The Great Fossil Enigma

by Simon J. Knell

  The mountain was truly shaken. A fault line now ran through it and intersected with the long-ignored animal. Published just four years after the first Treatise, this new thinking would render that earlier work completely obsolete if it took hold. Rhodes, who was visiting professor at Ohio State University in 1966, liked the paper and “did a lot to calm the ire of the vast majority.”19 Ziegler, however, was annoyed. He could not help but see this as a challenge to a lifetime's work. It complicated a system that worked so perfectly, easily, and usefully, and it came at a moment when Glenister and Klapper had replaced Ziegler's names with numbers and, as we shall see, when other studies were threatening to topple Ziegler's position. Ziegler must have feared that he would need to start his research from scratch, but he also knew that his collections did not hold sufficient evidence to reconstruct Devonian assemblages. The threat to his position was very real.

  Although Bergström and Sweet's stand was greeted with fierce opposition from nearly every quarter, a few looked at what they had done and hardly blinked. Chris Barnes, at the University of Waterloo in Ontario, for example, used Bergström and Sweet's new species to interpret an unexpected cluster of elements from the Middle Ordovician. These convinced him that these four variously sized conodonts of the same type, all oriented in the same way and stacked one upon the other, belonged to the same animal and seemed to reflect their positions in life. The discovery appeared to confirm that Ordovician assemblages were rather different from those that had been found in the Carboniferous. Two years later, in 1969, Austin and Rhodes reported a strange fused cluster from the Carboniferous of the Avon Gorge, near Bristol, England. It was made up of an interlocking arrangement of conodonts of the same kind but of different sizes. They, too, thought these elements had belonged to the same animal and reasoned – like Barnes – that the differing sizes possibly reflected continual replacement of old elements by new or, more likely, the presence of different sized elements in the same animal.20

  The most significant discovery, however, was published by Friedrich-George Lange in 1968. It reported the finding of seventy clusters bound together with bituminous material. These all came from the Upper Devonian Kellwasserkalk of Steinbruch, Germany, and revealed for the first time the complete natural assemblage containing the element known as Palmatolepis.21 Lange had visited Lund in the mid-1960s, showing Bergström his assemblages. Bergström was convinced that they were apparatuses preserved in their original three-dimensional form. He recalls that Lange was of the same opinion but had to back down in order to get his paper published when Ziegler refused to accept them as anything other than coprolites. This was probably the most important and influential discovery of assemblages since the 1930s, for it showed more precisely the positioning of the elements in Schmidt's reconstruction and revealed that these Devonian assemblages were very much like those seen in the Carboniferous. Ziegler's position was now untenable, and workers asked why his collections did not contain the full range of elements now known to accompany Palmatolepis.

  Assemblages in the Silurian were only known from Walliser's associations and Rexroad and Nicoll's clusters. However, in 1969, Charles Pollock of Indiana University discussed some fifty-four clusters, mostly from the same Kokomo Limestone. These confirmed and extended earlier discoveries and suggested the presence of two different natural assemblages in the rock. Pollock felt this supported Rhodes's belief that early natural assemblages were composed of fewer element types and arranged fairly simply. (It was not until 1972, when the Soviet Union's Tamara Mashkova found the first complete Silurian assemblage, that Walliser's associations would be demonstrated to be biological species).22

  Meanwhile, Richard Lane, a young worker at the University of Iowa, was continuing the kind of thinking Lindström had begun. In a paper just five pages long, Lane reviewed the symmetry of conodont pairings. He showed that one could not prove that midline elements existed or, indeed, that any element was unpaired, regardless of symmetry, unless it was found in a well preserved assemblage. Most elements came in mirror-image pairs that were presumed – no doubt based upon the jaw analogy – to have been arranged in life with concave sides facing each other. But by tracing the evolutionary development of these paired elements, Lane could show that some developed from mirror-like pairs into asymmetrical pairings in which each element was morphologically distinct. He pointed out that mirror pairs, asymmetrical pairs, and unpaired asymmetrical elements were known to occur in some modern worms and other marine animals. Lane was not suggesting that the conodonts came from one of these groups of animals but instead pointing out that one could infer relatively little from asymmetrical assemblages. A swimming lifestyle required a bilaterally symmetrical body but not necessarily a symmetrical arrangement of elements. In what became one of the most read papers of the period, Lane fundamentally extended and rationalized the acceptable possibilities.23

