A Seismic Shift in Human Origins Research, or a Downward Slide?
A new research paper is out which has created a lot of media buzz. “Evolutionary History and Adaptation from High-Coverage Whole-Genome Sequences of Diverse African Hunter-Gatherers, “ by Joseph Lachance et al. reports “archaic admixture” in three African hunter-gathering populations (Pygmies, Hadza and Sandawe). On the basis of modern samples, geneticists suggest that an archaic species of hominins – different but related to Neandertals – interbred with African humans some time between 80,000 and 20,000 years ago.
This post is not about the technical results of this paper but about the politics of science around it. Applying anthropologists’ favorite tool of participant observation to web discussions of human origins papers is a growing theme on this blog.
Science journalist Nicholas Wade covered the research in the NYT under the provocative header “Genetic Data and Fossil Evidence Tell Differing Tales of Human Origins.” According to Wade, “claims of interbreeding have opened up a serious discordance between geneticists and paleoanthropologists.” All human fossil remains in Africa for the last 200,000 years are of anatomically modern humans, providing no support for a coexistent archaic species. He quoted the foremost paleoanthropologist Richard Klein as saying that geneticists are acting “irresponsibly” by publishing genetic findings about human origins without even trying to show how they may fit in with the existing fossil and archaeological evidence. Klein said that the new claim of archaic and modern human interbreeding
“is a further example of the tendency for geneticists to ignore fossil and archaeological evidence, perhaps because they think it can always be molded to fit the genetics after the fact.”
Klein’s reaction has sent waves of controversy across the blogosphere.
Razib woke up, read Wade’s article and filed a complaint on his widely-read but bias-ridden “science blog”:
“[Wade’s article] aggravated me. I believe that it’s totally misrepresenting the tensions in the scientific process here, and misleading the public. The standard conflict/”two views” format is used, and to disastrous effect.”
Razib believes that Klein’s position reflects his personal grievance against the challenges posed by new genetic, paleobiological and archaeological evidence to his theory of symbolic revolution in Africa that resulted in the founding migration of behaviorally modern humans from Africa to the rest of the globe. He also believes that other paleoanthropologists (such as the original Multiregionalists Milford Wolpoff and his student John Hawks as well as Erik Trinkaus) are fully onboard with the new genetic evidence. Consequently, Wade did a disservice to the public by confusing what is “really a within field controversy into one between two scientific teams.”
Reader Wesley Pegden retorted:
“Shouldn’t you have interviewed one of the paleontologists your talking about, so that you could include a quote in your article supporting what you’re saying? Otherwise it seems weird to be criticizing the journalism standards of the Times article.”
It’s a fair point: although Wade’s article was likely inaccurate in some details (such as the imputation of an idea that Neandertals lived in East Africa to geneticists), Razib’s response showed even less journalist integrity by belittling Klein (“ticked off”) and correcting Wade without conducting any formal interviews of his own but on the basis of “talking” to some unnamed scientists and knowing some of the anthropologists “in passing.”
A lively discussion ensued and we heard John Hawks flogging Wade for misinterpreting – in a typical media fashion – the true realities of what’s going on in academia. Hawks, of course, took the opportunity to congratulate himself on being one of the first “cranks” who anticipated the recent turn toward “archaic admixture”:
“The story of the overall transformation of human origins is such a good one, and Wade (among other journalists) totally misses it. I mean, we’ve come to the point where Sarah Tishkoff is publishing strong evidence for archaic gene flow and intermixture. A mere five years ago, the idea you’d find something like this in African genetics was the idea of a few lone cranks (like me). Now everybody’s on the bandwagon.”
Enter Dienekes. This amateur blogger patted himself on the head for “predicting” earlier this year that full genome sequencing would turn up “archaic admixture” in Africa. So, Hawks anticipated the discovery and Dienekes predicted it. But how robust is the discovery and how prepared are we to interpret it to warrant priding on anticipating and predicting it?
The problem is that this “archaic admixture” in Africa is a hypothesis built on genetic data and genetic data only. It hasn’t been – and likely will never be – confirmed by paleoanthropology. There are no clearly identifiable clusters of skulls in Africa between 200,000 and 100,000 YBP that can be conceived of as realistic proxies for two discrete but potentially interbreeding populations. As Klein’s student Timothy Weaver (hi, Tim) has recently argued, there’s no paleobiological evidence that something special (a population split or a bottleneck) happened in Africa between 200,000 and 100,000 YBP that set apart modern populations from the archaic ones. Geneticists detected a statistical pattern, as Klein said, but this is not a historical fact. Ancient DNA doesn’t survive in hot climates and, according to a paper co-author Joshua M. Akey, it’s unlikely to be detected in skulls because the 2.5% of the genome affected by this “archaic admixture” confers no selective advantage. So, it’s like Freud’s Oedipus complex: since it’s always repressed, you can’t hope to find any evidence for it. Consequently, you just have to believe in it.
