My January 2016 Scientific American column, “Murder in the Cave: Did H. naledi behave more like H. homicidensis?”on the controversial hominin discovery Homo naledi, has generated much heat on online, with commentators (including that of John Hawkes, one of the principle investigators involved in the discovery) challenging my skepticism. Allow me to address a few of the points raised. The original column has been updated to reflect the one error regarding age and classification of fossils, but this essay provides a longer discussion of this and the other issues.

First, I have nothing against the publication of the papers in eLIFE instead of, say, Nature (where the original papers were submitted and rejected). I consider this part of the “democratization of science” that I recently suggested is one of the most important science news stories of our time (in this year’s Edge Question). And fitting this trend, I applaud the posting of all the data online for anyone to read and even download the 3D images of the fossils for analysis. 

Second, the authors’ decision to publish their findings much sooner than most paleoanthropological papers has led to even more of the process Karl Popper called “conjecture and refutation,” in the sense that the conjectures proposed in the papers have generated refutations by professional paleoanthropologists such as Tim White, who has also expressed skepticism of the authors’ claims. This is how normal science works, but if you fast-track the conjecture stage in a field known for glacially-slow evaluation of data (White spent 15 years analyzing the fossils of Ardipithecus ramidus, discovered in 1994 but not published until 2009), you can expect the refutation stage to be equivalently accelerated.

An excellent overview of the ongoing scientific debate by Nathan Lents, “Paleoanthropology Wars,” includes quotes from the major players in the drama and is well worth reading. White, for example, believes that the fossils represent “a small, primitive H. erectus” and describes the authors’ claim of a new species “an example of artificial species inflation in palaeoanthropology.” John Hawkes responded forcefully: “It’s just a poor match to H. erectus, so that the only way to make the H. naledi fossils fit within Homo erectus is to stretch that species beyond any other ever defined in the human lineage.” This is an example of normal science at work. Conjecture…Refutation…Counterconjecture... and so on.

Third, regarding my suggestion that “the authors are downplaying an all too common cause of death in our ancestors—homicide in the form of war, murder or sacrifice,” this too is a conjecture that the authors considered and rejected. My point in suggested it in the original column, accompanied as it was by several examples of violence from the fossil record, is that we don’t drop that hypothesis too swiftly in the additional analysis stage. The social sciences in general, and the anthropological sciences in particular, have a long history of ignoring signs of violence in fossils remains because of a prevailing worldview throughout the second half of the twentieth century that think of our species as a peaceful noble savage. This is why I referenced Lawrence Keeley’s War Before Civilization: The Myth of the Peaceful Savage, Steven LeBlanc’s Constant Battles: The Myth of the Peaceful, Noble Savage, and Steven Pinker’s The Better Angels of Our Nature.

In this regard I also recommend Pinker’s The Blank Slate for a history of the assumptions about human nature—and the flawed science that accompanied them—that lie behind the peaceful noble savage myth. I devote an entire chapter to this debate in my 2015 book The Moral Arc, which includes a lengthy discussion of the criticisms of Keeley, LeBlanc, and Pinker, and why the data overwhelmingly support the theory that human violence extends deep into our Paleolithic past. Thus, in my opinion the hypothesis of homicide/sacrifice for the H. naledi fossils is as good as the others also proposed by the authors (occupation, water transport, predators, and death trap), and still fits their preferred conjecture of “deliberate body disposal” (although in the media hype it was exaggerated to suggest that the finds might represent some sort of ritual burial, which is not what the authors were suggesting).

Finally, I was simply wrong in my statement that since the age of the fossils is underdetermined, “it is impossible to conclude where in the hominin lineage the fossils fit.” I had always assumed that age and morphology were both considered in the taxonomic classification of a species (since fossil taxonomy almost always includes fossil dates in the secondary literature), but according to the renowned paleontologist Dr. Donald Prothero, whom I queried on the matter, that assumption is incorrect. Here is his very interesting and useful explanation about the relationship between fossil age and taxonomy:

"This debate was held in the 1970s, and settled in every area except anthropology. Strictly speaking, the taxon must be defined only on morphological features alone, no matter what its age. If you use the age to define the taxon, then you're running into a circular argument (“It is species A because it is xx m.y. old, but the fossil of a different age looks the same, it is species B”). A famous paleontologist of the 1950s literally couldn't tell you what species his oreodont was unless you gave him the age first, which is a clear no-no. Otherwise, if your species is partially defined by its age, it cannot be used to date the rocks (circularity). Most biologists and paleontologists abide by this, although when they have large samples, they sort them by age first to compare variability between levels before deciding how many species there are.


Sadly, anthropology has always been on the trailing edge or fringe of these debates, often using concepts long rejected or considered outdated in the rest of systematics. They were still clinging to the "one species of hominid at a time" argument years after the fossils showed it couldn't be true, and lots of other questionable practices, since very few of them are trained in systematics of other organisms. This is another classic case of anthropologists being out of the loop in systematics. If H. naledi is based on complete enough material to establish that it is Homo, then the age is irrelevant. Homo could occur at any time if the specimens are good enough. But my recollection is that the skull and face are badly broken and partial, so it’s not so clear cut. If it does turn out to be older, lots of people will interpret the Homo-like features of this scrappy material as within variation of Australopithecus. That is the crux of the argument. As usual, the anthropologists are using outdated systematic concepts, and often they are arguing about questions that cannot be resolved by such poor material. If it were rhinos or any other fossil mammal, the paleontologist would put the poor material aside. But since it is us, even the poorest specimens are given names and over interpreted and argued about, no matter how much it can really be diagnostic material or not."

            —Dr. Donald R. Prothero, Dept. Geological Sciences, Cal Poly Pomona Research Associate in Vertebrate Paleontology, Natural History Museum of L.A. County


In conclusion, I remain both excited by and skeptical of the claims made about H. naledi, and find much wisdom in the words of the find’s principle investigator, Lee Berger, who told Nathan Lents: “You know, I don’t know where things will end up if we ever reach a new synthesis. We probably need more fossils to really do that, more skeletons especially.” In the always contentious field of paleoanthropology, more fossils always generates more conjectures…and refutations.