Leonardo’s Mountain of Clams and the Diet of Worms

by Stephen Jay Gould

  In both extreme cases, human ancestors begin at 300 cc, and peak today at 1,300 cc. Both schemes invoke a metaphor of struggle and persistence—the slow climb up a ladder in the traditional linear view, and success in hanging on through every adversity in the “bush pruning” alternative. Both views are also clearly wrong in their exclusive versions. Why, then, do we tend to feel comfort and affinity for the linear scheme, while regarding the “bush pruning” alternative as laughable and inexplicable nonsense, no doubt introduced by yours truly to satisfy some perverse and personal whim.

  Yet I wish to argue (1) that both views express important partial truths; (2) that we have favored the linear view primarily in obedience to the disabling cultural bias illustrated by the earlier examples in this essay; (3) that the history of twentieth-century ideas about human evolution can be epitomized by growing strength of the “bush making and pruning” view—and the retreat of the linear view—all leading to a proper balance; and (4) that a new discovery, announced in December 1996 (and inspiring this essay), provides strong and unexpected support for bushiness as the usual condition of the human lineage. (I shall, for the rest of this essay, refer to the two modes of thought as “linear” and “bushy” accounts of evolutionary trends.)

  My friend and closest colleague, Niles Eldredge, has labeled these two approaches to trends as “taxic” and “transformational”—or “based on the production of many separate species” (formally named groups of organisms, such as species and genera, are called “taxa”), versus “propelled by the advantages of certain traits” (big brains, for example) in competition among varying individuals within a single group. The two views differ most significantly in their primary “motors” for generating trends. In the bushy, or taxic, theory, trends require a substantial production of independent species, for net change in a lineage depends upon a differential survival and further proliferation of some species versus the extinction of others. In the linear, or transformational, theory, trends require no bush of species, but arise by the competitive success of favorable traits in a gradually progressing unit. (Of course, supporters of the linear view do not deny that lineages may also produce new species by branching, but these scientists tend to separate the progressive carrier of the trend from doomed side branches. In other words, for the linear transformationist, the production of numerous species does not contribute to the major progressive trends of life’s history.) Ernst Mayr, the dean of American evolutionists, and a strong supporter of copious speciation as a central ingredient in evolutionary trends, expressed the contrast well by writing:

  I feel that it is the very process of creating so many species which leads to evolutionary progress. Species, in the sense of evolution, are quite comparable to mutations. They also are a necessity for evolutionary progress, even though only one of many mutations leads to a significant improvement of the genotype . . . Seen in this light, it appears then that a prodigious multiplication of species is a prerequisite for evolutionary progress . . . Without speciation, there would be no diversification of the organic world, no adaptive radiation, and very little evolutionary progress. The species, then, is the keystone of evolution.

  Nonetheless, the linear view has, until recently, strongly dominated traditional thinking about human evolution. For example, the hoary and clichéd concept of a “missing link” presupposes linearity—for links are joining points in a sequence. Evolutionary bushes may be riddled with all the absences and uncertainties imposed by our poor fossil record, but a bush cannot feature a single and crucial “missing link.”

  Moreover, the linear view has not just been accepted passively or unthinkingly—as a simple expression of an unquestioned bias. The idea of coexistence among several hominid species has been actively denied, attacked, and even stigmatized as bad biological reasoning. For example, when I was a graduate student in the 1960s, an idea called “the single species hypothesis” still enjoyed strong, probably majority, support among students of human evolution. According to this theory, only one hominid species could, in principle, occupy a single region at any one time. Thus, since most of our evolutionary history had unfolded on the single continent of Africa, our trends must arise by linear transformation, with only one species living at any moment, slowly perfecting itself toward the next stage. Advocates cited (I would say mis-cited) the ecological principle that only one species can occupy any “niche”—or suitable environment for “making a living.” Beetles have “narrow” niches, so several species can live in one area—some on bark, some on the ground, some high in trees. But hominids, with our unique invention of “culture” (however primitive at first), occupy such a “broad” niche that no single place can house more than one species.