  When the Pander Society met in Iowa in May 1968, it was clear that many had tried but struggled to locate the kinds of natural association Bergström and Sweet had found so easily.24 But now Sweet made matters even more challenging for his colleagues by bringing along his doctoral student, Joe Kohut. Kohut had, in his studies, come under the influence of Howard Pincus, chairman of Ohio State University's Geology Department. Pincus was a great advocate of numerical methods in geology, and “Joe had Howard's ear and Howard had Joe's firm attention.”25 Kohut soon discovered that he had a natural aptitude for numbers, and he had, by the time of this meeting, already reworked Bergström and Sweet's vast data. Through Pincus he had gained access to the IBM 7094 mainframe computer in the School of Business Administration next door. This was one of the biggest and fastest mainframe computers available in the mid-1960s. Although extraordinarily underpowered compared to a modern desktop machine, a basic system cost more than three million dollars. Kohut converted Bergström and Sweet's form-species names into code numbers and used punched cards and a Fortran IV program to perform a succession of calculations to reveal the joint occurrences for every possible pairing of elements. By this means groups of elements were pieced together, confirming the natural species Bergström and Sweet had found simply by looking: “Application of the quantitative procedures that are outlined in this study have revealed faunal and genetic associations that have been determined previously by careful observation and tedious work over a period of many years. The obvious advantage of the quantitative approach is its rapidity and objectivity, but, like empirical procedures, it must be based on large collections that have been systematically tabulated.”26

  John Huddle heard Kohut's paper at the Iowa meeting and was completely befuddled by it.27 It pushed what was already a radical argument beyond the everyday concepts and visual evidence that occupied most paleontologists. Few conodont workers were so numerate, and rather than see Kohut's conclusions as affirming the efficacy of numerical methods, they might have felt, instead, that these methods revealed that looking and judging were sufficiently accurate.

  Kohut's novel analysis forced many conodont workers to reflect on their own skill set. Some saw an opening door – a new technology capable of advancing the science – and attempted to follow Kohut's lead. By the time of the North American Paleontological Convention in 1969, Sweet and Bergström could claim that Kohut's program had permitted the generation of Rhodes's assemblages from Ellison's 1941 revision of elements.28 However, this mathematical moment would not last. Kohut found no firm foothold in the tiny field of conodont studies, and with reluctance he entered librarianship. With a sigh of relief, the science slipped back into its more descriptive ways. The discipline had lost its numerate edge.

  In May 1969, when the Pander Society met to discuss conodont stratigraphy at Sweet and Bergström's university in Columbus, it was apparent that conodont studies remained locked in a halfway house, neither one thing nor the other.29 Here converts like Lindström made their commitment to the scheme known; Lindstrom now spoke the new language – as much as he was able. But in what is at the best of ti
mes an arcane science, those outside this small group of specialists would have been at a loss to understand precisely what was going on. With hindsight, the language used at this meeting appears to be in transition when really it existed in two opposing camps. Yet very little of this discrepancy appeared on the surface, at least in the published account. Thus Lindström could talk of “faunas” – a word all who attended would understand – as representing groups of conodont form species occurring at one horizon. Yet Lindström would know, and occasionally say, that these were associated with each other because many of them belonged to the same biological species; they had come from the same animal. Others would use the same word but make no reference to this underlying cause of association, believing that such information was not essential to stratigraphic study. Walliser simply updated his earlier work; his apparatuses remained unnamed. Some spoke of “conodont species” and meant species of animal, while others used to same words to mean “species” of element. These differences, however, only become apparent if close attention is paid to the text; no one made a great fuss about them. Those few who, like Bergström, were attached to the new way would describe their new species with a language that had left the everyday. His Eoplacognathus robustus, for example, was to be understood as composed of “sinistral and dextral ambalodiform and polyplacognathiform elements.” Ziegler, by comparison, could speak of single objects using a single name, such his Polygnathus varcus, which required little more than illustration. For the geologist who simply wanted a practical tool, the simplicity of the old way could not be beaten.

