Flores skeletons: Irrational scientists or imaginary species?

Robert B. Eckhardt, Ph.D.
Professor of Developmental Genetics and Evolutionary Morphology
Department of Kinesiology
The Pennsylvania State University
University Park, PA 16802 USA

Maciej Henneberg, Ph.D., D.Sc., FAIBiol
Wood Jones Professor of Anthropological and Comparative Anatomy
Medical School
University of Adelaide
Adelaide, SA 5005

Ah, the synergies of the Scientists’ Bookshelf!

Reviewing Thinking, Fast and Slow, Steven Sloman (2012) wrote “In the face of evidence, we are more concerned with how good a story we can tell about it than we are with coming up with alternative accounts to explain it, remembering other events that contradict it, or thinking about how things might have been otherwise…. Wrapping events into a good story that supports one’s case is far more persuasive than poking holes in somebody else’s story or making a fair and balanced argument.” Along the way Sloman notes, almost presciently for the style of the next review, “…that we humans have a tendency to focus only on what is staring us in the face….” Half of the next review essay by Greg Laden (we omit comment on Martin Meredith’s Born in Africa, which is treated in the same piece) focuses rather heavily on two skulls that continue to stare paleoanthropology in the face: Taung, for the better part of a century, and LB1, for the worse portion of nearly a decade.

Since 2004, we and our colleagues (originally several Indonesians, Australians, Polish, and U.S. citizens and British subjects, often referred to as “the pathology group” (probably suggested by the title of the paper by Henneberg and Thorne, 2004), now reduced by the deaths of Teuku Jacob and Alan Thorne but augmented by a growing number of young investigators) have faced much opposition during our efforts at “poking holes in” (we prefer “testing hypotheses about”) the story that Laden and others commonly call “one of the most important finds of the past hundred years.” Several of us (MH, RBE) are among the relatively few scientists to have studied firsthand the skeletons from Liang Bua Cave, Flores. Our study period was limited, and subsequently we have been denied access even to casts since early 2005, despite promises to the contrary. We concur that these bones are important, but not because they represent a new species, which we feel is a misreading of the evidence, a view shared by scientists who are independent of our group (e.g. Martin, et al., 2006). Rather, we stress that the Flores unfossilized remains are of major consequence for a different reason: They provide a chance to scrutinize some key operating assumptions and procedures of an entire sub-discipline, paleoanthropology, that split from human biology in the last century, developing its own methods and ways of arriving at conclusions. We note here that there are some individual scientists working in this domain who observe the rules of evidence and inference that are followed more commonly elsewhere in science. As sometimes is said in such situations, “some of our best friends are” paleoanthropologists, but the field as a whole has methodological idiosyncrasies. If we are successful in catalyzing what we see as overdue disciplinary introspection, the result should be the renaissance of a unified science of human biology, with shared methods used in the study of human populations past as well as present. If we fail to convince our colleagues why this objective is vital, then human biology will remain in its currently fragmented state, with a set of well understood evolutionary principles governing the study of genetic relationships among living populations, while discoveries from the past will continue to flicker briefly as episodic media images illustrating a disjointed narrative that is, for all practical purposes, atheoretical and deeply rooted in ancient Platonic essentialism (Eckhardt, 2007a, Henneberg 2009).

Establishing the necessary common principles that unite the study of human microevolution and macroevolution will require a firm grasp of the difference between nonce explanations reminiscent of Kipling’s “Just So Stories,” and standard tests of hypotheses that take into account the wealth of evidence and principles for its evaluation that have accrued for more than a century. These procedural contrasts parallel distinctions between the blogger and the working scientist. The first might say, as Laden does here, “One wonders, given the reality of the Flores find, whether the world may once have been (partly) populated by additional species as different from the main line of human evolution as Hobbit….” There are many pertinent responses to this statement, of which we will offer only two. First, Laden’s use of word “reality” is misleading. No one questions that real skeletons were found in Liang Bua Cave on Flores in 2003 – indeed, numerous prehistoric human burials have been unearthed on that island since the 1950s (van der Plas, 2002; Villa, et al., 2012); the evolutionary and taxonomic interpretations of the Liang Bua Cave skeletons are in question, not their physical existence. Does Laden take interpretations for “reality”? Second, Laden’s is not even a particularly original speculation about these remains. On January 29, 2007, Karl Zimmer wrote in his blog, The Loom, “To discover a chain of hominid species stretching from the mainland out to Flores would be cool beyond belief. When did the dwarfing begin, one wonders? Was there a Homo sulawesiensis before a Homo floresiensis? Or were the only hominids exploring these tropical islands members of our own species? Here’s hoping 2007 provides some answers.”