  The leading basic textbook in physical anthropology at the time (Human Evolution, by C. L. Brace and M. F. Ashley Montagu, 1977, first edition 1965), held that “the known fossils are most realistically placed in a linear evolutionary relationship.” The authors specified four sequential stages—australopithecine, pithecanthropine, Neanderthal, and modern—and justified their sequence by the “single species hypothesis.” They wrote:

  Culture as a major means of adaptation is unique in the world of living organisms, and for all important purposes can be considered an ecological niche in itself—the cultural ecological niche. There is an evolutionary principle based on the logic of efficiency which states that, in the long run, no two organisms can occupy the same ecological niche. In the end, one will out-compete the other and retain sole possession of the niche in question. Applied to the primates, this should mean that no two forms could occupy the cultural ecological niche for any length of time.

  C. Loring Brace, one of the authors of this text, has continued to resist the notion of bushiness in hominid evolution. In the 1991 edition of his popular text The Stages of Human Evolution, Brace acknowledges only one side branch in the entire history of human evolution—and he calls this substantial lineage of robust australopithecines a “twig”!

  My own view, however, is represented by the final unilinear arrangement, where the Australopithecines evolved into the Pithecanthropines which in turn evolved into the Neanderthals throughout the whole of the inhabited Old World, and these finally became transformed into the various modern populations alive today. I have left off the Australopithecine twig that became hyper-robust and died out . . . just to give a streamlined version of my general view.

  Brace dismisses the idea that two (or more) human species might have interacted in one place. He even invents the label of “hominid catastrophism” to stigmatize the view (now favored by most paleontologists, particularly for the replacement of Neanderthals by moderns in Europe) that a temporal transition from one species to another might arise by immigration of the later species from another region (followed by local extinction of the original inhabitants), rather than by linear evolutionary transformation. Brace writes:

  The result is remarkably like the picture presented by Cuvier’s catastrophism early in the nineteenth century which regards change as occurring suddenly, for undiscoverable reasons, and away from the region under examination. The new form, which spreads by migration, then prevails until the next sudden change.

  Yet, of all alterations in thinking about human evolution that have occurred during my professional lifetime, none has been more transforming, or further ranging in implications, than the increasing documentation of substantial bushiness throughout most of hominid history. Our present reality of one worldwide species represents an oddity, not the norm—and we have been fooled by our bad habit of generalizing a transient and contingent present.

  I would summarize this fundamental change from the linear to the bushy view of our evolutionary history in five chronological discoveries and arguments, with the latest news as the fifth finding.

  1. TWO BRANCHES OF AUSTRALOPITHECINES. When South African scientists described Australopithecus, the genus ancestral to our own Homo, in the 1920s, they designated two major branches or species, Australop
ithecus africanus and A. robustus (known in later literature as the gracile and robust forms). Thus, a bushy theory for our early days enjoyed some support from the start. But proponents of the single-species hypothesis either viewed the two names as improperly given to males and females of a single species, or (as in the quote from Brace previously cited) regarded the robust lineage as a doomed and insignificant side branch, probably driven to extinction by our superior forebears, the graciles.

  However, in 1959, Mary Leakey found a key specimen with robust features so exaggerated that sexual variation within a single species became implausible as an explanation for the extent of difference. The probable coexistence of two australopithecine lines could no longer be denied—and the purest version of the single-species hypothesis died. (Mary Leakey originally called this skull Zinjanthropus; we now generally designate this form as a separate, so-called hyper-robust species, Australopithecus boisei.)

