Research Article

Small-Bodied Humans from Palau, Micronesia

  • Lee R. Berger mail,

    Affiliation: Institute for Human Origins and the Bernard Price Institute for Palaeontology, School of GeoSciences, University of the Witwatersrand, Johannesburg, South Africa

  • Steven E. Churchill,

    Affiliation: Department of Biological Anthropology and Anatomy, Duke University, Durham, North Carolina, United States of America

  • Bonita De Klerk,

    Affiliation: Institute for Human Origins and the Bernard Price Institute for Palaeontology, School of GeoSciences, University of the Witwatersrand, Johannesburg, South Africa

  • Rhonda L. Quinn

    Affiliation: Department of Earth and Planetary Sciences, Rutgers University, Piscataway, New Jersey, United States of America

  • Published: March 12, 2008
  • DOI: 10.1371/journal.pone.0001780

Reader Comments (7)

Post a new comment on this article

Comments from Referee 2 (Robert B. Eckhardt)

Posted by PLoS_ONE_Group on 11 Mar 2008 at 10:19 GMT

Palau: Independent test of the Flores hypothesis
[Subtitle: No hobbits in this shire (Palau), either*]

To begin with, I suspect that many of the comments that will be written by others about "Small-bodied humans from Palau, Micronesia" by Berger, Churchill, De Klerk and Quinn will be devoted to critical comments focusing on what the Palau material described here is not: Very likely it will be said by more than a few paleoanthropologists that the Palau sample is not pertinent to tests of hypotheses about the Liang Bua Cave skeletons from Flores, particularly that of the most complete specimen found there, LB1. I would be surprised, in fact, if the majority of the comments on this paper are not negative. Since the beginning late in 2004 of the controversy over the Flores skeletons, my estimate is that roughly 80% of those who consider themselves to be paleoanthropologists think that “Homo floresiensis” is a valid new species of hominin. Judging from the array of papers and posters presented at the 2007 Annual Meeting of the American Association of Physical Anthropologists, this percentage distribution has remained fairly constant for about three years.

Against this background, from the experience of the last several years, members of our own international research group (for membership in which see Berger, et al., reference 4) often have encountered such illuminating scientific comments on our work from Morwood group collaborators as “rubbish” (too often to bother tabulating), and such fascinating morphological assessments as “Robert Eckhardt is thick as a plank,” (Peter Brown, January 2006 Discover magazine [this characterization has been falsified, however, since in a subsequent scientific meeting at the University of Pennsylvania, 10 February 2006, my wife, Carey, used an anthropometer to demonstrate that I am, in fact, thicker than two short planks]). Just to make sure that everyone in the game understood that what sport aficionados refer to as “trash talk” was officially endorsed, Nature (31 August 2006) “warmly welcomed” [their phrase] our group’s detailed paper in Proceedings of the National Academy of Sciences of the United States (PNAS) [reference 4 in Berger, et al.] under the editorial title “Rude paleoanthropology.” Against this background of experience, I suggest that Berger, et al., as well as readers of this journal, be ready for all sorts of attempts at dismissal of the work at hand and its importance. Forewarned is forearmed.

Rather than what it is not, though, we should begin with what the paper by Berger, et al., is; their own words serve just fine in this regard: “We feel that the most parsimonious, and most reasonable, interpretation of the human fossil assemblage from Palau is that they derive from a small-bodied population of H. sapiens (representing either rapid insular dwarfism or a small-bodied colonizing population), and that the primitive traits that they possess reflect either pliotropic [sic] or epigenetic correlates of developmental programs for small body size.” Much of the rest of their paper describes the geographic and temporal settings, plus some detailed, professionally competent, morphological descriptions of the Palau skeletal material. There is no need to repeat those descriptions here, but they are well worth reading, and re-reading.

Beyond its own intrinsic descriptive and analytical value deriving from the data provided,"Small-bodied humans from Palau, Micronesia" also is a most important and timely paper in the controversy that has arisen over the interpretation of the skeletal material recovered from Liang Bua Cave, Flores, Indonesia and first published by Nature in October, 2004. From the beginning of this matter, our research group (see references 4 & 27 in the manuscript by Berger, et al.) has stressed that there are two major elements in the refutation of the morphological bases of the contention by Brown and Morwood that "Homo floresiensis" is a valid taxon. First, the collection of skeletal material from Flores presents the expected regional (Australomelanesian) characteristics of extant humans from the geographic area (which we now see obviously includes Palau) in which they were recovered. Second, the most complete individual specimen, LB1, is recognizably abnormal developmentally (not only is the endocranial volume far too small even for its relatively short stature, as documented repeatedly by Robert Martin and his colleagues [e.g. reference 28 in Berger, et al., here] against vigorous but contradictory asseverations by Dean Falk and her helpers, but the LB1 specimen manifests marked craniofacial and postcranial asymmetries, as well as other skeletal signs of abnormal development such as tubulated femora, low humeral torsion, etc.). Furthermore, both stature and brain size were exaggerated downward in the initial and most subsequent reports [see Eckhardt, et al., 2005, cited as a footnote within reference 4 in Berger, et al.].

