Wonderful Life: The Burgess Shale and the Nature of History

by Stephen Jay Gould

  But Walcott had been downright circumspect compared with later reconstructions that added more and more arthropod features with less and less compunction. In 1931, the great ecologist G. Evelyn Hutchinson, driven to paleontology by the fascinating problem of how anostracans could change their environmental preferences from Cambrian oceans to modern freshwater ponds, reconstructed Opabinia in the standard upside-down position of a swimming anostracan (figure 3.22). He turned the lateral flaps into long bladelike appendages neatly fitted to the side of an arthropod carapace.

  The climax of this imaginative tradition arrived with the aesthetically lovely but fanciful reconstruction of Simonetta (1970).* Opabinia has become an ideal arthropod (figure 3.23). The frontal nozzle is shown with a longitudinal suture (entirely imaginary), indicating its origin as a pair of antennae, now fused. Simonetta “found” two additional pairs of short arthropod appendages on the head—one constructed from a pair of eyes, the other from a bump on the carapace. On each segment of the body itself, Simonetta drew a strong and fully biramous appendage—a bladelike gill branch above a small but firm leg branch. Whittington faced this unchallenged tradition when he began his work on the ten precious specimens of Opabinia.

  I now come to the fulcrum of this book. I have half a mind to switch to upper case, or to some snazzy font, or to red type, for the next page or two—but I desist out of respect for the aesthetic traditions of bookmaking. I also refrain because I do not wish to fall into the lap of legend (having already dispersed one for the discovery of the Burgess Shale). My emotions and desires are mixed. I am about to describe the key moment in this drama, but I am also committed to the historical principle that such moments do not exist, at least not as our legends proclaim.

  Key moments are kid stuff. How can such a story as this, involving so many people engaged in complex intellectual struggles, proclaim any moment as a single focus, or even as most important? I have labored to master all the details and to arrange them in proper order. How can I now blow all this effort on the myth of eureka? I suppose that one can discover a single object—say, the Hope diamond—at a particular moment, but even such a pristine event has a tangle of inevitable antecedents in geological training, political intrigue, personal relations, and good luck. But I am talking about an abstract and far-reaching transformation in our view of life’s pattern and the meaning of history. How can such a complex change possess a moment before, when it wasn’t, and a moment after, when it was? Does natural selection, or laissez-faire economics, or structuralism, or the rationale for the Immaculate Conception of Mary, or any other complex moral or intellectual position, owe its formulation to a single person, place, or day?*

  3.22. Hutchinson’s reconstruction of Opabinia as an anostracan swimming upside down in the modern position (1931).

  Still, as Orwell said about his metaphorical Russia in a farmyard, some animals are more equal than others. We need heroic items and moments to focus our attention—the apple that hit Newton and the objects that Galileo did not drop from the Leaning Tower. The beat goes on, but we may discern a high spot in the continuity.

  I believe that the transformation of the Burgess Shale did have a Rubicon of sorts, at least symbolically—a key discovery that can separate a before and an after.

  So we return to Harry Whittington, facing the entire world’s supply of Opabinia. Everyone had always identified this animal as an arthropod, but no one had found the smoking gun, the segmented appendages that define the group. But then, no one before Whittington had possessed the techniques needed to seek out small appendages hidden under an external carapace. A few years before, Harry had made the central methodological discovery that the Burgess Shale fossils are three-dimensional objects (however crushed), with top layers that one can dissect away, to reveal the structures underneath. Harry had already resolved Marrella, Yohoia, and the Burgess trilobites with this method.

  Opabinia virtually clamored for its crucial experiment under the new techniques: dissect through the carapace to find the body appendages and their attachments, dissect through the head shield to find the frontal appendages. So Harry dissected, in full confidence that he would find the jointed appendages of an arthropod. Harry dissected—and he found nothing under the carapace.

  Opabinia was not an arthropod. And it sure as hell wasn’t anything else that anyone could specify either. On close inspection, nothing from the Burgess Shale seemed to fit into any modern group. Marrella and Yohoia at least were arthropods, even if orphaned within this giant phylum. But what was Opabinia ?