The year 2007 passed without the answer hoped for by Zimmer. Indeed, five years later, despite substantial research support provided to the “new species” advocates by the National Geographic Society, The Australian Research Council, and others, we’re still waiting for resolution that does not involve stifling of disagreement. This situation might be remedied if instead of asking about a new data set, not “Wouldn’t it be “cool” if [fill in your own speculation here]…?” but instead “What are some alternative hypotheses about the sample, and how might evidence be used to test them?” (which is, we think, what a Peter Medawar or a Richard Feynman would ask). This is where data, and use of accepted philosophical principles for evaluating those data, come in if one really is doing “zealous research” per the Sigma Xi motto, rather than just trolling for blog hits.

Perhaps by way of establishing his own credentials beyond those of a blogger, Laden notes that he had the opportunity to examine the original Taung fossil under the guidance of Philip Tobias. One of us (MH) in 1990 actually took over from Phillip Tobias headship of the department holding these fossils as well as a profusion of others and for the next five years had an opportunity to familiarize himself with these rich fossil collections as well as to conduct his own excavations in the fossil-bearing region rather than having a passing glimpse of one specimen under the guidance of a respected senior figure in the field. What Laden does not mention, perhaps because he is unaware of the fact, is that Professor Tobias (very recently deceased, a notable loss to the field) was one of the referees who read and supported our group’s paper in PNAS containing the rejection of the putative new species “Homo floresiensis.” Instead, we identified LB1 as a developmentally abnormal (e.g. unusually asymmetrical craniofacially, with femora strikingly short, straight and tubulated, etc.) individual member of a modern Australomelanesian population with multiple characteristics (gnathic, dental, postcranial) that are found routinely among populations in its region, including people (the Rampasasa) still living near the cave (Jacob, et al., 2006). Phillip Tobias also wrote an enthusiastically supportive introduction to our last book on the subject of Flores finds (Henneberg, Eckhardt, Schofield 2010).

What about Laden’s qualifications to express an opinion about this problem in human evolution, implied in his use by The American Scientist as a reviewer? He hints at inside knowledge of the field, as by the statement “I thought that I had already heard all of the scuttlebutt….” However, there is a lot more to interpreting evidence responsibly, which is more important in science than familiarity with rumor and innuendo. For example, one wonders how extensive was his examination of the Taung fossil. Laden states “Taung’s brain and teeth were more apelike…but Taung’s body was more humanlike in that it was apparently configured for upright walking.” Really? What body? The Taung find comprises much of the face and mandible plus a partial brain endocast – but no postcranial remains, i.e. no “body.” Raymond Dart’s prescient interpretation of upright posture for Taung was all the more brilliant and intellectually courageous for its being an inference from Taung’s extremely limited craniofacial remains. Actual postcranial remains of australopithecines were decades in coming to confirm his insightful extrapolation (and never have been found for Taung). Evidence and inference both are important, but they are not the same. Laden’s failures to distinguish among data, reasoning, and rumor (“scuttlebutt”) seriously flaw his review. We suspect that Meave Leakey might agree with us about his familiarity with the field after her complexion loses its light purple hue (“Mauve” to Laden).

