Monday, January 11, 2016

Homo naledi shows how biological anthropology is not biology, and can't be, and shouldn't be

            I once read somewhere that the most interesting thing about human evolution is how everbody thinks they understand it.

            I suspect it's because everyone thinks they own a piece of it.  It's the story of where we came from, after all!  And not just any story of where we came from - it's the authoritative, scientific story.

            The authoritative origin stories are not like other stories.  They are value-laden in ways that other scientific stories are not.  Archaeology is routinely used in the service of nationalism, for example.  Rather moreso, at least, than fruitfly genetics is, so a fruitfly geneticist, or a general biologist, might be excused for not being an appropriately critical reader of the literature on human evolution or diversity, where there is rather more at stake.  It is a different and unfamiliar literature to them, and consequently requires some additional intellectual effort for a trained biologist to make sense of. Some don't bother.

            Now, there have been some very insightful contributions to the scholarly literature on human variation and evolution from biologists, even fruitfly geneticists, over the years.  I can think of three off the top of my head.

Julian Huxley collaborated with the Cambridge social anthropologist Alfred Cort Haddon for this important early critique  of race, We Europeans. Rob DeSalle collaborated with biological anthropologist Ian Tattersall on their recent book, Race? Debunking a Scientific Myth.  And Doby was a friend and collaborator of several anthropologists, including Sherwood Washburn, Ashley Montagu, and Margaret Mead.

            In science, our answers to the question of where we came from are stories that center around a descent from the apes.  And our characters are already there for us: The human lineage is composed of species, just like the units of paleontology and ecology. A recent ethnographic paper by Eben Kirksey begins, "Taxonomists, who describe new species, are acutely aware of how political, economic, and ecological forces bring new forms of life into being." That is probably true, but I think generally not in the first person. That is to say, the taxonomist working with "political, economic, and ecological forces" is usually somebody else; I'm the taxonomist who is uncovering raw nature.

            Back in 1945, paleontologist George Gaylord Simpson was reviewing the literature on  mammal taxonomy, but when he got to humans, he found it impenetrable.  He had an idea why it was so impenetrable to him, as well: “A major reason for this confusion is that much of the work on primates has been done by students who had no experience in taxonomy and who were completely incompetent to enter this field, however competent they may have been in other respects”.[1] 
GGS in 1983

            Granted that many of the workers in the field may have been trained principally in medical anatomy rather than in evolutionary paleontology, Simpson thought it was reasonable to expect that an expert on  the species of other kinds of mammals should be able to translate freely to the literature on human evolution, because the units ought to be the same.  But he misunderstood the species in our own lineage, for these taxonomic entities are not like the taxa of biology.  Simpson hoped to study his ancestors dispassionately and rationally, as perhaps Vulcans  contemplate their ancestors.[2] But a purely rational and logical Vulcan approach to ancestry involves not dividing people into relatives and non-relatives, for they acknowledge that rationally and logically, everyone is related. They also do not consider ancestry beyond the twelfth generation (approximately 300 earth-years, because in the 12th generation, every sexually-reproducing organism had 4096 ancestors, which is rather a lot to track; and each contributed less than 1/40 of 1% of the genome, so none of them on average is particularly genetically significant).  But we aren’t Vulcans, we are Earthlings, and we treat our kinship and descent in all kinds of meaningfully irrational (but nevertheless coherent and logical!) ways, even in science.

            The classification of our ancestors is still vexed.  Sure, scientists acknowledge some of our colleagues to be “lumpers” or “splitters” – interpreting anatomical diversity among the fossils to be the result of age, sex, pathology, deformation, and microevolution,  thus “lumping” the fossils into few species; or conversely “splitting” them into many species by interpreting the anatomical diversity taxonomically.  But there is something else going on here.  This is participation in the construction of an authoritative story of our ancestry.  There is simply more at stake than in the narrative of clam or deer ancestry. The units here, the species, are not comparable to the species the zoologist is familiar with, for these are not units of ecological genetics, but units of story.
[from a forthcoming paper in Philosophy, Theology, and the Sciences]

           The lumper story is one of the continuity and survival of the lineage; the splitter story is one of diversity and extinction of different lineages. Is the story of our ancestry like a tree trunk, or like a bush? The lumper inclines to the former; the splitter inclines to the latter. But those are significantly different shrubbery metaphors to be imposing upon the same sample of fossils. 