  3.23. Attractive but fallacious restoration of Opabinia as an arthropod by Simonetta (1970). (A) Top view. (B) Side view. Simonetta showed the frontal nozzle as formed by fused antennae, and drew biramous appendages on each supposed body segment.

  Whittington’s conclusion may have been confusing, but it was also liberating. Opabinia did not have to conform to the demands of arthropod, or any other, design. Whittington could come as close as any paleontologist ever had to the unattainable ideal of Parsifal—the perfect fool, with no preconceptions. He could simply describe what he saw, however strange.

  Opabinia is peculiar indeed, but not inscrutable. It works like most animals. Opabinia is bilaterally symmetrical. It has a head and a tail, eyes, and a gut running from front to back. It is an ideal creature for any eager scientist—not so crazy as to be intractable, but weird enough to thrill any curious person.

  Whittington began his monograph by chiding his predecessors for their unquestioning allegiance to the arthropod model, and for their consequent tendency to rely more on expectations of the model than on observation of the specimens: “Continuous interest in Opabinia has not been accompanied by critical study of the specimens, so that fancy has not been inhibited by facts. The present work aims to provide a sounder basis upon which to speculate” (1975a, p. 3). With characteristic understatement (his personal tendency added to the British norm), Whittington then wrote: “My conclusions on morphology have led to a reconstruction which differs in many important respects from all earlier ones” (1975a, p. 3).

  These “many important respects” led to an animal that might grace the set of a science-fiction film, if considerably enlarged beyond its actual length of 43–70 mm (less than three inches at most). Consider the major features of Whittington’s reconstruction:

  1. Opabinia does not have two eyes, but, count ’em, five! These are arranged as two pairs on short stalks, with a fifth eye, probably unstalked, mounted on the midline (see figure 3.20).

  2. The frontal nozzle is not a retractable proboscis or a product of fused antennae (the two favorite interpretations consistent with arthropod design). It is attached to the bottom front border of the head and extends forward. It is a flexible organ, built as a cylindrical striated tube—literally like the hose of a vacuum cleaner, and perhaps bendable by the same principles. Its end is divided longitudinally into two halves, each with a group of long spines directed inward and forward. The tube may have contained a central, fluid-filled canal—a good device for requisite stiffness with enough flexibility.

  3. The gut is a single tube running straight along the center of the animal for most of the body’s length (see figure 3.24). However, at the head, the gut makes a U-shaped bend, and turns sharply around to produce a backward-facing mouth. Interestingly, the frontal nozzle has just the right length to reach, and appropriate flexibility to bend around and pass food to, the mouth. Whittington suggests that Opabinia fed primarily by capturing food in the “pincers” formed by the spiny parts at the front of the nozzle, and then bending the nozzle around to the mouth.

  4. The main portion of the trunk has fifteen segments, each segment bearing a pair of thin lateral lobes, one on each side of the central axis. These lobes overlap, and are directed downward and outward (see figure 3.20).

  5. Each lobe except the first bears on its dorsal surface a paddle-shaped gill attached near the base of the lobe. Although the bottom surface of the gill is flat, the upper surface
consists of a set of thin lamellae, overlapping like a deck of cards spread out.

  6. The last three segments of the trunk form a “tail” built by three pairs of thin, lobate blades directed upward and outward (see figure 3.20).

  Whittington needed all his special methods of dissection, varied orientations, and part–counterpart to resolve the morphology of so peculiar a beast. He also discovered that a failure to appreciate these methods had provided a major argument to support the arthropod model. Walcott had confused part and counterpart in one important specimen. He thought that he was viewing the bottom surface of the animal; in fact, he was looking down upon the upper surface. Raymond, accepting this upside-down interpretation, had made the perfectly reasonable claim that the gills of Opabinia lay below the outer carapace—as in the standard arthropod arrangement, with gill branches as the upper limbs of biramous appendages located just under the carapace. But in the correct orientation, the gills lie above the body lobes in a most unarthropod-like orientation.

