Eight Little Piggies
These extreme positions, of course, are embraced by very few thinkers. They are caricatures constructed by the opposition to enhance the rhetorical advantages of dichotomy. They are not really held by anyone, but partisans think that their opponents are this foolish, thus fanning the zealousness of their own advocacy. The possibility for consensus drowns in a sea of charges.
The central claim of each side is correct, and no inconsistency attends the marriage once we drop the peripheral extremities of each attitude. Science is, and must be, culturally embedded; what else could the product of human passion be? Science is also progressive because it discovers and masters more and more (yet ever so little in toto) of a complex external reality. Culture is not the enemy of objectivity but a matrix that can either aid or retard advancing knowledge. Science is not a linear march to truth but a tortuous road with blind alleys and a rubbernecking delay every mile or two. Our road map is not objective reality but the patterns of human thoughts and theories.
My position, as a variety of apple pie, is easy to state. It is also empty and tendentious as an abstract generality. This middle way, this golden mean, can only permeate our understanding by example. Cephalaspis provides one of the best demonstrations I know because this fish played a central role in three important and sequential views of nature’s order. Each view embodied its cultural context, but each also provided a framework for new and genuine objective knowledge about Cephalaspis. The new knowledge then helped to establish a revised view of natural order. Speaking of rhetoric in the best American tradition, culture and knowledge are rather like liberty and union—one and forever, now and inseparable.
Cephalaspis, as its name implies, enclosed its head in a thick, bony shield. Much thinner scales covered everything behind, from front fins to tail. Since the scales usually disarticulate at death and are rarely preserved at all, most fossils of Cephalaspis include only the head shield. By itself, the shield is a peculiar and decidedly unfishlike object. It looks much like the head end of many trilobites (fossil arthropods), and was so classified until Louis Agassiz established the true affinity of Cephalaspis in his great monograph Les poissons fossiles (Fossil Fishes), published in five large volumes between 1833 and 1843.
Agassiz confessed his wonder and puzzlement in his first paragraph on Cephalaspis:
These are the most curious animals that I have ever observed; their features are so extraordinary that I had to make the most careful and scrupulous examination…in order to convince myself that these mysterious creatures are really fish.
Agassiz reached the correct solution to his puzzle because his collection included some unusually well-preserved specimens, with the characteristic head shield indubitably attached to an undeniably fishy posterior (see figure). Yet while Agassiz began the modern history of Cephalaspis by placing this genus properly among the vertebrates, he could never resolve its relationship with other fishes for lack of crucial evidence. Agassiz particularly bewailed his failure to find any specimen exposing the lower surface of the head shield, where, he surmised (correctly), the mouth would be located. Thus Agassiz could never recognize the chief feature of jawlessness in Cephalaspis and could not identify the ostracoderms as structural precursors of all later vertebrates (jaws evolved from bones that supported gill arches behind the mouth of these jawless fishes). Cephalaspis, to Agassiz, remained an unplaceable oddball among fishes.
Although Agassiz could not fully resolve the status of Cephalaspis, he used this most peculiar of fishes as a linchpin for his theory of biological order. Les poissons fossiles is no simple list of old fishes; it is, perhaps most of all, a closely reasoned brief for Agassiz’s creationist world view—a theory that embodied the cultural consensus of 1830, but that Agassiz maintained doggedly to his death in 1873, long after its scientific demise in Darwin’s favor.
A figure from Agassiz’s Les poissons fossiles proving the vertebrate affinities of the head shield of Cephalaspis. Photograph by Rosamond W. Purcell.
Agassiz rooted his version of creationism in a complex analogy with his favorite subject, comparative embryology. Agassiz viewed embryonic growth as a tale of differentiation—more complex and specialized forms develop from simpler and more generalized precursors. These later specializations may proceed in several directions from a common initial form. Thus, a single (and simple) early embryo, representing a vertebrate prototype, might differentiate along several pathways into advanced fishes, reptiles, or mammals.