            Some decades after Simpson lodged his complaint, paleobiologist Tim White reiterated it, while reviewing a book on the history of "the" 22 species in our lineage: “Many of the putative species are chronotaxa; others are not even valid species in that sense. No one really thinks that available hominid fossils represent 22 separate species lineages in the last six million years.”[3]  Except, possibly, for the authors of the book under review.

           Or perhaps they didn’t really believe it either.  The assumption here that needs to be interrogated is that the fossil taxa of other groups of animals  are comparable – are made the same way, for the same reasons, of the same elements – as the fossil taxa of our own ancestors.  And that is the key error: Fossil animal species are units of biology; fossil human ancestors are bio-cultural units of narrative.  This is not to say that they don’t overlap, and that there were no zoological species in our ancestry.  The problem is that those zoological species are inaccessible to us, and so – rather like the angels sitting on the pinheads – we can see different numbers of species and tell quite different stories from the same empirical database.  This is consequently not an empirical issue at all, but a hermeneutic issue.

A recent book by a historian tells readers on its cover that 100,000 years ago “at least six human species inhabited the earth.”  Yet few practicing biological anthropologists would come up with the number six as the target number of species in the human lineage that inhabited the earth 100,000 years ago; and far fewer would acknowledge the particular six that the author does: Homo sapiens, H. neanderthalensis, H. erectus, H. soloensis, H. denisova, and H. floresiensis.  After all, “H. denisova” has not been formally named, and is based on the genome of a Siberian finger bone, which is itself simply a variant of the Neanderthal genome, which is not clearly a different species in the first place, since recreational genomic ancestry services (for about $200) will now identify the circa 5% of your genome that ostensibly comes from Neanderthalswhich sounds very un-species-like.

            Is the tally right or wrong, then? It is actually neither.  We can’t say, because the zoological answer is inaccessible to us.  These are units of mythology, not of zoology.

            Another of the six presumptive species 100,000 years ago is Homo soloensis.  That name is a linguistic marker, denoting a particular set of Indonesian fossils, anatomically continuous with Homo erectus before and with Homo sapiens after.  As such, it is a named place-saver for a part of the human lineage – a rivulet, or capillary, or rhizome that better represents its elements metaphorically than a tree-limb does.  But in the words of Tim White, “no one really thinks” that this set of fossils represents a valid zoological species of their own.  There is no Homo soloensis. In other words, the ontological status of Homo soloensis is the same as that of Mother Corn Spirit. Neither is a unit of nature, but a unit of meaning or narrative which, to a believer, is perfectly sensible in the context of a story about origins. Homo soloensis is something, but it is not a zoologically familiar species, a fact of nature, so to speak. It is a named fictive ancestor, with more symbolic than naturalistic properties.  In the most fundamental way, human ancestry is self-consciously a story, and taxa like Homo soloensis and H. denisova are the components of this particular historical account.

            And likewise, Homo naledi, the newest major addition to our family tree. “But is it real?“ some journalists queried. Of course it’s real, you didn’t just imagine it.  “But is it real biologically?” they persist.  And that is my point: It doesn’t matter; Homo naledi is not an element of biology; it is an element of our origin story.  It is part of the bricolage of origin story-making.[4]  There is no true or false answer to Homo naledi as a zoological species; for the category of zoological species does not apply to things like Homo naledi.  The mistake here lies in assuming that Homo naledi designates a unit of zoology; that there is an underlying natural taxonomy in human ancestry that will be revealed by the proper ratiocination.  If you're looking for zoological reality, look for it at the genus level. It isn't there at the species level.