So much for Laden’s reliability as a reviewer. What of his main conclusions about The Fossil Chronicles? Falk’s selective autobiographical volume has two quasi-related foci: a long-running controversy, chiefly with Ralph Holloway, about the interpretation of fossil hominin endocasts, and her opinions on the Flores skeletons. Any reviewer of the book could choose from a wide array of perspectives. Among various alternatives, she or he could summarize these topics neutrally, or provide an independent perspective based on additional evidence, or offer informed criticism of the author’s potentially subjective positions. Laden does none of these; instead he simply echoes Falk’s viewpoints on both subjects. His biases, as Falk’s, are set forth clearly. “The academic fight between Holloway and Falk was typical of the 1980s. Falk was a woman working in a largely male field, and some people wondered whether Holloway might be trying to marginalize her for that reason…. Anyone interested in a study of both gender bias and the ad hominem argument would do well to look at these papers.” To those statements we would reply: Many people may wonder about many things, but unsupported conjecture doesn’t – at least shouldn’t – count for anything in science in general, or in The American Scientist in particular. Nowhere in Laden’s review does he present any evidence in support of his serious professional disparagement of Ralph Holloway, unless one counts his own unpublished grad school term papers to which he alludes. We had thought that Sigma Xi’s motto was “Companions in zealous research,” not something more like “collaborators in undocumented defamation.” In a journal published by our Society one might have expected not innuendo, but evidence (though we have been here before: see Eckhardt, 2007b, 2007c). There are plenty of data that bear on this point, allowing investigators to get beyond the superficially journalistic approach of “He said, she said.” For interested readers, here are two papers among many that can be referenced, one published before Laden’s review (Vannucci, Barron, and Holloway, 2011), the other after (Holloway, 2012).

Just as Laden echoes Falk’s opinion of Holloway’s endocast research, so does he repeat uncritically her conclusions about the Flores remains, even to the specific arguments in support of their status as a new species of human. “Because the female Hobbit and the fossils found with her are so unusual, it has been suggested that their peculiarities are attributable to some disease or abnormality, such as microcephaly. Homo floresiensis is normal, although very different than [sic: from] expected, just as Taung was different than [sic: from] expected.” Who is supposed to have expected what about the Flores remains, and why?

In terms of formal logic, argument by analogy, as with Falk and Laden extrapolating from the reception of Taung to that of the Flores skeletons, is an instance of reasoning from the particular to the particular. A problem here is that in the reconstruction of human evolution, the data set has become so large that there are not merely two particulars for evaluation, but a great many. As drawn by Laden, comparison is simply between two skulls, Taung and LB1 (since the LB1 specimen has the only known cranium from Liang Bua Cave). They lived about 2 million years apart on two different continents. At about 400 ml the LB1 endocranial volume is only about one-third the average in normal extant humans, but whether it is microcephalic or not depends on the reference population, which is part of the matter at issue. As far as the argument from one particular specimen to another particular specimen, some context is necessary to avoid analyses that are simplistic, as are those of Falk and Laden. Over the nearly two centuries that human fossils have been accepted, thousands of specimens have been recovered (the numbers differing depending on how one counts fragmentary and partial finds and the temporal and taxonomic boundaries that are used, but one can say that some 210 fossil hominin skulls are known well enough to reconstruct brain sizes (Henneberg and Saniotis 2009, Saniotis and Henneberg 2011). Throughout this time the skulls have occasioned much discussion and debate on various grounds. Comparing the reception of just two skulls, in the presence of a hundredfold more that appear to be deliberately ignored by people claiming familiarity with the field, may at best suggest a bias.

There is some truth to the contention that novel discoveries sometimes are resisted (for documentation of this point see Eckhardt, 2000, particularly the chapter titled “A century of fossils”). However, resistance does not always occur because the new find is held to be pathological. Taung was dismissed by Sir Arthur Keith and others as a human ancestor, though on other (quite specious) grounds than pathology. Similarly, Keith excluded the earliest known fossils of Javan pithecanthropines from human ancestry because they were, in his chronology, too primitive and too recent to qualify as ancestors. Although Rudolf Virchow and others did hold that the Neanderthal remains discovered decades earlier were those of a modern human afflicted with arthritis, rickets and syphilis, the previous earlier description, by Fuhlrott and Schaffhausen, held them to be normal, and Thomas Huxley championed them as ancestors of extant humans. Schaffhausen and Huxley have been proven correct. In other cases, diagnoses of morphology-altering pathology have been substantiated, notably without the specimen in question being proposed as the type specimen of a new human species (Walker, et al., 1982).

