The Panda’s Thumb (3 page)

Darwin then shows how the same labellum in other orchids evolves into a series of ingenious devices to ensure cross-fertilization. It may develop a complex fold that forces an insect to detour its proboscis around and past the pollen masses in order to reach nectar. It may contain deep channels or guiding ridges that lead insects both to nectar and pollen. The channels sometimes form a tunnel, producing a tubular flower. All these adaptations have been built from a part that began as a conventional petal in some ancestral form. Yet nature can do so much with so little that it displays, in Darwin's words, “a prodigality of resources for gaining the very same end, namely, the fertilization of one flower by pollen from another plant.”

Darwin's metaphor for organic form reflects his sense of wonder that evolution can fashion such a world of diversity and adequate design with such limited raw material:

We may not be flattered by the metaphor of refurbished wheels and pulleys, but consider how well we work. Nature is, in biologist François Jacob's words, an excellent tinkerer, not a divine artificer. And who shall sit in judgment between these exemplary skills?

W
ORDS PROVIDE CLUES
about their history when etymology does
not
match current meaning. Thus, we suspect that emoluments were once fees paid to the local miller (from the Latin
molere
, to grind), while disasters must have been blamed upon evil stars.

Evolutionists have always viewed linguistic change as a fertile field for meaningful analogies. Charles Darwin, advocating an evolutionary interpretation for such vestigial structures as the human appendix and the embryonic teeth of whalebone whales, wrote: “Rudimentary organs may be compared with the letters in a word, still retained in the spelling, but become useless in the pronunciation, but which serve as a clue in seeking for its derivation.” Both organisms and languages evolve.

This essay masquerades behind a list of curious facts, but it is really an abstract discourse on method—or, rather, on a particular method widely used but little appreciated by scientists. In the stereotyped image, scientists rely upon experiment and logic. A middle-aged man in a white coat (most stereotypes are sexist), either shyly reticent, but burning with an inner zeal for truth, or else ebullient and eccentric, pours two chemicals together and watches his answer emerge in a flask. Hypotheses, predictions, experiments, and answers: the scientific method.

But many sciences do not and cannot work this way. As a paleontologist and evolutionary biologist, my trade is the reconstruction of history. History is unique and complex. It cannot be reproduced in a flask. Scientists who study history, particularly an ancient and unobservable history not recorded in human or geological chronicles, must use inferential rather than experimental methods. They must examine
modern results
of historical processes and try to reconstruct the path leading from ancestral to contemporary words, organisms, or landforms. Once the path is traced, we may be able to specify the causes that led history to follow this, rather than another, route. But how can we infer pathways from modern results? In particular, how can we be sure that there was a pathway at all? How do we know that a modern result is the product of alteration through history and not an immutable part of a changeless universe?

This is the problem that Darwin faced, for his creationist opponents did view each species as unaltered from its initial formation. How did Darwin prove that modern species are the products of history? We might suppose that he looked toward the most impressive results of evolution, the complex and perfected adaptations of organisms to their environments: the butterfly passing for a dead leaf, the bittern for a branch, the superb engineering of a gull aloft or a tuna in the sea.

Paradoxically, he did just the opposite. He searched for oddities and imperfections. The gull may be a marvel of design; if one believes in evolution beforehand, then the engineering of its wing reflects the shaping power of natural selection. But you cannot demonstrate evolution with perfection because perfection need not have a history. After all, perfection of organic design had long been the favorite argument of creationists, who saw in consummate engineering the direct hand of a divine architect. A bird's wing, as an aerodynamic marvel, might have been created exactly as we find it today.

But, Darwin reasoned, if organisms have a history, then ancestral stages should leave
remnants
behind. Remnants of the past that don't make sense in present terms—the useless, the odd, the peculiar, the incongruous—are the signs of history. They supply proof that the world was not made in its present form. When history perfects, it covers its own tracks.

Why should a general word for monetary compensation refer literally to a profession now virtually extinct, unless it once had some relation with grinding and grain? And why should the fetus of a whale make teeth in its mother's womb only to resorb them later and live a life sifting krill on a whalebone filter, unless its ancestors had functional teeth and these teeth survive as a remnant during a stage when they do no harm?

No evidence for evolution pleased Darwin more than the presence in nearly all organisms of rudimentary or vestigial structures, “parts in this strange condition, bearing the stamp of unutility,” as he put it. “On my view of descent with modification, the origin of rudimentary organs is simple,” he continued. They are bits of useless anatomy, preserved as remnants of functional parts in ancestors.

The general point extends both beyond rudimentary structures and beyond biology to any historical science. Oddities in current terms are the signs of history. The first essay of this trilogy raised the same subject in a different context. The panda's “thumb” demonstrates evolution
because
it is clumsy and built from an odd part, the radial sesamoid bone of the wrist. The true thumb had been so shaped in its ancestral role as the running and clawing digit of a carnivore that it could not be modified into an opposable grasper for bamboo in a vegetarian descendant.

In a nonbiological musing, I found myself wondering last week why
veteran
and
veterinarian
, two words with such different meanings, should have a similar root in the Latin
vetus
, or “old.” Again, an oddity suggesting a genealogical approach for its solution. Veteran presented no problem, for its root and its modern meaning coincide—no indication of history. Veterinarian turned out to be interesting. City dwellers tend to view vets as servants of their pampered dogs and cats. I forgot that the original veterinarians treated farm and herd animals (as do most modern vets, I guess—pardon my New Yorker's parochialism). The link to
vetus
is through “beast of burden”—old, in the sense of “able to take a load.” Cattle, in Latin, are
veterinae
.

This general principle of historical science should apply to the earth as well. The theory of plate tectonics has led us to reconstruct the history of our planet's surface. During the past 200 million years, our modern continents have fragmented and dispersed from a single supercontinent, Pangaea, that coalesced from earlier continents more than 225 million years ago. If modern oddities are the signs of history, we should ask whether any peculiar things that animals do today might be rendered more sensible as adaptations to previous continental positions. Among the greatest puzzles and wonders of natural history are the long and circuitous routes of migration followed by many animals. Some lengthy movements make sense as direct paths to favorable climates from season to season; they are no more peculiar than the annual winter migration to Florida of large mammals inside metallic birds. But other animals migrate thousands of miles—from feeding to breeding grounds—with astounding precision when other appropriate spots seem close at hand. Could any of these peculiar routes be rendered shorter and more sensible on a map of ancient continental positions? Archie Carr, world's expert on the migration of green turtles, has made such a proposal.

A population of the green turtle,
Chelonia mydas
, nests and breeds on the small and isolated central Atlantic island of Ascension. London soup chefs and victualing ships of Her Majesty's Navy found and exploited these turtles long ago. But they did not suspect, as Carr discovered by tagging animals at Ascension and recovering them later at their feeding grounds, that
Chelonia
travels 2,000 miles from the coast of Brazil to breed on this “pinpoint of land hundreds of miles from other shores,” this “barely exposed spire in mid-ocean.”

Turtles feed and breed on separate grounds for good reasons. They feed on sea grasses in protected, shallow-water pastures, but breed on exposed shores where sandy beaches develop—preferably, on islands where predators are rare. But why travel 2,000 miles to the middle of an ocean when other, apparently appropriate breeding grounds are so much nearer? (Another large population of the same species breeds on the Caribbean coast of Costa Rica.) As Carr writes: “The difficulties facing such a voyage would seem insurmountable if it were not so clear that the turtles are somehow surmounting them.”

Perhaps, Carr reasoned, this odyssey is a peculiar extension of something much more sensible, a journey to an island in the middle of the Atlantic, when the Atlantic was little more than a puddle between two continents recently separated. South America and Africa parted company some 80 million years ago, when ancestors of the genus
Chelonia
were already present in the area. Ascension is an island associated with the Mid-Atlantic Ridge, a linear belt where new sea floor wells up from the earth's interior. This upwelling material often piles itself high enough to form islands.

Iceland is the largest modern island formed by the Mid-Atlantic Ridge; Ascension is a smaller version of the same process. After islands form on one side of a ridge, they are pushed away by new material welling up and spreading out. Thus, islands tend to be older as we move farther and farther from a ridge. But they also tend to get smaller and finally to erode away into underwater seamounts, for their supply of new material dries up once they drift away from an active ridge. Unless preserved and built up by a shield of coral or other organisms, islands will eventually be eroded below sea level by waves. (They may also sink gradually from sight as they move downslope from an elevated ridge into the oceanic depths.)