            Such is the long-standing taxonomic fallacy in grappling with the science of who we are and where we come from.  On a bio-political terrain, a preparation in biology is inadequate to comprehend the taxonomy, for it is not biological taxonomy.  To the extent that our ancestry is populated by species, those species are attempts to impose a taxonomic structure, which we assume ought to be there, upon an assortment of fossils from various times and places, with diverse anatomies, representing distinct lineages different from one another and yet connected in complex ways. There are a lot of ways of doing it, and they are all very sensitive to the conditions under which the science itself is practiced.

            If the Neanderthals and the Denisovans are not like zoological species, then what might they be like? And here we return to Linnaeus. They would be at most subspecies, as Linnaeus considered unfamiliar peoples to be.  In other words, the classification of extinct humans intergrades into the classification of extant humans. This fallacy – imposing taxonomic structure upon our ancestry, and mistaking the bio-political categories of our story for natural units – is the same fallacy we find at the heart of race.  For race, the meaningful story is “Who are we?” rather than “Where did we come from?” but the problem is the same, mistaking bio-political units of people for zoological units of people.  And those two questions are invariably intertwined, whether the answer comes from science or from any other system of explanatory narrative.

* This blog post is cobbled together from some forthcoming work, mostly Why is Science Racist? (Polity Press, 2017).

[1] Simpson, G. G. (1945) The principles of classification and a classification of mammals Bulletin of the American Museum of Natural History 85:1-349, quotation from p. 181.
[2] Vulcan is the “Star Trek” planet, notable for the overbearing rationality of its inhabitants.
[3] White, T. D. (2008) Review of The Last Human: A Guide to Twenty-Two Species of Extinct Humans, by G J Sawyer and Viktor Deak. Quarterly Review of Biology 83:105-106, quotation from p. 105.
[4] Claude Lévi-Strauss (1962) used the term “bricolage” to refer to the available elements a mythmaker draws on, while tinkering with them to construct a resonant story. It was borrowed by molecular biologist François Jacob (1977) to argue that evolution is more like a tinkerer than like an engineer. 


  1. Interesting. I hadn't thought before of the similarity of thought about "races" and hominid "species." Compelling point. Allows me to be comfortable again with a single-species (er, genus?) perspective.

  2. Weird essay. Some thoughts:

    1. It is indeed difficult to agree on what the hominin species are, but this is primarily a matter of having a lot of specimens (hundreds of partial specimens among dozens of putative species), but not enough (hundreds of specimens within each putative species) to make the kinds of fine statistical determinations that would determine for once and for all what the discrete populations are, and (even harder) when they merge, going back in time. Without that, we fall back on intuition.

    2. These problems are not unique to humans, they are common amongst fragmentary fossil records. The main differences are (a) there are way more workers in anthropology, so you get many different interpretations, whereas there are only a few people working on fossil canids, giraffes, etc.; and (b) the media and everyone else gives much more attention to "new species", versus "new specimens of an old species", so there is a clear motivation in that direction. That, more than grand metanarratives and storytelling and whatever, is what seems to be producing so many species.

    3. Genera are not more objective, they are a Linnaean rank with no real definition. What is clearly definable, and objectively confirmable or disconfirmable as data improve, is clades. Some genera are clades, but many fossil genera (especially) are paraphyletic and thus not phylogenetically real. At best a paraphyletic genus describes some expert's sense of morphological grouping. This is a pretty strange way to describe a continuous evolutionary process, but the Linnaean system is pre-Darwinian so this conflict shouldn't be surprising.

    Anyhow, I would encourage serious consideration of the phylogenetics revolution in paleontology in general (which has been massive; including abandoning of Linnaean ranks), and then consider its uptake in anthropology (which appears to be a few decades behind). I am convinced that the real solution to these sorts of taxonomy issues will come once the questions are matters of replicable statistical analysis -- rather like statistical phylogenetics has wholly replaced the old-fashioned method of drawing phylogenies by intuition.

