Leonardo’s Mountain of Clams and the Diet of Worms

by Stephen Jay Gould

  While all other Mammals have both the anterior and posterior limbs as organs of locomotion, in Man the anterior are transferred from the locomotive to the cephalic series. They serve the purposes of the head, and are not for locomotion. The cephalization of the body—that is, the subordination of its members and structure to head uses—so variously exemplified in the animal kingdom, here reaches its extreme limit. Man, in this, stands alone among Mammals.

  To buttress his second dominating idea of created type, Dana developed an idiosyncratic taxonomy within a popular pre-Darwinian genre that evolution would soon render incoherent—a numerological system, with a fixed number of subgroups in each larger group. Dana favored either two or four subgroups as the key to correct classification.

  Historians of taxonomy have often argued—quite incorrectly—that the development of evolution inspired little change in the structure of classification, for the order that had once been attributed to God could easily be shifted to evolution without any alteration of content. This claim supposedly carries the message that theories should be viewed as mental constructions sitting lightly upon nature—or, to put the argument another way, that nature’s evident factuality must be rendered in the same manner by any chosen mode of explanation. In either formulation, the role of theory, or worldview, gets demoted and the proper balance between inside theory and outside nature becomes distorted by deemphasizing the internal aspect of a truly intricate and equal pairing.

  But this common claim must be rejected. Evolution made a world of difference in classification. The major groups may have retained their definitions (arthropods are arthropods, and vertebrates vertebrates, whether created by God or developed by evolution), but a hundred other significant details had to change because the geometry of evolution differs fundamentally from the structure of created systems. Numerological schemes like Dana’s disappeared forever as soon as Darwinism triumphed—and most modern taxonomists don’t even know that such systems ever existed (and therefore fail to appreciate the power of evolution to alter the practice of their profession), because numerology suffered such a complete and sudden death. If God made all species, and their order reflects the nature of his thought, then why not search for an arcane numerological system that might embody divine wisdom? But if organisms are tied together by genealogy on an evolutionary tree of life, then success or failure becomes a question of contingent history, and no rationale for fixed numbers of subgroups within groups can possibly be devised.

  When Dana worked by twos, he divided each group into a “typical” class defining the essence, and a “hemitypical” class specifying a departure. Thus, for example, he regards terrestriality as typical for vertebrates (don’t ask me why, for fishes came first—but I do suspect an a priori desire to define the group containing humans as typical). The twofold division of vertebrates therefore contrasts Tetrapoda (all terrestrial forms) as typical, with Pisces (fishes) as hemitypical.

  When Dana worked by four, he specified three degrees of typicality in descending order—alphatypic, betatypic, and gammatypic—with a fourth group as a true departure named “degenerative.” In this version, mammals are alphatypic (for standing tall), birds betatypic (lovely creatures, but not on defining terra firma), reptiles gammatypic (as slitherers on the ground), and fishes degenerative (as living in the “wrong” place for vertebrates).

  In a revealing article, titled “Thoughts on Species” (including mineralogical as well as biological entities), and published in 1857 as Darwin composed his magnum opus, Dana defended his numerological system as embodying an unchangeable, Platonic, and universal truth:

  Fixed numbers, definite in value and defiant of all destroying powers, are well known to characterize nature from its basement to its top-stone . . . The universe is not only based on mathematics, but on finite determinate numbers in the very natures of all its elemental forces.

  He even argued that the human soul requires fixed numbers, both to avoid despondency by perceiving order, and to adore God even more. (This passage also expresses Dana’s hostility to any notion of graded evolutionary transition between groups.)

  Were these units capable of blending with one another indefinitely, they would no longer be units, and species could not be recognized. The system of life would be a maze of complexities; and whatever its grandeur to a being that could comprehend the infinite, it would be unintelligible chaos to man. The very beauties that might charm the soul would tend to engender hopeless despair in the thoughtful mind, instead of supplying his aspirations with eternal and ever-expanding truth. It would be to man the temple of nature fused over its whole surface and through its structure, without a line the mind could measure or comprehend.

  Darwin’s candid reaction to Dana’s two theories of cephalization and numerological taxonomy provides a striking illustration of the unbridgeable discordance between their worldviews. On February 17, 1863, in the interval between Dana’s key letter on evolution and Darwin’s emotional reply, Darwin also wrote to his guru and confidant Charles Lyell about his unhappiness with Dana’s paper on classification of mammals by principles of cephalization and numerical order. Darwin, with his usual insight, nails Dana for constructing a system whose absurdity would be apparent to anyone not firmly committed to ranking humans as the crown of creation. Dana’s whole scheme, Darwin correctly notes, becomes one long, forced rationale for human centrality:

  The same post that brought the enclosed brought Dana’s pamphlet on the same subject. The whole seems to me utterly wild. If there had not been the foregone wish to separate man, I can never believe that Dana or any one would have relied on so small a distinction as grown man not using fore-limbs for locomotion, seeing that monkeys use their limbs in all other respects for the same purpose as man. To carry on analogous principles . . . from crustacea to the classification of mammals seems to me madness. Who would dream of making a fundamental distinction in birds, from fore-limbs not being used at all in some birds, or used as fins in the penguin, and for flight in other birds?