It also should not be overlooked that some finds have been acclaimed widely as ancestral humans for decades, only to be rejected much later because the evidence eventually failed to stand up to close scrutiny. Piltdown, mentioned by Laden, is a well known example; its fraudulently fabricated physiognomy was the reverse of Taung’s total morphological pattern, thereby constituting a barrier to acceptance of the Dart’s evidence and reasoning. Another unfortunate example, seemingly rooted in wishful thinking rather than fraud, was Ramapithecus. This taxon, which was advocated as a hominid by Elwyn Simons (1961), misled most professionals and, by becoming standard textbook fare, students as well for about two decades. The error persisted even though its status was based on very little physical evidence, was at variance with growing molecular data, and was separated from the otherwise earliest known humans, australopithecines, by a gap of about ten million years. Examples as disparate as Piltdown and Ramapithecus share a common element: they illustrate the episodic occurrence of a “bandwagon effect” in the reconstruction of human ancestry, with many professionals who have little knowledge of the evidence wishing not to be seen as lagging behind the latest intellectual fashion wave. My favorite firsthand example of this phenomenon was a conversation that I had several years ago at a professional meeting with an anthropologist who is a member of the National Academy of Sciences. I had expressed strong doubts that the Liang Bua Cave skeletons represented a new human species, noting that nearly all of the supposedly defining features of the new taxon were based on the single skull that existed. Her reaction was to insist that I had to be wrong because there were multiple skulls, all sharing the same distinctive anatomy. She knew, because she had been teaching about the find. Chivalry prevents me from identifying this person, but perhaps we can add the incident to the undocumented scuttlebutt of this colorful field, since that seems acceptable in the context of The American Scientist.

What protects scientific disciplines against manifestly erroneous examples of the sort discussed here? Laden seems to rely for validation on consensus, without any discussion of how widespread agreement might be reached other than spontaneous disappearance of the mental illness of a minority: “The history of the study of human evolution shows that surprising findings make well-educated and otherwise rational people behave irrationally for a time, until everyone eventually settles on a new view.” Statements of this sort in fact help us to understand why there is a bandwagon effect: The implication that scientists who do not accept the new view are irrational. The “irrationality” label carries a price to those onto whom it is pinned. Just seeming to be with the majority can carry more ready access to funding and publication, as has been the case so far for the advocates of “Homo floresiensis.”

Are there reasonable criteria for judging the intrinsic correctness of hypotheses other than popularity? Sure. One is the principle commonly attributed to Carl Sagan that “extraordinary claims require extraordinary evidence” (which is an updated version of Hume’s Maxim). But of course this principle raises another question: Which of several not very likely possibilities is the more extraordinary claim? This question lies at the heart of the continuing controversy surrounding the skeletons from Liang Bua Cave: Do they all uniformly represent an entirely new, hitherto unknown human species or is there evidence that the sample might include one or more unusual members of our own?

Experienced scientists and philosophers of science know that the system of data collection and analysis in which we operate is not one that, particularly in unusual situations, inevitably yields correct answers (by which we mean those that are in accord with how nature has operated in the past and continues to operate, to make a complex matter simpler than it sometimes is), particularly in the short run (Dart waited three decades for vindication). Again, however, there are operating principles that furnish some guidelines. One is the hallowed “Occam’s Razor.” Here I am paraphrasing and summarizing several cogent presentations readily available on Wikipedia (re Occam’s Razor) and Skeptico (re Burden of Proof). In principle Occam’s Razor commonly is taken to mean that, other things being equal, a simpler explanation is preferable to a more complex one; realistically, in practice one seeks the simplest explanation that does not reduce the explanatory power of a hypothesis. Of critical importance but often overlooked is the principle that the implicit burden of proof is on the person asserting the claim (in this case, that the Liang Bua Cave skeletons represent a new species). When this principle is violated the consequence is that the burden of proof is shifted unfairly to critics of the claim. This shift introduces yet another logical fallacy — that a proposition is true if it has not yet been proven false. In the case of the Flores skeletons Bertrand Russell’s version of Occam’s Razor is particularly apposite: “Whenever possible, substitute constructions out of known entities for inferences to unknown entities.” It seems that, paradoxically for how science is thought to operate, some very bizarre aspects of the “new species” interpretation have heightened interest in it rather than raising appropriate skepticism. Situations of this sort are compounded further if attempts at hypothesis testing are impeded by restricting access by the critics to the primary data whose interpretation is in question, as demonstrably is the case here. A notable parallel case in physics recently was documented tellingly by Eugenie Samuel Reich in Plastic Fantastic (2009). Instances of this sort occur when, in the face of astonishing and exciting claims (the sort that inspire phrases such as “Wouldn’t it be cool if…) critical judgment is suspended by journal editors who exchange their appropriate roles as gatekeepers for the more glamorous ones of advocate and promoter.