  Figures 3.24–3.26 provide a striking illustration of the power of Whittington’s methods. These are his camera lucida drawings of three specimens, in varying orientations, each combining features from the part and counterpart of the same specimen. Figure 3.24 provides a view from above (dorsal). We see the position of the eyes and nozzle, the full sequence of lateral lobes, and the gills lying above the lobes. The gut runs as a straight tube down the middle of the body. Figure 3.25 is a side view and reveals several features that could not be seen from the top. We now discern the point of attachment for the nozzle, and we note that the gut bends in a U to form the rearward-facing mouth. (In top view, the bend and rearward section collapse upon the straight portion and cannot be distinguished at all.) The top view also tells us nothing about the relative positions of lateral lobes and tail fins, for these are collapsed into the same plane. But the side view of figure 3.25 shows the lateral lobes pointing downward and away from the body, while the tail fins stand high and point upward—in good positions, respectively, for oars and rudders.

  3.24. Camera lucida drawing for a specimen of Opabinia in the conventional position, viewed from the top. On each side, gills (labeled g) and lobes (l) are clearly distinguishable; the trace of the gut runs along the midline. Two pairs of eyes are visible, and the nozzle extends forward from the front end.

  Figures 3.24 and 3.25 provide the two basic orientations, but they still leave several questions unanswered—and further specimens are needed. For example, neither shows the full complement of five eyes (they are delicate, and often collapse together into a jumble). Figure 3.26 fills some crucial gaps: five separate eyes are visible, and the frontal nozzle bends around to the area of the mouth.

  Marrella and Yohoia had challenged Walcott’s shoehorn, but these genera were only orphaned within the Arthropoda. With Opabinia, the game cranked up to another level, and changed unalterably and forever. Opabinia belonged nowhere among the known animals of this or any former earth. If Whittington had chosen to place it within a formal classification at all (he wisely declined), he would have been forced to erect a new phylum for this single genus. Five eyes, a frontal nozzle, and gills above lateral flaps! Walcott’s shoehorn had fractured. Whittington wrote with characteristic brevity in the passive voice: “Opabinia regalis is not considered to have been a trilobitomorph arthropod, nor is it regarded as an annelid” (1975, p. 2). Harry may be a measured man, but he knew what Opabinia implied for the rest of the Burgess fauna. “The Burgess Shale,” he remarked laconically, “contains other undescribed segmented animals of uncertain affinities” (1975, p. 41).

  3.25. A specimen of Opabinia preserved in a more unusual orientation, on its side. Here lobes and gills of the right and left sides are jumbled together and difficult to distinguish. But many features not visible in the conventionally positioned specimen of figure 3.24 can now be understood: the orientation of the tail fins (labeled fins relative to the side lobes, the point of insertion for the nozzle, and the rearward bending of the front end of the gut.

  3.26. A third specimen of Opabinia, again in the conventional position. Several features not apparent in the other specimens can be distinguished: the fifth eye (labeled m, for “middle eye”) is visible at the upper right, and we note that the nozzle can bend around to the level of the mouth.

  I believe that Whittington’s reconstruction of Opabinia in 1975 will stand as one of the great documents in the history of human knowledge. How many other empirical studies have led directly on to a fundamentally revised view about the history of life? We are awestruck by Tyrannosaurus; we marvel at the feathers of Archaeopteryx; we revel in every scrap of fossil human bone from Africa. But none of these has taught us anywhere near so much about the nature of evolution as a little two-inch Cambrian oddball invertebrate named Opabinia.

  ACT 3. The Revision Expands: The Success of a Research Team, 1975–1978

  SETTING A STRATEGY FOR A GENERALIZATION

  Think of all the accumulation songs in the English folk tradition. The first item never amounts to much—a partridge in a pear tree, or a paper of pins. “Green Grow the Rushes, Ho” puts it best: “One is one and all alone and ever more shall be so.”