  (Darwin’s valid complaint provides another example of how theory can control the arrangement of nature. Dana used the functional status of the fore-limb as a key character for making major taxonomic divisions among birds because, under his theory of cephalization, the head controls use of the forelimbs, and their status therefore defines the degree of domination by the head. Darwin, in the last sentence of his letter to Lyell, regards such a scheme as absurd because different uses of forelimbs, in his evolutionary version of life, must be interpreted as immediate adaptations to varying modes of life, not as fundamental divisions on the genealogical tree of birds.)

  As a chief ingredient in the mythology of science, the accumulation of objective facts supposedly controls the history of conceptual change—as logical and self-effacing scientists bow before the dictates of nature and willingly change their views to accommodate the growth of empirical knowledge. The paradigm for such an idealistic notion remains Huxley’s famous remark about “a beautiful theory, killed by a nasty, ugly little fact.” But single facts almost never slay worldviews, at least not right away (and properly so, for the majority of deeply anomalous observations turn out to be wrong; every fact about the revolution of the earth can be paired with a hundred claimed observations for cold fusion, perpetual motion, or the transmutation of gold).

  Rather, at least for a first approach, anomalous facts get incorporated into existing theories, often with a bit of forced stretching to be sure, but usually with decent fit because most worldviews contain considerable flexibility. (How else could they last so long, or be so recalcitrant to overthrow?) The best test for the power of a worldview to order and interpret facts—and, therefore, an excellent illustration of the fascinating and intricate interaction between theory and data in science—arises when someone discovers an absolutely pristine and unanticipated bit of novel information. Fortunately, I practice a profession—paleontology—particularly well supplied with superb test cases, for nothing can be quite
so out-of-the-blue as a newly discovered fossil. Therefore, if we consider a discovery that, in modern hindsight, points unambiguously toward the validity of a new worldview, we achieve our ideal test case: if everyone bows to the fact immediately, and accepts the implied reconstruction of nature, then Huxley’s dictum triumphs. But if most members of the old guard manage to embrace the new fact comfortably enough within their conventional worldview, then major changes of theory in science require a more complex push involving social context as well as factual impetus.

  In the early 1860s, as Darwin and Dana debated evolution in their letters, the best possible example of an unanticipated fact burst upon the scene—the discovery of Archaeopteryx, not only the oldest bird but also, apparently, beautifully intermediate between reptiles and birds in its retention of teeth, reduced coating of feathers, and basically reptilian anatomy. Score one knockout blow for evolution.

  Darwin, of course, read the discovery in exactly this sensible light. He wrote to Dana on January 7, 1863:

  The fossil bird with the long tail and fingers to its wings . . . is by far the greatest prodigy of recent times. This is a great case for me, as no group was so isolated as birds; and it shows how little we knew what lived during former times.

  But wait. The old fighter rises at the count of nine. He circles back; he feints, he bides his time; the bell rings. He rests and recoups; and he comes out fighting for the next round. In November 1863, Dana published his reply to Archaeopteryx in an article titled “On Parallel Relations of the Classes of Vertebrates, and on Some Characteristics of the Reptilian Birds.” Archaeopteryx, he proclaimed, provides no evidence for evolution, but becomes instead the best possible discovery for validating his own creationist numerology of classification based on cephalization!

  Since we are primates, and primates are visual animals, we often epitomize our worldviews in iconographic form. And nothing can be quite so powerful as a picture for summarizing and solidifying a view of life. In his article, Dana presents the classification of vertebrates as a picture, and thereby upholds a crucial role for Archaeopteryx in completing the geometry of divine numerology. As his picture shows (see the following figure), Dana wishes to classify each of the three terrestrial classes—mammals, birds, and reptiles—into his customary twofold division of typical and hemitypical. In each case, the hemitypical group should point toward the inferior class below. Mammals, of course, stand on top. Ordinary placentals form their typical group, while marsupials and the egg-laying monotremes (duck-billed platypus and echidna) build the hemitypical group below. He calls these hemitypical mammals “ooticoids”—and they clearly point to birds and reptiles in their laying of eggs.

  Reptiles also come in twos—the typical snakes, turtles, lizards, and others of that ilk; and the hemitypical frogs and salamanders among the amphibians. (We now classify amphibians as a separate class, but taxonomists of Dana’s day often put all terrestrial coldblooded creatures together into an expanded class Reptilia.) As hemitypic egg-laying mammals pointed to the reptilian class below, so also do hemitypic amphibians point to the fishes below in their initial aquatic phase of tadpole life.