Against this background, we see the “Liang Bua Cave skeletons” controversy as being a perfect scientific storm that brings together several winds, each of gale force:
1. Disregard for data. Many of the supposedly “unique” features of the supposed new species (very short stature, low humeral torsion, etc.) are not unique to the Liang Bua Cave skeletons at all, but occur, individually and in combination, in other extant human populations, others (e.g. mandibles lacking external chins, rotated premolar teeth) are notably common in extant Australomelanesians, including the nearby Rampasasa, while the tiny (400 ml) endocranial volume, abnormal asymmetry, and unusually short femora of LB1 offer evidence of individual developmental abnormality. Published data disproving the supposedly unique features are not addressed, while illustration of asymmetry and other abnormality has been denied, thereby going beyond the negative phenomenon of shifting the burden of proof, to patently ignoring documented disproof. This aspect of the problem further concerns the predictive value of hypotheses.
2. Indifference to hypothesis formulation and testing. Our group’s hypothesis (Jacob, et al., 2006) has remained unchanged from the beginning of our work on this problem (LB1 is an abnormal individual member of the expected modern human regional Australomelanesian population still represented on Flores). In contrast, supporters of the new taxon not only have multiplied hypotheses, but the original one (the new species evolved in isolation on Flores over more than 800,000 years from a Homo erectus ancestor) was changed in 2007 (to hold that the new species evolved its small brain and short stature before reaching Flores, and is derived from an African taxon that lived more than a million years earlier, at the pre-erectus level, sharing a common ancestor with Homo habilis). These formulations appear to us as mutually exclusive, but the contradictions do not appear to yet have been either reconciled by the proponents of the new species or the difficulty commented on by science writers and other journalists.
Our hypothesis originally was formulated in response to the data known at the time of the discovery. Subsequently, Berger, et al. (2008) described a sample of relatively recent (about 940 to 2890 years bp) small-bodied human skeletons from Palau, combining relatively short stature with reduced or absent external bony chins, but with all known partial crania being around the low end of the distribution of small-bodied humans (very roughly, about 1200 ml). Thus, for example, absence of an external bony chin manifestly is not necessarily primitive or even unusual, while on the other hand a single endocranial volume of about 400 ml (LB1) falls outside the normal range for another relatively recent small-statured human population from the same geographic and cultural area that includes Flores. The response to the Palau find from supporters of the view that the Liang Bua Cave remains must represent a new species is not data analysis but denial: “There’s nothing new about these little guys; … this offers zero window into Homo floresiensis…” (W. Jungers, quoted in Culotta, 2008). Jungers’ response echoes uncannily the trope often used by police officers at crime scenes: “Move along, there’s nothing to see here.”
3. Preference for myth over method. “Unknown entities” in the sense advised against by Bertrand Russell abound in the widely but uncritically accepted new species hypothesis. In fact the supposedly inexplicable elements are emphasized, to the obvious delight of journalists, most of whom have no notable scientific background. Particular examples of this gambit can be seen in the PBS special titled “Alien from Earth.”