  Opabinia carries the full weight of the Burgess message for a new view of life. It is as bizarre, as different from all living creatures, as anything else in the Burgess Shale. But one is all alone and ever more shall be so. The fossil record contains other oddities here and there—like the Tully Monster of Mazon Creek (see page 63). Opabinia, just one case, is a shrug of the shoulders, not a discovery about life in general. This example did not establish an incontrovertible new interpretation. Quite the opposite; it only hinted at a possibility worth exploring—especially with Marrella and Yohoia indicating that something similar, at a lower level, was running rampant among the Burgess arthropods.

  All interesting issues in natural history are questions of relative frequency, not single examples. Everything happens once amidst the richness of nature. But when an unanticipated phenomenon occurs again and again—finally turning into an expectation—then theories are overturned. Opabinia would not earn its status as primer and flagship for a new view of life until its message of taxonomic uniqueness became ordinary within the Burgess Shale, however exquisitely rare for later times.

  This need for numbers of examples—for an assessment of the relative frequency of oddballs within the entire Burgess fauna—makes the myth of the hero, grade B Western movie style, inapplicable to this story in principle. Harry Whittington could not be a lone lawman subduing saloonful after saloonful of reprobates. Marrella had taken more than four years. The Burgess arthropods alone would require several lifetimes. Whittington could either intone the lament of the frustrated Mercedes—“So many pedestrians, so little time”—or he could enlist a fleet to help. He chose the second alternative. Science is a collective enterprise in any case.

  After selecting the genera that would provide a focus for his personal studies, Whittington divided the remaining arthropods into three groups, each suitable for an extensive research project by a collaborator. In addition, and growing both more troubling and crucial since the identification of Opabinia as an oddball outside any established phylum, stood the many genera that Walcott had classified as annelid worms (191lc). If Walcott’s shoehorn had hidden a general theme, of taxonomic uniqueness, the story would probably emerge (if not explode) even more clearly from the annelids than from the arthropods. Arthropods have clear and complex defining characters. Walcott might have wrongly shoehorned his arthropods into conventional groups within the phylum, but most were genuine arthropods at least (with Opabinia and, later, Anomalocaris as exceptions). But anything soft, segmented, and bilaterally symmetrical might be called a worm. The potential for oddballs loomed largest among Walcott’s “annelids.”

  Whittington doubted that the three arthropod groups were coherent taxonomic assemblages. Each shared some features of superficially similar appearance, but Marrella and Yohoia had already
taught caution about such externalities. Still, the three groups formed convenient divisions for research efforts, and the postulate of coherence could become a focal question for testing. (All three groups turned out to be heterogeneous—an important conclusion that confirmed the status of Burgess arthropods as spectacularly disparate compared with all later faunas.)

  The three groups, all generally recognized in Burgess classifications from Walcott to Størmer, were (1) the large assemblage of arthropods with bivalved carapaces, always assumed to be true malacostracan crustaceans; (2) the “merostomoid” species, generally oval in shape and with a large discrete head shield that seemed to recall the great group of fossil eurypterids and their cousins the horseshoe crabs; and (3) apparent crustaceans with simple carapaces not divided into two parts, or valves.

  When Whittington began his work in the late 1960s, two junior colleagues agreed to take on the smaller projects in this list. David Bruton of the University of Oslo received the “merostomoids” (I have discussed his work on Sidneyia in my section on techniques, early in chapter III, and shall report his conclusions in proper chronological sequence, in Act 5). Chris Hughes of Cambridge tackled Burgessia and Waptia, third and fourth most common Burgess arthropods, and forming the group of apparent crustaceans with simple carapaces. The monograph on Waptia has yet to appear, but Hughes’s 1975 treatment of Burgessia provided an important affirmation of the growing pattern already indicated by Marrella and Yohoia. Burgessia, with its oval carapace, and long tail spike (almost twice the length of the body), was not a notostracan branchiopod, as Walcott had believed, but yet another arthropod orphan of unique design (figure 3.27). Hughes declined to make a formal taxonomic place for Burgessia, because he regarded this genus as a peculiar grabbag, combining features generally regarded as belonging to a number of separate arthropod groups. He concluded: