The Flamingo’s Smile

by Stephen Jay Gould

  Thus, flamingos and Cassiopea—two animals that could scarcely differ more in design and evolutionary history—share the common feature of feeding upside down. As a general message amidst the particulars, they have both redesigned conventional anatomy to match reversed life style. The flamingo’s upper bill has changed radically—in size, shape, and motion—to look and work like the lower beak of most birds. The structural top of Cassiopea’s umbrella has inverted its shape, all the better to work properly as an ecological bottom.

  Adaptation has a wonderful power to alter an anatomical design, widespread and stable among thousands of species, for the reversed requirements of an odd life style assumed by one or a few aberrant forms. Yet, we should not conclude that Darwinian adaptation to local environments has unconstrained power to design theoretically optimum shapes for all situations. Natural selection, as a historical process, can only work with material available—in these cases, the conventional designs evolved for ordinary life. The resulting imperfections and odd solutions, cobbled together from parts on hand, record a process that unfolds in time from unsuited antecedents, not the work of a perfect architect creating ab nihilo. Cassiopea co-opts a band of muscles ordinarily used in swimming and forms a raised rim to grasp the substrate. Flamingos bend their bill in a curious hump as the only topological solution to a new orientation.

  These adaptations to life upside down are not just funny facts. They help us to comprehend the solution to a major, and classical, dilemma in evolutionary theory (hence my decision to unite them in this essay). We can easily understand how flamingos and Cassiopea work; their unusual features do fit them for their unconventional lives. But how do these odd structures arise if evolution must proceed through intermediate steps (no one will seriously suggest that the first proto-flamingo turned its head upside down and then produced offspring with a complete set of complex adaptations to reversed life).

  In pre-Darwinian years of the early nineteenth century, when evolution was new, and when early exponents of such a radical idea were trying to work out its ramifying implications, two schools emerged and carried out an interesting (and largely forgotten) struggle until Darwin resolved their debate. Both sides admitted the good fit that usually exists between form and function—adaptation in its static, non-historical meaning. Structuralists, like Etienne Geoffroy Saint-Hilaire argued that form must change first and then find a function. Functionalists, like Jean Baptiste Lamarck, held that organisms must first adopt a different mode of life to trigger some sort of pressure for a subsequently altered form.

  The nature of this “pressure” inspired another famous (and better remembered, but no more important) debate. Lamarck held that organisms respond creatively to the needs imposed by their environments and then pass the resulting changes directly to offspring—“inheritance of acquired characters” in the usual jargon. Darwin argued that environments do not impose their adaptive requirements directly. Rather, those organisms that vary, by good fortune, in directions better suited to local environments leave more surviving offspring by a process of natural selection.

  Since Darwin won this argument about the nature of signals that pass from environment to organism, Lamarck has been eclipsed and still, despite many efforts by historians to set the record straight, suffers from an imposed reputation as a loser not to be taken seriously for any of his ideas.

  But Lamarck had the right answer (the same as Darwin’s) to the larger dispute between structuralists and functionalists. (He only proposed the wrong mechanism for how environment gets its message to organisms.) Geoffroy’s structuralist solution poses an obvious dilemma. If structure changes first, according to unknown “laws of form,” and then finds the environment best suited to its altered state, how can precise adaptation arise? We might allow that some very basic and general changes could precede any functional meaning or advantage—an animal might, for example, get larger and then exploit the inherent advantages of increased size. But can we seriously believe that something so complex, so multifarious, and so intimately suited for an unusual ecology as the flamingo’s bill might arise before the fact and without relationship to its usefulness—permitting the flamingo to discover only later how nicely such a beak worked upside down?

  Lamarck’s functionalist solution has an elegant simplicity accepted by nearly all evolutionists today (but usually attributed to Darwin, who also supported it. However much I revere Darwin, I want to advance a plea for recognizing this basic principle as Lamarck’s primary contribution. It does not appear as an incidental footnote in Lamarck’s Philosophie zoologique of 1809, but as a central theme of his book. Lamarck knew exactly what he was arguing and why.). Lamarck simply recognized that change of behavior must precede alteration of form. An organism enters a new environment with its old form suited to other styles of life. The behavioral innovation establishes a discordance between new function and inherited form—an impetus to change (by creative response and direct inheritance for Lamarck, by natural selection for Darwin). The protoflamingo first inverts its normal bill—and it doesn’t work very well. The proto-Cassiopea turns over, but its convex umbrella doesn’t clutch the substrate. Lamarck wrote:

  It is not the shape either of the body or its parts, which gives rise to the habits of animals and their mode of life; but it is, on the contrary, the habits, mode of life, and all the other influences of the environment, which have in course of time built up the shape of the body and of the parts of animals.

  The direct evidence for Lamarck’s solution cannot emerge from such “completed” adaptations as the flamingo’s beak or Cassiopea’s umbrella—though the inference even here becomes quite compelling (for why should flamingos, uniquely among birds, develop such a peculiar beak if not to exploit their chosen, odd environment). We must catch the process at its beginning stages—by finding upside down animals that have already altered their behavior, but not their form.