  The Ohio meeting indicated that conodont studies had entered a kind of limbo. There were tensions, but they never appeared in the written accounts. One night, Sweet, Ziegler, and Klapper stood on the street corner in front of the students’ union until 2:00 AM. “We argued and argued and argued that night,”30 Sweet recalled. Positions were becoming entrenched. Ellison, for example, refused to let his students consider Bergström and Sweet's scheme.

  When the conodont workers met again, at East Lansing, Michigan, in August 1970, there was much talk of the Marburg symposium planned for the following year. Here the conflicting approaches were to go head to head. Rhodes was worried: “Neither the dual nomenclature employed by Scott and Rhodes, nor the assemblage Linnean nomenclature of Schmidt, Bergström and Sweet is entirely satisfactory…. The first is illegal, but useful; the second is legal, but ambiguous and sometimes illogical…. There is no reason why the change in general taxonomic practice, which recent data now justify, needs to involve a nomenclatorial bloodbath.”31 Ever ready to assume the role of mediator and diplomat, he nevertheless feared “Conodontological fratricide.”

  In September 1971, between sixty-five and seventy conodont workers arrived in Lindström and Ziegler's home territory of Marburg. They had come from fourteen countries. Huddle arrived expecting the bloodletting Rhodes had predicted, but to almost universal surprise it did not come. Lennart Jeppsson, a relatively new Swedish entrant to the field, thought this owed much to the way the meeting was organized, as presentations were made in stratigraphic order, beginning with those dealing with the oldest rocks. This gave the stage first of all to the leading advocates of change: Lindström, Bergström, and Sweet. There were no keynote speakers, but the seniority of these three men and their positioning in the program assured them of this status. Possessing incomparably huge collections and having pushed hard to progress their interpretations since 1966, they now thought it possible to piece together the evolution of apparatuses in some detail. As a result of this careful planning, by the end of the first day the argument for change had effectively been won. Jeppsson, who spoke toward the end of the conference, felt powerless. His rocks turned up relatively few conodonts and there was little chance he could emulate the Ohio workers. A young Englishman, Dick Aldridge, also found himself at the tail end of proceedings. He stood up, walked to the front without his papers, and said he would no longer give his paper; he had changed his mind.32

  No blood had been spilt. And despite the fear that the utilitarian mountain would collapse, it did not. Ziegler – who never became fully committed to the new scheme – was satisfied because the proposals seemed to endorse the view, first expressed in the 1950s but implicit in the work of Branson and Mehl, that the most rapidly evolving elements might both define and name the assemblage and also perform in stratigraphic studies. This removed the need for Ziegler to truly know apparatuses.33 Other apparatuses seemed to be composed of single element types, and these too posed no problem. Ziegler must have been as relieved as anyone: a juggernaut had driven into his university and he had managed to grab the wheel. But those around him were also keen to see the new scheme incorporate rather than alienate him. If he, as the champion and most successful exponent of the old way, was converted, then they knew the opposition would be fatally wounded.

  The conodont workers had now committed themselves to pursue a path that twenty years before had seemed impossible. Then they imagined this course of action would result chaos. As it was, they resigned themselves to learning new names and understanding the fossils as components rather than species in their own right. Old words would acquire new and different meanings, and other words would simply disappear. David Clark feared that this “new language” would need to be “mastered in a short time.” Lindström and Ziegler, however, preferred to underplay the significance of this change: “We all learned numerous names for fossils when we were students, and a great proportion of these are now obsolete, victims of healthy revision. This is true for all important fossil groups, it must apply in the case of the conodonts, as well.”34 With a little sense, they knew they could rescue a good deal from the ashes of the old regime.