  But what about birds? Now Dana encounters a problem, and a threat to the numerical beauty of his system. Flying birds are clearly typical, but what birds can be called hemitypical? Hemitypes must point to the class below—in this case, again to fishes. One might label flightless ostriches and emus as hemitypical, but in what possible way can these creatures be pointing to fishes? On the contrary, their terrestrial life seems to point upward to mammals (or at least sideways to reptiles), and therefore threatens the entire system. Imagine, then, Dana’s delight in the discovery of Archaeopteryx—for he could argue that this fossil’s retention of teeth leads downward to the hemitypic shark, which, as Mack the Knife so pointedly noted, has “pretty teeth, dear, and he shows them pearly white.” Almost gleefully, therefore, Dana portrayed Archaeopteryx, the supposedly final messenger of evolutionary truth, as the salvation of his own numerological system of creation! For Archaeopteryx—“erpetoids” in Dana’s terminology—stands forth as the missing hemitypic bird, and Dana’s system becomes healed and whole. He wrote:

  The discovery of the Reptilian Birds has brought the general law to view, that, among the four classes of Vertebrates, ordinarily received, each, excepting the lowest, consists of, first a grand typical division, embracing the majority of its species, and secondly, an inferior or hemitypic division, intermediate between the typical and the class or classes below.

  Dana actually drew two arguments against evolution from his new icon of vertebrate classification. First, as noted above, Archaeopteryx completed a numerical geometry that could only arise by divine intent and imposition. The organization of fishes provides a second anti-Darwinian argument. Typical fishes are teleosts, the bony fishes that include almost all modern species. But fishes also include two hemitypic divisions—with the crucial difference that fish hemitypes point upward toward the higher terrestrial vertebrates rather than downward. The hemitypic sharks (Selachians on Dana’s chart) point to the hemitypic Archaeopteryx and then up to typical birds; and the hemitypical lungfishes (Ganoids on Dana’s chart) point up to the hemitypical amphibians and thence to the typical reptiles.

  The whole system is therefore rounded and complete within itself. The upper divisions point down through their hemitypical groups; while the lower divisions point up through their own hemitypes. What else but a static and created order could be so self-contained and self-defining? Dana concludes, with explicit refutation of Darwin:

  It is plain from the preceding that the subkingdom of Vertebrates, instead of tailing off into the Invertebrates, has well-pronounced limits below, and is complete within itself . . . We find in the facts no support for the Darwinian hypothesis with regard to the origin of the system of life.

  This scheme may be madness, as Darwin might have said, but it is surely divine madness.

  Historians, like most decent people, tend to be patriotic. Dana represented America’s best, and who wants to saddle him with a reputation as an old fogey and holdout against the truth of evolution? Several scholarly articles have therefore focused upon Dana’s belated, strictly minimal, and reluctant “conversion” to evolution—as first stated in the 1874 edition of his Manual of Geology and, two years later, in his last paper on cephalization. Rather defensively, Dana now holds (in the 1876 article on cephalization) that evolution may have become the preferred mode of change, but that progress by cephalization still marks the result. Dana now seems to say, “I was right about what happens, but perhaps not about how it happens. What happens is more important anyway. Pattern reveals divine intent; mechanism is only a means to an end.” In his own words:

  Whatever the types of structure in course of development, there was also a general subordination in the changes to the principle of cephalization . . . These views may hold whatever be the true method of evolution. The method by repeated creations through communications of Divine power to nature should be subordinated, as much as any other, to molecular law and all laws of growth; for molecular law is the profoundest expression of the Divine will . . . But the present state of science favors the view of progress through the derivation of species from species, with few occasions for Divine intervention. If then there has been derivation of species from species, we may believe that all actual struggles and rivalries among animals, leading to a “survival of the fittest,” must tend, as in Man, to progress in cephalization.

  But we should not cite such grudging passages as Dana’s last hurrah and ultimate redemption—though traditional interpretations have followed this route. Dana made his minimal move toward evolution in order to preserve as much as possible of his crumbling system, not as a zealous, born-again crusader who had finally seen the light. By owning evolution as a mechanism, Dana could preserve his deeper convictions about progress by cephalization.

  The “heroic” approach does great disservice to Dana’s powerful intellect, and p
erpetuates a silly doctrine of validation by redemption in late conversion to a current truth (almost like an apostate Christian who reconciles with Jesus on his deathbed and expires in grace, despite the ignominy of his former existence). Dana’s last hurrah for evolution was a little blip, not the definition of his scientific life. We should honor and respect him for the power of his lifelong view, ably and honorably defended over decades, though now judged wanting. Surely, in science, it is no sin to be wrong for good reasons.

  If we dismiss those scientists now judged wrong, only valuing them if they eventually saw the light, we will miss a grand opportunity to address one of the most elusive and portentous questions in scholarly life. What is the nature of genius; why, among brilliant people, do some make revolutions and others die in the dust of concepts whose time had begun to pass in their own day? What is the crucial difference between Darwin’s transcendent greatness and Dana’s merely ordinary greatness? (Ordinary greatness is not an oxymoronic concept, but a definition of leadership in old guards throughout history.)

  I do not know the answer to this question of questions, but we can surely specify a key ingredient. Somehow, for some reason of psyche or quirk of mind, some impetus of social life or some drive of temperament, Darwin was driven to challenge, to be fearless in bringing down an intellectual universe, to be joyful in trying out each thrilling and lovely bit of furniture in a reconstructed world. Dana, for other properties of the same attributes, could not, or dared not, abandon the traditional hope and succor of centuries: Rock of ages, cleft for me; let me hide myself in thee.