  African catfishes of the family Mochokidae include several species that characteristically swim upside down (see G. Sterba, in bibliography). Behavior has already changed radically, and we even have good hints about the triggers in some cases. (Synodontis nigriventris, for example, eats algae by grazing the undersides of leaves on water-dwelling plants.) But form has altered scarcely, if at all. A few species have reversed the usual pattern of cryptic coloration for fish swimming near the surface. The light bellies of most fish render them invisible to predators looking up through the water into sunlight above. But S. nigriventris, as its name (black belly) implies, is dark on its anatomical underside, and light on its structural top. Since this fish swims upside down, the light side lies below, as usual. Yet, beyond this switch in color, most upside-down mochokids look just like their upright relatives. Size, shape, and position of fins have not changed. The trigger (presumably recent) is behavioral. We shall wait to see what changes in form might still ensue.

  As a final point, readers might acknowledge my argument, but dismiss the examples as trivial or peripheral. We all love flamingos, and Cassiopea might prick our interest (our bodies too, if we get in the way). Mochokids are amusing in aquaria. But can we view life upside down as any more than a funny little corner of natural history? All my examples are the dead-end adaptations of a few species; can turning upside down lead to anything fundamental and expansive?

  As an important illustration from history (though almost surely an incorrect idea), life upside down once compelled attention as a leading speculation for the origin of vertebrates—the “worm that turned” theory, so to speak. Annelids and arthropods, the most complex of segmented invertebrates, develop ventral (bottom) nerve cords; the esophagus pierces the nerve cords and connects an even more ventral mouth to a central alimentary (gut) canal lying above the nerve cords. In vertebrates, the major nerve cord runs fore and aft in a dorsal (top) position, and the alimentary canal, including mouth and esophagus, lies entirely below.

  These two designs seem quite incompatible and unrelated. But, and ironically in the context o
f my contrast between structural and functional views, the greatest of all structuralists, Geoffroy Saint-Hilaire himself, noted that an annelid turned on its back would look more than a bit like a vertebrate—for the ventral nerve cord would then become dorsal and lie above the alimentary canal. In solving one problem, others emerge: the mouth now opens atop the inverted worm. Geoffroy suggested, as an ad hoc solution straining credulity, that the old mouth and nerve-piercing esophagus simply disappeared, and that an entirely new opening (the vertebrate mouth) developed below the dorsal nerve cord, connecting directly with the gut canal, and no longer piercing the nervous system. (So many other differences plague the comparison—lack of any annelid structure resembling the notocord or gill slits of vertebrates, fundamental disparities in embryological development between the two groups, for example—that the worm theory never commanded general assent, though it remained a leading contender for nearly a century.)

  Geoffroy never intended his comparison of vertebrate to inverted worm as an evolutionary speculation, but only as a structural comparison to buttress his remarkable theory that all animals shared a common architectural plan. (He also argued that the segments of an insect’s external skeleton matched our internal vertebrae—and that insects literally lived within their own vertebrae. This comparison compelled the additional and astonishing conclusion, forthrightly maintained by Geoffroy, that insect legs are vertebrate ribs.)

  Geoffroy also did not advance his comparison as a functional hypothesis about adaptation—he did not argue (as Lamarck might have done) that a worm’s innovative behavior (in turning over) triggered an adaptive pressure for redesign. Quite the contrary. As a structuralist, he contended that belly and back are meaningless terms of human invention to describe a superficial orientation utterly without significance to what really matters—abstract structural laws of form and permitted pathways of change.

  Today, we reject Geoffroy’s speculation along with his approach to form and function. Life upside down affirms Lamarck’s claim that substantial change in morphology usually arises as a consequence of behavioral triggers. The famous fourteenth-century motto of that upstart institution, New College, Oxford, seems to embody an essential truth about history as well as conduct: manners makyth man.

  2 | Only His Wings Remained

  THE CONVENTIONAL PROSE of twentieth-century science is lean and spare. But our Victorian predecessors delighted in leisurely detail, in keeping perhaps with the gingerbread on their houses and the shelves of bric-a-brac inside. Consider, for example, this extended (but most entertaining) description of sex and death in praying mantises, published by L.O. Howard in 1886:

  A few days since, I brought a male of Mantis carolina to a friend who had been keeping a solitary female as a pet. Placing them in the same jar, the male, in alarm, endeavored to escape. In a few minutes, the female succeeded in grasping him. She first bit off his left front tarsus, and consumed the tibia and femur. Next she gnawed out his left eye. At this the male seemed to realize his proximity to one of the opposite sex, and began to make vain endeavors to mate. The female next ate up his right front leg, and then entirely decapitated him, devouring his head and gnawing into his thorax. Not until she had eaten all of his thorax except 3 millimeters did she stop to rest. All this while the male had continued his vain attempts to obtain entrance at the valvules, and he now succeeded, as she voluntarily spread the parts open, and union took place. She remained quiet for 4 hours, and the remnant of the male gave occasional signs of life by a movement of one of the remaining tarsi for 3 hours. The next morning she had entirely rid herself of her spouse, and nothing but his wings remained.

  I cite this passage not merely for its style, but primarily for its substance—since it represents the first account I know of an all-time favorite among nature’s curious facts. We have all heard that some animals can live after losing large portions of themselves, but we think of them as just scraping by in such a limited state, not as improving their skills. Our cliché about “running around like a chicken with its head cut off” underscores this reasonable assumption that reduced anatomy entails diminished competence. Yet male mantises, beheaded by a rapacious mate, not only continue their act of courtship and copulation but actually perform more persistently and successfully.

  I want, as usual, to discuss the larger message behind this paramount oddity, but adequate treatment requires a long digression right back to Darwin himself. So bear with me, and we’ll eventually get back to mantises and much more of what the biological literature calls “sexual cannibalism.”

  The Descent of Man is, without doubt, Darwin’s most misunderstood book. Many people suppose that it represents Darwin’s attempt to fit the facts of human evolution into his evolutionary perspective. But no direct facts existed when he published in 1871, for besides Neanderthal (a race of our own species, not an ancestor or any form of “missing link”) no human fossils were discovered until the 1890s. Rather, the Descent of Man is an extended essay on the close biological relationship of humans with great apes and the possible modes of our physical and mental evolution from this common ancestry. But Darwin abhorred speculation; he never wrote a purely theoretical treatise. Even the Origin of Species is a compendium of facts pointing to a powerful conclusion. He would not have written a naked account of how it might have been, no matter how much he yearned to extend his evolutionary perspective to what he once called “the citadel itself”—the human mind.

  The key to the Descent of Man is its situation as a relatively short preface to a large, two-volume work, The Descent of Man and Selection in Relation to Sex. Darwin could weave wonderful and extensive tapestries about central themes—so much so that his readers often lose the core in its extensive mantling. But all his books are solutions to specific puzzles; the rest, for all its brilliance, is superstructure. The coral reef book is about historical inference from contemporary results, the orchid book about imperfect adaptation based on parts available, the worm book about large effects accumulated by successive small changes (see essay 9 in Hen’s Teeth and Horse’s Toes). But because he loved detail, Darwin tells you more than you want to know about how insects fertilize orchids or how worms pull objects into their burrows—and you easily lose the kernel, the paradox, the gem of a problem that started the whole edifice.

  The Descent of Man is a preface to such a problem. By 1871, twelve years after the Origin of Species, Darwin no longer needed to convince people of good will and mental flexibility that evolution had occurred; that battle had been won. But how does evolution work, what kind of world do we inhabit, and how can we know? Darwin’s radical message lay in his claim that the beauties and harmony of nature are all byproducts of one primary process called natural selection: organisms struggle to achieve greater personal reproductive success—in modern parlance, to pass more of their genes into future generations (since they cannot preserve their bodies)—and that is all. No overarching laws about the good of species or ecosystems, no wise and watchful regulator in the skies—just organisms struggling.

  But how can we know that the world is regulated by selection and not by some other evolutionary principle? Darwin’s answer is brilliant, paradoxical, and usually misunderstood. Do not, he cautions, rest your case on what might seem to be the most elegant expression of selection—the beautiful, optimally designed adaptations of organisms to their environments: the aerodynamic perfection of a bird’s wing or the streamlined beauty of a marlin. For good design is the expectation of most evolutionary theories (and of creationism as well, for that matter). There is nothing distinctively Darwinian about perfection. Instead, look for the oddities and imperfections that only occur if selection based on the reproductive success of individuals—and not on some other evolutionary mechanism—shapes the path of evolution.

  The largest class of such oddities includes those structures and habits that plainly compromise the good design of organisms (and the ultimate success of species) but just as clearly increase the reproductive prowess of in
dividuals bearing them. (My favorite examples are the tail feathers of peacocks and the huge, encumbering antlers of Irish elks, both adaptations in the struggle among males for access to, or acceptance by, females, but certainly not contributions to good design in the biomechanical sense.) Our world overflows with peculiar, otherwise senseless shapes and behaviors that function only to promote victory in the great game of mating and reproduction. No other world but Darwin’s would fill nature with such curiosities that weaken species and hinder good design but bring success where it really matters in Darwin’s universe alone—passing more genes to future generations.

  Darwin realized that natural selection in its usual sense—increasing adaptation to changing local environments—would not explain this large class of features evolved to secure purely reproductive benefits for individuals. So he christened a parallel process, sexual selection, to explain this crucial evidence. He argued that sexual selection might work by combat among males or choice by females: the first to produce overblown weapons and instruments of display; the second to encourage those adornments and elaborate posturings that impel notice and acceptance (the nightingale does not sing for our delectation).