of essentially similar segments, although each segment differs from the other, and all vary from their archetype” (1849, p. 119).

  For the naming and essence of this archetypal element, Owen agrees with Geoffroy in designating the vertebra. We must conceptualize Owen's “verte­bra” not only as a spinal disc, but as a set of highly generalized elements (a central disc surrounded by various bars and rods) ripe for modification along myriad pathways. Owen's archetypal unit (Fig. 4-13) operates as an abstract blueprint of bursting potential. (For example, in the “vertebra” that makes the shoulder girdle, the pleurapophysis lateral to the centrum becomes the scapula, while the haemapophysis below forms the coracoid, and the lower­most haemal spine makes the front of the sternum.) Owen writes: “I have sat­isfactorily demonstrated that a vertebra is a natural group of bones, that it may be recognized as a primary division or segment of the endoskeleton, and that the parts of that group are definable and recognizable under all their teleological modifications, their essential relations and characters appearing [Page 318] through every adaptive mask.” (Note the strong claim — the key and continu­ing relevance of formalism in critiquing Darwinian traditions — that any spe­cialization for utility, or “teleological modification,” imposes an “adaptive mask” upon the generating archetype. In such phrases, we grasp the essential difference between formalism and functionalism. Adaptive modification, the architect of morphology in Darwinian functionalism, becomes, in formalist thought, a secondary, superficial and confusing overprint upon the underly­ing essence.)

  For anyone wishing to explain the human skeleton by genesis from a verte­bral archetype, three great groups of bones must be resolved in different ways and with varying degrees of difficulty: the vertebral column itself, the skull, and the limbs with their associated girdles. Th
e archetypal model obviously works for the vertebral column, the empirical source of the theory in the first place. The skull and the limbs therefore become crucial experiments for test­ing the model of archetypal genesis.

  The attempt to depict the skull as a profound modification of a few verte­brae substantially predates Owen (see p. 283 on Goethe's allegiance, dating from observations on a sheep's skull made in 1790). The subject had been much aired and debated, with the number of proposed vertebrae ranging from one (Dumeril in 1808) to seven (Geoffroy). The most common resolu­tion proposed four vertebrae, a number popularized by Oken and accepted by Owen. Oken had named the four elements from back to front — occipi­tal, parietal, frontal, and nasal — and he had associated each with a primary sense: auditory, lingual, ocular, and olfactory. Owen accepted these four names. He argued that lateral and ventral elements of the occipital vertebra

  4-13. Owen's picture of the ideal or archetypal vertebra, interpreted by him as the ground plan for all parts of the vertebrate skeleton. From Owen, 1849. (Au­thor's collection.)

  [Page 319]

  formed the pectoral girdle. He kept Oken's names — parietal, frontal, and na­sal — for neural halves of the three anterior vertebrae, and designated the haemal (ventral) halves by their associated structures: hyoid, mandibular, and maxillary.

  Having thus resolved one of the two problematic bone groups by tradi­tional arguments, Owen turned to the single remaining issue for completion of the archetypal research program — the explanation of limbs and associated girdles as modified vertebral parts. In this sense, The Nature of Limbs should not be read as a specialized treatise on one part of the body, but as an attempt to complete the most radical version of formalism by bringing the last out­post of the vertebrate skeleton under the vertebral archetype.