Agassiz then argued that the geological history of a group should match the embryological development of its latest and most advanced members. Early (geologically oldest) forms should be few, simple, and generalized; later relatives should be specialized and differentiated versions of these primordial archetypes. This scheme might sound evolutionary, but Agassiz explicitly rejected such a heresy. The geological sequence of separate creations paralleled embryological growth within each group because God’s orderly and benevolent plan permeated all developmental processes in nature.
Agassiz remained loyal to the classification of his mentor, the great French zoologist Georges Cuvier. He arranged all animals in four great groups: radiates (a hodgepodge by modern standards, but including such radially symmetrical forms as corals and echinoderms); mollusks; articulates (segmented worms and arthropods); and vertebrates. The four trunks are coequal and do not coalesce at life’s dawn, for they represent separately created plans for anatomy, not ancestors and descendants. But since geological history mimics embryological differentiation, prototypes of the four trunks from the oldest strata should be more similar than their modern representatives—for embryology is a tale of divergence from generalized roots.
Agassiz used Cephalaspis as a primary illustration of his embryological vision for geological history. As a representative of the oldest fishes, Cephalaspis fulfilled all expectations for a primordial creature in Agassiz’s vision of differentiation as the guiding principle of history. Agassiz located two different supports for his theory of differentiation in “the bizarre characters of this genus” (les caractères bizarres de ce genre). First, he viewed the single solid head shield (not divided into separate cranial bones linked by sutures) and the few, simple scales covering the body as marks of a primitive generality—a source for later differentiation of separate bones and more complex scales. Second, he twisted to his advantage the old bugbear of superficial resemblance between the shield of Cephalaspis and the head of trilobites—for this similarity indicated that the major trunks of animal life did draw closer to a common simplicity at life’s source.
Finally, Agassiz delighted in the great age of Cephalaspis, for its antiquity proved that all four trunks lived simultaneously at the dawn of life. The most complex group of vertebrates did not arise later as a possible evolutionary descendant (heaven forfend) of a simpler trunk.
Agassiz’s theory of a God ordering his creation by embryological rules of differentiation was clearly not an interpretation logically entailed by objective facts of nature. It was a vision rooted in a cultural context still unable to embrace evolution, and in the personal psychology of Agassiz’s own interests and training. Agassiz imposed his theory upon Cephalaspis and highlighted only those facts most congenial with his preferred views. Yet his use of Cephalaspis cannot be read as a vindication of relativism. Agassiz may have exploited Cephalaspis in the interests of his vision, but he also unearthed the primary fact that fueled all later discussion. He proved that Cephalaspis was a vertebrate by discovering the body of a fish behind a head shield that had confused all earlier observers.
When William Patten used Cephalaspis as the centerpiece of an important theory eighty years later, the context of science had changed irrevocably. Evolution had triumphed, and Cephalaspis would now be invoked in the interest of genealogical claims. Our cardboard relativist might argue that since Cephalaspis had played no notable part in fomenting this great revolution in thought, any evolutionary interpretation must be viewed as a new convention impressed upon old informatio
n—a new set of rules like the annual revision of Mah-Jongg hands imposed upon the same old tiles. But a realist would rightly reply that the tiles had changed as well. Agassiz had not resolved the anatomical status of Cephalaspis among the fishes; he could only affirm that the genus was both old and aberrant. Several of the greatest nineteenth-century evolutionists then studied Cephalaspis—including T. H. Huxley, E. R. Lankester, and E. D. Cope. From all their arguments and disagreements, one strong theme emerged: Cephalaspis and the ostracoderms were not just a grab bag of peculiar fishes. They formed a coherent group, with a large and consistent set of features all pointing to an anatomically primitive status among fossil vertebrates. Cephalaspis therefore became a prime candidate for theories about the ancestry of higher vertebrates. Evolution set the context, but new information about Cephalaspis fueled the debate.
Patten presented the most sophisticated case for the oldest theory of vertebrate origins—the attempt, dating to Geoffroy Saint-Hilaire in the early nineteenth century, to derive vertebrates from an inverted annelid or arthropod, a “worm that turned,” so to speak. Arthropods run their main nerve cords along their ventral (bottom) surface. The gut lies above, and the esophagus must therefore pierce through nervous tissue to end in a ventral mouth. In vertebrates, on the other hand, the main nerve tract, the spinal cord, is dorsal (on top), and the gut lies below. Turn an arthropod upside down, and you get the right order for vertebrates—nerves above guts. You also obtain a set of additional correspondences that some scientists have read as superficial and analogical, and others as deeply meaningful signs of evolutionary affinity.
But this act of inversion also produces some horrendous problems for the theory of arthropod ancestry. In particular, the vertebrate mouth does not pierce the brain and open on top of the head—though the old arthropod mouth would take this path in its supposedly inverted position. Proponents of the arthropod theory must therefore argue that this original mouth atrophied, and that vertebrates opened a new ventral version below the brain. No one has ever provided a good explanation for how such a topological transformation might plausibly occur.
A drawing from Patten’s 1912 book showing the superficial resemblance of a head shield of an ostracoderm fish (left) to a trilobite (right). Courtesy of Department of Library Services, American Museum of Natural History.
The arthropod theory, though venerable, suffered another major impediment that Patten tried to remedy. This theory was little more than an abstract argument based on a theoretical transformation without hard evidence in the form of intermediary creatures from the fossil record. Patten therefore returned to the oldest folk wisdom about Cephalaspis—the basic observation that had been judged false and treacherous ever since Agassiz. Maybe that first idea about a relationship with trilobites had some validity after all. Maybe Cephalaspis looked like an arthropod because it wasn’t all fish, as its back end seemed to proclaim. Maybe the head shield truly possessed some arthropod characters. Maybe ostracoderms did represent that long-sought intermediary group between arthropods and vertebrates.
Patten eventually argued himself into this position. He identified the marine arachnids (eurypterids and horseshoe crabs) as arthropod ancestors, and he classified ostracoderms, not as primitive jawless fishes, but as a transitional group between the two great phyla. He wrote in 1912:
We may now confidently affirm that the ostracoderms belong neither to the arthropods nor to the vertebrates, but constitute a new class standing midway between them, the ancestors of the one and the descendants of the other, the long sought missing link between the vertebrates and the invertebrates.
Patten was more than a vertebrate anatomist; he also fancied himself a philosopher and moralist. As such, he used his theory of vertebrate origins as a centerpiece for one of the widest (and wildest) claims ever made for the sweep of evolutionary theory. In a series of works, including published class notes for Dartmouth courses in the late 1920s and in his general book The Grand Strategy of Evolution (1920), Patten tried to establish evolution as the source of all morality, proper conduct, and good human relations. He therefore becomes a convenient foil for our cardboard relativist who wishes to see little of the external world (if such a concept be intelligible at all) and much of social context in the claims of science.
I could not be more out of sympathy with Patten’s wider effort. I have never read a more tendentious or vainglorious attempt to establish a preferred social morality as the pathway and dictate of nature. I have no particular quarrel with Patten’s beliefs—a compendium of unassailable apple-pie virtues, featuring the value of service to others and the wisdom of self-restraint in a world of temptation. But I’ll be damned if nature can validate, or even address, such cultural hopes and preferences.
Patten argued that nature could instruct us if we learned her patterns and followed them in all our beliefs and dealings. He wrote in 1920:
The universal end, or purpose in life, and in nature, is to construct, to create, or grow. The ways and means of accomplishing that end are mutual service, or cooperative action, and rightness.
This universal growth occurs along three cosmic axes—time, space, and rightness. The three-dimensional result is linear and necessary progress:
There is an abiding compulsion to the action of all these factors which is cumulative, or progressive, producing that increasing architectural organization that we call nature-growth, or evolution.