  2. COEXISTENCE OF AUSTRALOPlTHFCUS AND HOMO. Linearists could still adopt a fallback position. They could brand the robust (and hyper-robust) australopithecines as an insignificant blind alley, regard the graciles as linearly ancestral to our own genus Homo, and then apply the single-species hypothesis to Homo alone, drawing a line from Homo erectus (“Java” and “Peking” man in the older texts) through Neanderthal to our current exaltation. But then, in the mid-1970s, Richard Leakey (Mary’s son) found hyper-robust specimens in the same strata that yielded bones of African Homo erectus (sometimes called Homo ergaster, but little different from the Asian Homo erectus of Indonesia and China). No one could possibly encompass this range of variation within the boundary of a single species. If the most extreme of the robust australopithecines coexisted with the most advanced members of our own ancestry, then the old line of progress had become an undeniably diverging bush.

  3. THE PLETHORA OF AFRICAN SPECIES BETWEEN 3 AND 2 MILLION YEARS AGO. Two branches destroy the linear theory, but don’t build a very impressive bush. In the twenty years since Richard Leakey’s discovery of these two irrefutably coexisting species, further research on hominid history has stressed one primary theme above all others: The bush gets bushier and bushier. To summarize a great deal of elegant research in too short a statement: We have no evidence for more than one species during the earliest period from 3.0 to more than 4 million years ago. (For most of this interval, we know only Australopithecus afarensis, the famous “Lucy” of our popular literature.) But between 3.0 and 2.0 million years ago (and mostly during the last half-million years of this interval), a virtual explosion of hominid species occurred, on both major branches of the hominid bush—that is, both within the ancestral genus Australopithecus, and within the derived genus Homo. The accompanying chart, presented in Donald Johanson and Blake Edgar’s recent book From Lucy to Language, shows as many as six coexisting hominid species during this period, three within our own genus Homo.

  4. BUSHINESS IN LATER HUMAN HISTORY: THE NEANDERTHAL ISSUE. Linear preferences die hard. I think that all major students of the subject now accept substantial bushiness, and coexistence of several species in Africa, during early hominid history, but a version of the old linear view still persists as a popular (though, I judge, dwindling) theory for later human history during the past million years or so, and especially for the origin of Homo sapiens. This debate has been prominently featured in the press (and treated in several of these essays) as a conflict between the “multiregional” and “out-of-Africa” theories for modern human origins. Multiregionalism will probably be remembered as the last post of the linear view. Under this model, all hominid evolution occurs in Africa (admittedly in a fairly bushy manner) until the origin of Homo erectus. This species then spreads out to all the Old World continents between 1.5 and 2 million years ago. The three major populations of Homo erectus, in Africa, Europe, and Asia, then evolve in parallel (abetted by a low level of migration and consequent mixing among the three groups) toward Homo sapiens. Such an idea represents linearity with a vengeance—as all subgroups within a single species move onward (and brainward) in the same optimal direction. In Europe, for example, Homo erectus evolves to Neanderthal, and Neanderthal transforms to Homo sapiens—one species at any time, but constantly on the upward move.

  The out-of-Africa alternative may best be understood as a particular version of the bushy perspective. Homo erectus moves out to all three Old World continents. Homo sapiens arises as a branch (the bushy view) from one of these populations, not as a terminus to a universal trend. Homo sapiens then spreads as a second diaspora from its place of origin, presumably Africa on both genetic and paleontological grounds. But Homo erectus (or its descendants) already inhabit Europe and Asia—so African Homo sapiens arrives as a second human species (bushy coexistence again), and eventually supplants the original form. Under this bushy view, Neanderthal and modern Homo sapiens are separate (and potentially coexisting) human species, not the before and after of a single linear transformation—for Neanderthal branched from European Homo erectus (or its descendants), while forebears of modern Europeans arrived from Africa after a separate origin from African Homo erectus populations.