Also worthy of note is that the supposed evolution in situ of a novel hominin species on Flores from Homo erectus ancestors due to island isolation from the beginning was a postulate (something assumed as a precondition for an argument) rather than a matter of convincing empirical proof. Never likely, but widely appealing to uncritical commentators, this part of the scenario now has been abandoned even by Morwood (personal communication, Jogjakarta, 25 July 2007), who now derives "H.floresiensis" from H. habilis or some other pre-erectus species, and asserts that that the small brain and body size of the Flores skeletons existed prior to arrival on the island. Thus the popularly appealing but unnecessary initial framing hypothesis to explain the supposedly “unique” Flores hominin morphologies has been abandoned in favor of one even more nebulous and less testable. Further undercutting the original claim for about 800,000 years of isolation by “H. floresiensis” are the independent observations that a new species of giant rat has been found on the not-so-tiny land mass of Papua by the Conservation International Expedition of 2006, and the strong doubt cast on the occurrence of dwarf stegodons with “H. floresiensis” (by N. Rolland & S. Crockford, in Antiquity 79, 14 June 2005). Giant rats and dwarf stegodons formerly had been widely accepted supporting bits of evidence for the protracted island endemism on Flores. As we already have shown from the geological evidence for sea level fluctuations, such endemism is, to be charitable, highly unlikely.

It is against this background that the Palau specimens and their analysis by Berger, et al. become so important. Note that, to begin with, Palau presents small human body size without any need to postulate isolation. Much more follows from this positive evidence, as well as the absence of any cranial volume outlier such as LB1 on Flores. The overlapping archeological and paleoanthropological communities for the most part largely have been unable to understand and/or accept the manifest biomedical evidence for the developmental abnormality in LB1 and have persistently interpreted regional features of this and the associated skeletons as "primitive" characteristics (with all of the unattractive archaic and socially repugnant concomitants of such conceptions of normality and modernity). But now Berger, et al. meet archeological-based speculation with archaeological-based evidence, analysis, hypothesis test, and refutation.

From the very beginning of the Flores hypothesis – a characterization that I prefer to “Homo floresiensis problem” or the cloyingly cutesy “hobbit” references – there has been an identifiable pattern: bold – one might even say “wild” -- claims based on inadequate background research, followed by refutation (by our group and others), then spin by the original claimants without answering the evidence that countered their original claims, succeeded by new claims also based on inadequate background research. It is impossible to follow all of the iterations of this pattern in the space available here, so I will provide only a few of the many possible examples. Bold claim: the Flores skeletons represent an entirely new species of hominin, in which small brain size (based on a single specimen) and small stature figure prominently. Inadequate research: the initial paper by Brown, et al. (Nature 2004) considered only a very few medical syndromes (IGF-related postnatal growth retardation, pituitry dwarfism, primordial microcephalic dwarfism), while there are literally hundreds of syndromes that include these features. Refutation: Jacob, et al., 2006, [ref. 4 here] account for all of the originally described morphological features of LB1 – cranial, facial, dental, post cranial, every one (including supposedly uniquely low humeral torsion). Spin: For one such feature, humeral torsion, an initial estimate of of 110º for LB1 (Morwood, et al., Nature 2005), a figure that – quite remarkably – was repeated in the abstract of the paper by Larson, et al. (Journal of Human Evolution 2007), thus: “…a nearly complete right humerus (LB1/50). Although the humerus appears fairly modern in most regards, it is remarkable in displaying only 110º of humeral torsion, well below modern human average values.” However, in the same article (p. 722), Larson, et al. subsequently wrote “The average of the two new torsion measurements for LB1 is 120º,” and their Fig. 4 shows that the 120º value falls within the 95% confidence level for 6 of their 15 samples of extant humans (!). Only very rarely does a paper contain an internal refutation of its own incorrect abstract. New claims based on inadequate background research: Those “primitive” little wrist bones that are not shown in any form other than idealized multicolor cartoon shapes in their original presentation (Tocheri, et al., Science 2007) -- this for structures that in extant humans commonly are characterized by orthopedic surgeons as highly variable, with some of these variants being referred to frankly as atavistic. Refutation in hand, with no pun intended.

This paper by Berger, et al. clearly is a preliminary report based on specimens that have not been prepared fully yet, and of course must be analyzed in greater detail in the future (incidentally, it should be noted that as of the last time that truly independent observers could examine them, the Liang Bua skeletons were far from fully prepared for study, either; this might account for the numbers of individuals reported in various accounts ranging widely, or wildly, from 6 to 11). The core findings presented here are that the Palau specimens exhibit roughly the same statures as the Liang Bua skeletons , yet have endocranial volumes that are about 2.5 times as large, thus falling at the lower limits of extant humans. These findings will be refined, though it is all but inconceivable that they will change materially, a point made correctly by the authors. They should be encouraged to proceed to the more detailed, possibly monographic, treatment that is merited by this research.

For those who wonder why, in the face of hypotheses that have been refuted again and again, “Homo floresiensis” is reified repeatedly, realize that the Flores find from the first was hailed as the most important discovery in the field for the last fifty years. Furthermore, as noted above, a majority of workers in the field persistently have supported the “new species” claim. Thus this hypothetical new species has become a proxy for the paradigm of an entire field: One skull is enough to serve as the type specimen of a new species. However, if LB1 proves to be developmentally abnormal, as repeatedly has been shown to be the case, this datum must be denied, for admitting it would disqualify LB1 as a type specimen according to The International Code of Zoological Nomenclature, Article 1 (b) (2) “for teratological specimens as such.” And there is no other skull that looks like LB1, not on Flores, and, as we know now thanks to Berger, et al., not on Palau, either – or anywhere else yet known.

*”No hobbits in this shire” was the science writer’s title for the press release at the time of publication of
Jacob, T., Indriati, E., Soejono, R. P., 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 the USA 103:13421-13426.

Robert B. Eckhardt
Professor of Developmental Genetics and Evolutionary Morphology
Laboratory for the Compartive Study of Morphology,
Mechanics and Molecules
Department of Kinesiology
The Pennsylvania State University
University Park, PA 16802