We close by noting that from the first publications in 2004 through the present time, the idea that the Liang Bua Cave remains represent a new species has – unlike Taung but like Piltdown and Ramapithecus – decidedly not been a minority viewpoint championed bravely by a few scientific pioneers against widespread opposition. Instead, from the first the new diminutive species was a popular, widely accepted interpretation (which, like the hominid status of Ramapithecus, already has entered the textbooks). Its relatively few critics, in contrast, are held, as by Laden here, to be irrational. It is our position that this situation will change, but only if there is more general realization that the approach of “wrapping events into a good story that supports one’s case is far more persuasive than poking holes in somebody else’s story” is alien to the best traditions of science, however appealing to journalists. We hope that we will not be characterized as irrational for as long a period as was Raymond Dart (roughly three decades; we’re already up to year eight) but find that situation, however difficult, is preferable to joining a broad consensus that has coalesced prematurely and for the wrong reasons. To quote Sloman in closing as in opening, isn’t it time for “a fair and balanced argument”?

REFERENCES

Culotta, E. 2008. Hobbit redux? (Science Now, 11 March).

Eckhardt, R. B. 2000. Human Paleobiology. Cambridge, England:Cambridge University Press.

Eckhardt, R. B. 2007a. Paleoanthropology: A science in need of a theoretical framework: The Liang Bua example. Pages 107-116 in Recent Advances in Southeast Asia, Paleoanthropology and Archaeology, Ed. E. Indriati. Yogyakarta, Indonesia:Yogyakarta Faculty of Medicine.

Eckhardt, R. B. 2007b. Letter to the bookshelf (response to review of A New Human). American Scientist Online, August 10, 2007.

Eckhardt, R. B. 2007c. Errata in review by Peter Andrews of A New Human.American Scientist 95(5):382.

Henneberg M., 2009 Two interpretations of human evolution: Essentialism and Darwinism, Anthropological Review 72:66-80.

Henneberg M., Eckhardt R. B., Schofield J. 2010 The Hobbit Trap, How New Species are Invented (revised second edition). Walnut Creek, California: Left Coast Press.

Hennberg M, Saniotis A, 2009, Evolutionary origins of human brain and spirituality. Anthropologischer Anzeiger 67(4):427-438.

Henneberg M, Thorne A. 2004. Flores may be pathological Homo sapiens. Before Farming 1:2–4.

Holloway, R.L. 2012. The issue of brain reorganization in Australopithecus and early hominids: Dart had it right. Chapter 9 in African Genesis: Perspectives on hominin evolution. Eds. S. C. Reynolds and A. Gallagher. Cambridge University Press. Cambridge, England.

Jacob, T., Indriati, I., Soejono, R., Hsü, K., Frayer, D. W., Eckhardt, R. B., Kuperavage, A. J., Thorne, A., and Henneberg, M. 2006. Pygmoid australomelanesian Homo sapiens skeletal remains from Liang Bua, Flores: population affinities and pathological abnormalities. Proceedings of the National Academy of Sciences of the USA 103:13421-13426.

Laden, G. 2012. Two journeys through the human past. American Scientist 100:75-78.

Martin R. D., Maclarnon A. M., Phillips J. L., Dobyns W. B. (2006) Flores hominid: New species or microcephalic dwarf? Anat Rec A. Discov. Mol. Cell Evol. Biol. 288:1123– 1145.

van der Plas, M. 2002. A New Model for the Evolution of Homo sapiens. Pp. 1-146. M.S. thesis, Department of Biology, Leiden University.

Reich, E.S. 2009. Plastic Fantastic. New York: Palgrave Macmillan.

Saniotis, A. Henneberg, M. 2011 An Evolutionary approach towards exploring altered states of consciousness, mind-body techniques and non-local mind. The Journal of General Evolution 67: 182–200.

Simons, E. L. 1961. The phyletic position of Ramapithecus. Postilla 57: 1-9.

Sloman, S. 2012. The battle between intuition and deliberation. American Scientist 100:73-75.

Vannucci, R. C., Barron, T. F. and Holloway, R. L. 2011. Craniometric ratios of microcephaly and LB1, Homo floresiensis, using MRI and endocasts. Proceedings of the National Academy of Sciences of the USA http://www.pnas.org/cgi/doi/10.1073/pnas.1105585108

Villa, C., Persson, L., Alexandersen, V., and Lynnerup, N. 2012. A small skull from Flores dated to the 20th century. Homo 63:12-20.

Walker, A., Zimmerman, M. R., and Leakey, R. E. F. 1982. A possible case of hypervitaminosis A in Homo erectus. Nature 296:248-250.