  In my reading, and as summarized elsewhere (perhaps best in the recent book by C. Stringer and R. McKie, African Exodus: The Origin of Modern Humanity), the balance of recent evidence tilts strongly (perhaps conclusively) to the out-of-Africa view, and therefore to the predominance of bushiness over linearity as a central theme in human evolution. (Incidentally, this new and emerging consensus is the very view that Brace so scornfully rejected and labeled as “hominid catastrophism”—the idea that Homo sapiens arrived from Africa as a second wave and supplanted Neanderthal, in contrast with the only reconstruction that Brace regarded as “evolutionary,” that is, the linear passage of Neanderthal to modern humans. In fact, both views are equally consistent with an evolutionary perspective. The contingent and empirical data of actual history, not preferences of theory [laden with a complex range of unconscious biases], must decide the issue.)

  5. MORE BUSHINESS IN LATER HUMAN HISTORY: NEW DATA FROM ASIA. If Neanderthal and Homo sapiens coexisted as independent species in Europe, thus refuting the linear view, what happened in eastern Asia, where Dubois first discovered Homo erectus in the 1890s, and where this ancestral species enjoyed long and widespread success? In the multiregional view, these Asian Homo erectus populations evolved directly into modern Asian groups of Homo sapiens. In the bushy alternative, Homo sapiens arrived (ultimately from Africa) as a second wave of migration, and may have coexisted for a time with Asian Homo erectus or its descendants. The obvious test between these starkly different views requires a fossil record, either of intermediacy or of coexistence, during the crucial time of transition between the two species. But such decisive data have not been available, because the youngest known Asian Homo erectus (from China) range from about 290,000 to 420,000 years old, while the oldest Asian Homo sapiens specimens are only about 40,000 years old. Thus, we had no evidence at all for the crucial intervening years.

  Eugen Dubois first discovered Homo erectus in Java during the early 1890s—and these specimens, from Trinil, remain the most famous Indonesian representatives of the species. But, in the early 1930s, Dutch geologists discovered a suite of twelve hominid calvaria (skull tops lacking the facial skeleton and upper jaws) from the nearby site of Ngandong on the banks of the Solo River. These specimens—variously known in the older literature as “Solo man” or “Homo soloensis”—have engendered a long and substantial debate about their identity, but a present consensus considers them as members of Dubois’s species, Homo erectus.

  Yet, while anthropologists had finally reached some agreement about their identity, the age of the Solo specimens remained unknown. This crucial issue may now have been resolved—and in a surprising manner—by an article that appeared in the December 13, 1996, issue of Science magazine: “Latest Homo erectus of Java: Potential contemporaneity with Homo sapiens in Southeast Asia,” by C. C. Swisher III, W. J. Rink, S. C. Anton, H. P. Sch
warcz, G. H. Curtis, A. Suprijo, and Widiasmoro (yes, the last author’s name is complete; most Indonesians, like former leader Suharto, or past boss Sukarno, use only one name). The curators of the Solo calvaria would not let these authors use the original material for dating (since the methods destroy parts of the specimens). So Swisher and colleagues collected bovid teeth (cattle and their relatives) from two sites in the same stratum that yielded the hominid calvaria. They applied two independent techniques of radiometric dating, and reached the same surprising conclusion—ever so gratifying for fans of the bush—that the Solo hominid specimens lived between 27,000 and 53,000 years ago. If these conclusions stand up to later scrutiny, then Homo erectus did not transform to modern humans in Asia—for the two species coexisted as independent entities about forty thousand years ago.

  Moreover, and moving to the general statement that inspired this essay, if we now consider the whole earth at forty thousand years ago, we note a bush of three coexisting human species—Homo neanderthalensis in Europe, surviving Homo erectus in Asia, and Homo sapiens continuing a relentless spread throughout the habitable world. This collection of three might not match the richness of an African bush of some half a dozen species about 2 million years ago, but the conclusion that three human species still coexisted as recently as thirty to forty thousand years ago does require a major reassessment of conventional thinking. Our modern world represents the oddity, not the generality. Only one human species now inhabits this planet, but most of hominid history featured a multiplicity, not a unity.