  Rhodes was given the role of diplomatic envoy and charged with drawing up and circulating the “Marburg Proposals,” which would, of course, reach those who were not there to witness this surprising outcome. Inevitably, there was little agreement among the forty-one responses he received, but Rhodes pressed on, editing these together into what became the third version and final agreement. The aim of the document, as he explained, was not to legislate for others but to provide “useful guidelines for those who choose to follow them.” It stressed the need for “great care, constraint and consultation.” It also required Rhodes to cook up a little linguistic spaghetti to cover all eventualities.35

  Having taken this giant leap without the outbreak of civil war, it must have been with a sense of irony that conodont workers read of the intention of the International Commission on Zoological Nomenclature to introduce a scheme of parataxa at the end of that decade. The scheme was proposed by ICZN secretary Richard Melville in 1978 and sought to resolve the same difficulties that had troubled conodont workers in the 1950s.36 To prepare his case, Melville had reviewed those earlier arguments and alighted upon the conodonts as perhaps the strongest cause with which to garner support. By doing so, he unleashed a hornets’ nest of often sharply worded and deeply argued criticism. Jeppsson, for example, felt compelled to send copious letters documenting in detail all that had happened since the 1950s and why the proposed scheme would damage conodont science. Dick Aldridge, who also fundamentally opposed the scheme, took a more dispassionate view. He told Jeppsson how “greatly impressed” he was that Melville, “who has to deal with ‘all’ systematic zoologists in the world, has involved himself so deeply and thoroughly to bring about a solution of these problems.” Aldridge corresponded with Melville, hoping to school him in the niceties of conodont taxonomy, but Melville tenaciously held his ground. While admitting that conodonts were no longer the group in his sights, Melville felt duty bound to serve others who would benefit from the scheme. Aldridge pressed on – and not without effect. The arguments became increasingly refined and arcane. “I am gradually educating myself,” Melville told him encouragingly, but then warned, “and must ask you to accept that an old dog has difficulty with new tricks.”37

  When Aldridge attended the Pander Society meeting in Vien
na and Prague in August 1980, he feared he was still losing. Among the seventy-one conodont workers from twenty-five countries, parataxa once again became a hot topic, but now it produced absolute opposition. Those present drafted a petition – known as the Wolayer Resolution – which read, “We unequivocally reject these amendments to the Code and urge you in the strongest possible terms to vote against them…. As a nomenclature for parataxa would legalise a dual system of names, paranomenclature is antithetical to the purpose of the Code.”38 Aldridge wrote to Melville to warn him, explaining that he did not know if this was the right way to go about things but that it was now so late in the day that they felt compelled to try anything.

  Melville was still willing to learn and asked Aldridge to give him a practical demonstration of the problem that October in London. When the two met, Aldridge found the ICZN secretary amiable and helpful. But as they were discussing the finer points of the proposed scheme, Melville's secretary entered the room with a letter from Walt Sweet. In seven neatly typed pages, Sweet set out a closely argued case against Melville's proposal. In fact, Sweet laid it on rather thick. This was not a letter seeking to forward science so much as win the argument and rebuff an unwelcome intrusion into a field that had so recently emerged from a difficult past. Sweet was manning the barricades again and ready for another civil war. He ended by warning, “Should the Commission approve the amendments involving ‘parataxonomy’ and ‘paranomenclature,’ you can expect a somewhat more extensive and more carefully composed response from me in the literature. I hope it will not come to that.”39 Sweet saw in the scheme the potential to undo all that had been achieved.

  Melville thought Sweet's letter “magnificent,” and he responded at length using a well-honed style that must have seen him through many such arguments.40 It was smooth and diplomatic, cleverly and no doubt wittily using abstraction and generality to throw a thin veil over his criticisms of Sweet and his other detractors. He told Sweet that conodont workers had made themselves remote from the ICZN, and as a result “the structure of current conodont nomenclature suffers from severe logical weaknesses and is highly vulnerable.” Rather pointedly he observed, “If conodont workers in the mid-60’s had taken the trouble to keep us informed of the exciting developments in taxonomy that were then taking place, we could have proceeded on a better basis of mutual understanding.”