Carr therefore proposed that the ancestors of Ascension green turtles swam a short distance from Brazil to a “proto-Ascension” on the late Cretaceous Mid-Atlantic Ridge. As this island moved out and sank, a new one formed at the ridge and the turtles ventured a bit farther. This process continued until, like the jogger who does a bit more each day and ends up a marathoner, turtles found themselves locked into a 2,000-mile journey. (This historical hypothesis does not deal with the other fascinating question of how the turtles can find this dot in a sea of blue. The hatchlings float to Brazil on the Equatorial Current, but how do they get back? Carr supposes that they begin their journey by celestial cues and finally home in by remembering the character [taste? smell?] of Ascension water when they detect the island's wake.)

Carr's hypothesis is an excellent example of using the peculiar to reconstruct history. I wish I could believe it. I am not troubled by the empirical difficulties, for these do not render the theory implausible. Can we be confident, for example, that a new island always arose in time to replace an old one—for the absence of an island for even one generation would disrupt the system. And would the new islands always arise sufficiently “on course” to be found? Ascension itself is less than seven million years old.

I am more bothered by a theoretical difficulty. If the entire species
Chelonia mydas
migrated to Ascension or, even better, if a group of related species made the journey, I would have no objection, for behavior can be as ancient and as heritable as form. But
C. mydas
lives and breeds throughout the world. The Ascension turtles represent only one among many breeding populations. Although its ancient ancestors may have lived in the Atlantic puddle 200 million years ago, our record of the genus Chelonia does not extend back beyond fifteen million years, while the species
C. mydas
is probably a good deal younger. (The fossil record, for all its faults, indicates that few vertebrate species survive for as many as ten million years.) In Carr's scheme, the turtles that made the first trips to proto-Ascension were rather distant ancestors of
C. mydas
(in a different genus at least). Several events of speciation separate this Cretaceous ancestor from the modern green turtle. Now consider what must have happened if Carr is right. The ancestral species must have been divided into several breeding populations, only one of which went to proto-Ascension. This species then evolved to another and another through however many evolutionary steps separated it from
C. mydas
. At each step, the Ascension population kept its integrity, changing in lock step with other separate populations from species to species.

But evolution, so far as we know, doesn't work this way. New species arise in small, isolated populations and spread out later. Separate subpopulations of a widely dispersed species do not evolve in parallel from one species to the next. If the subpopulations are separate breeding stocks, what is the chance that all would evolve in the same way and still be able to interbreed when they had changed enough to be called a new species? I assume that
C. mydas
, like most species, arose in a small area sometime within the last ten million years, when Africa and South America were not much closer together than they are today.

In 1965, before continental drift became fashionable, Carr proposed a different explanation that makes more sense to me because it derives the Ascension population after
C. mydas
evolved. He argued that ancestors of the Ascension population accidentally drifted on the Equatorial Current from west Africa to Ascension. (Carr points out that another turtle, the west African ridley,
Lepidochelys olivacea
, colonized the South American coast by this route.) The hatchlings then drifted to Brazil in the same east-to-west current. Of course, getting back to Ascension is the problem, but the mechanism of turtle migration is so mysterious that I see no barrier to supposing that turtles can be imprinted to remember the place of their birth without prior genetic information transmitted from previous generations.

I don't think that the validation of continental drift is the only factor that caused Carr to change his mind. He implies that he favors his new theory because it preserves some basic styles of explanation generally preferred by scientists (incorrectly, in my iconoclastic opinion). By Carr's new theory, the peculiar Ascension route evolved gradually, in a sensible and predictable fashion, step by step. In his former view, it is a sudden event, an accidental, unpredictable vagary of history. Evolutionists tend to be more comfortable with nonrandom, gradualistic theories. I think that this is a deep prejudice of Western philosophical traditions, not a reflection of nature's ways (see essays of section 5). I regard Carr's new theory as a daring hypothesis in support of a conventional philosophy. I suspect that it is wrong, but I applaud his ingenuity, his effort, and his method, for he follows the great historical principle of using the peculiar as a sign of change.