    1. Thanks for your comment. I’ve tried to argue that waiting for biological anthropology to “catch up” is problematic today for the same reason it was in 1945. You might as well be waiting for organic chemistry to catch up. What you’re actually dealing with in human ancestry are data that are different in kind from the units of data that paleontologists and general biologists are familiar with. They resemble familiar biological things, but they are bio-political units. That is why they are so confusing to you; it’s not so much the intellectual primitiveness of the scientists.

    2. Given the common ancestry of all life, when do paleontological taxa cease becoming bio-political units? Is Australopithecus a bio-political unit? Proconsul? Aegyptopithecus? Morganucodon? Tiktaalik? While few if any paleontologists would claim the last four as contributing directly to the human lineage, they often used as bricolage in our common myth-making, albeit on a larger scale.

      Perhaps you would argue that these are genera, not species. As the commentator above pointed out, this distinction, especially in paleontology, seems largely arbitrary. In any case, when it comes to questions of higher-taxa ancestry and phylogeny, these genera act much in the same way as species; people might argue which genus of stem-mammal is closest to crown mammal in the same way paleoanthropologists argue which species of stem-Homo (Australopithecus) is closest to Homo.

  3. Homo or Australopithecus naledi is about biology, but is not necessarily about human ancestry. There was no deliberate burial, of course: the skeletons fossilised naturally. If we simply use biology, it's not so diffecult (in fact, predicted, google: aquarboreal): curved hand-bones for vertical climbing, flat humanlike feet more penguin- than ostrich-like, small brains, fossilisation in mud-stone, i.e. in stagnant water, etc.: naledi apparently lived at least parttime like bonobos wading bipedally in forest swamps, and climbing arms overhead in the brances above the swamp, google illustraitons at: bonobo wading. All the anthropocentric interpretations of Berger et al. are wrong: naledi didn't bury their dead, they didn't run over plains, they didn't make tools more than e.g. chimps do, and they might have been closer relatives of bonobos than of people. In any case, naledi *is* about biology.

  4. I agree even though the homo naledi can't be fit into a scientific category it's still apart of our evolutionary history-there isn't an awnser for everything.

  5. Hi Tess. Well, IMO naledi is more about chimp or bonobo evolution than about human evolution. Naledi looked much more like a bonobo than like a human, except for its more humanlike feet (which are also present in Pan before birth) & its more humanlike hands (which are also more primitive than the elongated bonobo hands). IMO, naledi should be called Australopithecus naledi. Most paleo-anthropologists still believe that the australopithecines are closer relatives of Homo than of Pan or Gorilla, but this is statistically impossible: of the 5 or 6 extant hominid (African hominoid) species, 1 (H.sapiens) has virtually all fossils (possibly >1000) and the 4 or 5 other species (chimps, bonobo, lowland & highland gorilla) have virtually no fossil relatives. Orangutans do have several fossil relatives, e.g. Sivapithecus, Lufengpithecus & Gigantopithecus, but for some mysterious reason, Pan & Gorilla are believed to have no fossil relatives... This pre-darwinian anthropocentrism. In fact, all so-called "human" or "bipedal" features (e.g. flat feet) in the australopiths are not derived-human (uniquely-human), but are primitive-hominid and lost in chimps & gorillas. Most likely, the early hominids were wetland waders (all australopith fossils lay in well-wooded & well-watered regions, e.g. K.Reed 1997), google "bonobo wading" illustrations. Pan & Gorilla evolved from wetter to drier forests, but Pleistocene Homo followed the coastal forests and began colonising the continental shelves when sea-levels dropped during the glacials: archaic Homo dispersed intercontinentally along African & Eurasian coasts & rivers, walking, wading & diving for waterside & shallow-aquatic foods (which are extremely rich in brain-specific nutrients), google e.g. "econiche Homo".