The direction of evolution also sets a moral imperative, for it “compels man to accept nature’s constructive rightness as his ethical standard, and to adopt her constructive methods as his moral code.”
But if nature’s progress is the source of our morality, then we had better be able to find an unambiguous direction in evolution. Patten bravely surveyed the entire tree of life, with its complex and ramifying branches shooting forth in a thousand separate directions, and managed to realign this intricate meshwork, with bold upper case, as “The Great Highway of Organic Evolution which leads from the lower forms of animal life up to man.”
To produce this remarkable change, Patten had to extract one lineage from life’s tree and depict it as a straight, central highway. He then had to view all other groups as side roads leading to the dirt of nowhere. But how can a central artery be discovered (with humans on top) if, as generally held, vertebrates go back into the mists of time and do not arise directly from any other complex phylum (but share closest ties with the lowly echinoderms)? How can we specify a Great Highway if arthropods, representing some 80 percent of animal species and including the most structurally complex invertebrates, do not lie firmly upon this main route?
Obviously, then, the concept of a Great Highway demands a direct linkage of lower arthropod to higher vertebrate. And if no highway can be found, then we have no natural basis for morality and no primacy of evolution among the disciplines. Patten therefore called upon Cephalaspis and the ostracoderms to perform the greatest of all services—to form the link that would secure both the direction of life and the laws of moral conduct. Patten explicitly cited his arthropod theory of vertebrate origins as the key to his entire system:
It shows that the great vertebrate-ostracoderm-arthropod phylum forms the main trunk of the genealogical tree of the animal kingdom; that, emerging from unsegmented, coelenterate-like animals, as though driven by some mysterious internal power, moves with astonishing precision, through broad, predetermined channels—from which neither habit, nor environment, nor heredity, can cause it to diverge—towards its goal.
Patten even thought that he had finally found a mechanism, in his theory of necessary progress, for that most improbable claim of the arthropod theory—the closure of the old mouth above the brain and the opening of a new version below. He argued that progress must be marked by increasing size of the brain, and that expanding nervous tissue would choke off the old brain-piercing esophagus, forcing construction of a new mouth:
The progressive constriction of the esophagus, by the growth of the surrou
nding brain, ultimately compels all those with relatively larger brains to suck their food in liquid form through the narrowest possible opening, or give up eating altogether…. Without this closing up of what had come to be a very inconvenient gateway to the gut, the growth of the brain, as we see it in the higher vertebrates and in man, would have been a physical impossibility.
An illustration of Patten’s Great Highway of animal evolution from his 1920 book. Note that he places ostracoderms between arthropods and vertebrates. Courtesy of Department of Library Services, American Museum of Natural History.
Our cardboard relativist may now exult. Patten’s personal need to find moral answers in evolution, and the early twentieth-century vogue among paleontologists for reading life’s history as a tale of linear progress, surely fueled his improbable interpretation of ostracoderms as transitional between arthropods and vertebrates. But our realist shouts “wait!” A factual question must still be resolved. People may believe correct things for the damnedest and weirdest of flawed reasons. We still have to know the zoological status of Cephalaspis—for this genus ranks somewhere in life’s genealogy no matter what cultural blinders we may wear at any moment. We still must find out whether Patten was right or wrong.
This question can be answered definitively, for Cephalaspis then had the good fortune to become the subject of this century’s greatest work in observational paleontology—a treatise so stunning in care and detail that I thrill every time I pick it up, even though its unremitting technical detail scarcely forms the usual stuff of inspirational literature. In 1927, Erik Andersson Stensiö, professor of paleontology at Stockholm, published his monograph on “The Downtonian and Devonian Vertebrates of Spitsbergen, Part I, Family Cephalaspidae” (Downtonian is an old name for strata now termed Upper Silurian).