Arguing about species: Is it evidence, or ego?

For some people who follow human evolution news, recognizing “species” is really just about whether you’re a lumper or a splitter. Many people assume that the names of species are about ego, not evidence.

But nature presents us with real challenges, which still cause different scientists to approach the past with different assumptions. Let me give some examples.

Just today, I got notification of a new paper by Walter Neves and colleagues, in which they suggest that Australopithecus sediba and Homo naledi are actually South African representatives of Homo habilis. Some people might scoff at this—after all, the Dinaledi fossils are only 236,000–335,000 years old, while the latest-known H. habilis is around 1.6 million. But a young date for some fossils doesn’t bar them from from membership in a species with much older fossil representatives. Identity is tested with morphological evidence, not geological age.

Now, I disagree with the idea that H. naledi is the same species as H. habilis—Neves and colleagues have come to this taxonomic conclusion by neglecting all the morphological evidence showing H. naledi is different from H. habilis. But it’s not so easy to reject the idea that these species might be close relatives. As we pointed out earlier this year, H. naledi might even conceivably be a descendant of H. habilis or Au. sediba. On the other hand, Mana Dembo and colleagues showed last year that H. naledi seems to be closer to modern and archaic H. sapiens than to H. erectus (and much closer than H. habilis). These are stark differences of interpretation, from similar parts of the skeleton.

At the other extreme, this week Jeffrey Schwartz is set to present results of his own examination of the Dinaledi Chamber sample. According to his abstract, all the teeth belong to one species, but some of the skulls represent another—two species in this assemblage, not just one. I haven’t seen the details of this analysis, but I’m pretty sure I disagree with this one, too.

I admit that it would be fatuous to say that ego plays no role in paleoanthropology. Scientists express provocative opinions that will draw attention from the press.

Still, the trouble with taxonomy isn’t just about new fossil discoveries like H. naledi or Au. sediba. We have seen similarly broad and vociferous diversity of opinions in the last few years about H. erectus, Au. deyiremeda, Au. anamensis, H. floresiensis, Denisovans, Neanderthals, H. heidelbergensis, Kenyanthropus platyops, and others. These are species new and old, and the same issues keep arising again and again.

Many people would say that taxonomic debates just reflect basic philosophy about variation—again, lumping versus splitting. But that’s really only one of the dimensions:

  1. How much variation should a species include? Broadly, all of us recognize that some species are polytypic (as humans are today), but small and fragmentary samples make it very hard to distinguish polytypy from distinct species. There are living polytypic species that include very extensive variation, and living sister species that barely differ from each other, making model selection difficult.

  2. What kind of data provide evidence of similarity or difference? Some researchers rely mainly on phenetic similarity measures, using geometric morphometrics, principal components or canonical variates approaches. Others examine discrete (or threshold) traits, counting shared derived traits as evidence of similarity and ignoring shared primitive traits. This group was once dominated by cladists, but in recent years Bayesian approaches have become more and more common.

  3. What temporal or geological information is sufficient to justify pooling fossil specimens into a single sample? Some scientists are willing to assume that fossils from the same 500,000-year period belong to a single population, even if they preserve different parts of the skeleton or minimally overlap. Others draw trees that separate every specimen into its own “operational taxonomic unit”. The concept of a paleodeme is based on lumping specimens by date and geography, an approach that has come more and more into question as the fossil record increases.

Anthropologists may get a bad rap from other biologists for arguing about taxonomy so much, but in reality many areas of taxonomy are undergoing seismic shifts following more widespread application of genetics and phylogeographic analyses.

For example, bovid systematists have been arguing for the last few years about whether to double the number of species they recognize—a debate about living species with abundant morphological samples and genetic data. Meanwhile, living and fossil elephants are on the verge of a complete revision of relationships, based on ancient DNA and the appreciation of deep diversity between forest and savanna African populations. Similar examples are unfolding across mammalian systematics.

Neandertal and Denisovan DNA has shown us that hominins also exchanged genes by introgression, occasionally but recurrently despite hundreds of thousands of years on their own trajectories. Genetic evidence of African “ghost lineages” means other long-lasting Pleistocene populations once existed.

H. naledi might potentially be one such lineage. I have no idea what the closest relative of H. naledi will prove to be. Whether it reproduced with human populations or not, it shows that many cherished human features may not have been uniquely derived evolutionary developments.

I can’t help but feel that we are standing at a special moment in the history of paleoanthropology. New data give us the opportunity to make progress on old areas of disagreement about species and phylogeny. We have to start by taking what we now know about the later Pleistocene, and seriously appling these lessons to earlier periods of human evolution. Our assumptions about the past really are changing.

Whatever we choose to call species won’t change their nature. But our assumptions determine the way we frame our future studies, including our attempts to find more fossil evidence. That makes it important to communicate clearly about what these ancient species mean, both with each other and with the public.