Gould, Eight (or Fewer) Little Piggies

The great Swedish paleontologist Erik Jarvik closed his two-volume magnum opus on vertebrate structure and evolution with a telling point about pentadactyl limbs and human possibilities. He noted how many "advanced" mammals modify the original pattern by loss and specialization of digits: horses retain but one as a hoof; whales lose practically the whole hind limb. Jarvik noted that an essential coupling of a multidigited hand, fit for wing tools, with an enlarging brain, well suited to devising new and better uses for such technology, established the basis and possibility of human evolution. If the ancestor of our lineage had lost the original flexibility of the "primitive" pentadactyl limb and evolved some modern and specialized reduction, human intelligence would never have developed. In this important sense, we are here because our ancestors retained the full archetypal complement of five and had not substituted some newfangled, but ultimately more limiting, configuration. Jarvik wrote:

The most prominent feature of man is no doubt his large and elaborate brain. However, this big brain would certainly never have arisen--and what purpose would it have served--if our arm and hand had become specialized as strongly as has, for instance, the foreleg of a horse or the wing of a bird. It is the remarkable fact that it is the primitive condition, inherited from our osteolepiform ancestors [fishes immediately ancestral to tetrapods] and retained with relatively small changes in our arm and hand, that has paved the way for the emergence of man. We can say, with some justification, that it was when the basic pattern of our five-fingered hand for some unaccountable reason was laid down in the ancestors of the osteolepiforms that the prerequisite for the origin of man and the human culture arose.

I don't dispute Jarvik's general point: the retention of primitive flexibility is often a key to evolutionary novelty and radiation. But is the five-fingered limb a constant and universal tetrapod archetype, interpreted in Darwin's evolutionary way as an ancestral pattern retained in all descendant lineages?

Erik Jarvik is maximally qualified to address this question (his rationale, of course, for raising it in the first place), for he has done by far the most extensive and important research on the earliest fossil tetrapods--the bearers and perpetrators of the five-fingered archetype in any evolutionary interpretation. (Fish fins are constructed on different principles, although the lobe-finned ancestors of tetrapods built a bony architecture easily transformable to the forelimbs and hind limbs of terrestrial vertebrates. In any case, fish do not display the pentadactyl pattern, and this central feature of canonical design arises only with the evolution of the Tetrapoda.)

The oldest tetrapods were discovered in eastern Greenland by a Danish expedition in 1929. They date from the very last phase of the Devonian period, a geological interval (some 390 to 340 million years ago) often dubbed the "age of fishes" in books and museum exhibits that follow the silly chauvinism of naming time for whatever vertebrate happened to be most prominent. The Swedish paleontologist Gunnar Säve-Söderbergh collected more extensive material in 1931 and directed research until his untimely death in 1948. Erik Jarvik then took over the project and, during the 1950s, published his extensive anatomical studies of two genera that share the spotlight of greatest age for tetrapods--Ichthyostega and Acanthostega. Although no specimens preserved enough of the fingers or toes for an unambiguous count, Jarvik reconstructed the earliest tetrapods with the canonical number of five digits per limb.

Our confidence in this evidence-free assumption of an initial five began to crumble in 1984, when the Soviet palcontologist O. A. Lebedev reported that the newly discovered early tetrapod Tulerpeton, also of latest Devonian age, bore six digits on its limbs. This find led anatomist and embryologist J. R. Hinchliffe to suggest in 1989, prophetically as we have just learned, that five digits represent a secondary stabilization, not an original state. Hinchliffe entitled his article "Reconstructing the Archetype: Evolution of the Pentadactyl Limb" and ended with these words: "Restriction to the pentadactyl form may have followed an evolutionary experimental phase."

Hinchliffe's suspicion has now been confirmed--in spades. Just last September, M. I. Coates and J. A. Clack reported on new material of Ichthyostega and Acanthostega, collected by a joint Cambridge-Copenhagen expedition to East Greenland in 1987 ("Polydactyly in the Earliest Known Tetrapod Limbs," Nature, September 6, 1990, pp. 66-69). Some remarkable new specimens--a complete hind limb of Ichthyostega and a forelimb of Acanthostega--permit direct counting of digits for the first time.

In an admirable convention of scientific writing that maximizes praise for past work done well and minimizes the disturbing impact of novelty, Coates and Clack write: "The proximal region [closest to the body] of the hindlimb of Ichthyostega corresponds closely with the published description, but the tarsus [foot] and digits differ." In fact, the back legs of Ichthyostega bear, count 'em, seven toes!--with three smallish and closely bound digits corresponding to the hallux (big toe, in human terms) of ordinary five-toed tetrapods. Acanthostega departs even more strongly from a model supposedly common to all; its forelimb bears eight digits in a broad arch of increasing and then decreasing size.

The conclusion seems inescapable, and an old "certainty" must be starkly reversed. Only three Devonian tetrapods are known. None has five toes. They bear, respectively, six, seven, and eight digits on their preserved limbs. Five is not a canonical, or archetypal, number of digits for tetrapods--at least not in the primary sense of "present from the beginning." At best (for fans of pentadactyly), five is a later stabilization, not an initial condition.

Moreover, in the light of this new information, an old fact may cast further doubt on the primacy of five. The naïve "ladder of life" view depicts vertebrate evolution as a linearly ascending series of amphibian-reptile-mammal-human (with birds as the only acknowledged branch). But ladders are culturally comforting fictions, and copious branching is the true stuff of evolution. Tetrapods had a common ancestor to be sure, but modern amphibians (frogs and salamanders) represent the termini of a large branch, not the inception of a series. Moreover, no fossil amphibian seems clearly ancestral to the lineage of fully terrestrial vertebrates (reptiles, birds, and mammals), called Amniota, to honor the "amniote" egg (with hard covering and "internal pond"), the evolutionary invention that allowed, in our usual metaphors, "complete conquest of the land" or "true liberation from water." (The point is tangential to this essay, but do pause for a moment and consider the biases inherent in such common "descriptions." Why is the ability to lay eggs on land a "liberation"; why is water tantamount to slavery? Why is exclusive dwelling a "conquest"? Who is fighting for what? Such language only makes sense if life is struggling upward toward a human pinnacle--the silliest and most self-centered view of evolution that I can imagine.)

The first reptile fossils are just about as old as the first amphibians in the group that eventually yielded our modern frogs and salamanders. Thus, rather than a lad-

406