Paleoanthropology and genetics are clearly at odds here (this doesn’t mean that, in addition to the cross-disciplinary contradiction, there is no dissent between schools within paleoanthropology and genetics), and Wade captured it correctly. It’s consistent with another recently transpired instance when genetic and paleobiological/archaeological data don’t match. Geneticists calculated that the exit from Africa took place around 60,000-50,000 years and proceeded along the coast to South Asia, Southwest Asia and the Sahul, while archaeologists firmly believe that there is no material evidence for this migration at this time. Archaeologists rather suggest an earlier exit at 120,000-100,000 years BP but there’s no evidence in genetics for such an early migration.
So, there are no archaic hominin vs. modern human skulls in Africa 200,000-100,000 YBP. African anatomically modern skulls can be seen as a gradual extension of earlier archaic variation in Africa from 400,000 YBP down, and not a clearly identifiable source for populations outside of Africa. And there is no archaeological evidence for a coastal migration of modern humans out of Africa at 50,000 YBP. On the other hand, on the basis of the data coming from living humans, modern humans admixed with archaic hominins in Africa between 80,000 and 20,000 YBP, while migrating out of Africa in the same window of time. This could be interpreted as an expansion by a new species within and outside of Africa (say, from a narrow source in Africa) but there is nothing in archaeology or paleobiology that can corroborate it. If, on the other hand, modern humans expanded 200,000-100,000 YBP, how come we don’t see genetic traces of earlier and deeper admixture levels in Africa and of an earlier migration out of Africa?
It’s not at all surprising that creationists such as David Klinghoffer are gloating over all of this. And Razib’s sucker punching him in a top-of-the-post update is a poor defense of the Titanic of a science of human origins that has indeed hit an iceberg. For many years, academic science has touted a consensus across genetics and paleobiology without establishing a clear standard of what an interdisciplinary synthesis in modern human origins research should be like and what is actually proven and what is just hypothesized on the basis of inherently fragmentary paleobiological data and inherently synchronic population genetic data. As the Denisova Cave discovery demonstrated, the true synthesis of genetics and paleobiology (as in DNA extracted from a 40,000 year-old pinkie and a tooth) is capable of unsettling the consensus between paleobiologists and geneticists and open up a rift that just keeps widening with every study.
My view is that there was a large-scale migration of behaviorally modern humans (our ancestors) into Africa at 50,000-40,000 YBP (the Hofmeyr skull in South Africa dated at 36,000 YBP clusters with Upper Paleolithic Eurasians), and what geneticists have mislabeled as “archaic admixture” in modern African foragers comes from the so-called “anatomically modern humans” in Africa. This is fully consistent with Klein’s claim that there are no archaic skulls in Africa between 200,000 and 100,000 YBP. And we don’t need them for genetics and paleobiology to align. All we need is to re-interpret anatomical modernity in Africa between 200,000 and 100,000 YBP as product of convergence. AMH were not behaviorally modern and when they tried to exit Africa around 100,000 YBP, they were replaced by Neandertals in the Levant. So, they are not our genetic and cultural ancestors, no matter how similar their skulls look to ours. But they could interbreed with us. They were the product of an overlooked indigenous evolution of African archaics into a species that didn’t survive. We see a similar phenomenon in lithic industries in Europe, Middle East and Asia: Mousterian technologies survive until recent times, they undergo independent evolution toward blade industries and they spill out of Africa but not beyond Arabia or India, and there are no anatomically or unambiguously behaviorally modern humans around until 45-40,000 YBP. Y-DNA supports a massive expansion of Eurasians into Africa because African hg E is a subset of the Eurasian CT clade. mtDNA phylogenies probably need adjustments to reflect a similar topology.
This is the way to make genetics and paleobiology tell the same tales. But for this synthesis to prevail, both disciplines need to abandoned two of their most sacred but thoroughly flawed claims, namely that African “anatomically modern humans” are the ancestors of living humans and that modern Africans are phylogenetically basal to modern non-Africans. It’s precisely on these two claims that the previous out-of-Africa consensus was built. Now it’s time to revise them.
There is indeed a seismic shift going on in human origins research: data derived from living human populations can drive theories about the ancient past that successfully compete with theories derived from tools and bones. (My out-of-America theory reflects this broader shift.) Students interested in human origins are jumping the paleobiological ship and becoming geneticists (as it was noted on Razib’s blog, “a google scholar search suggests that John Hawks is mostly doing genetics now”). But this shift is accompanied by the erosion of meaning behind the questions academic and popular science is asking and the answers it is providing. Let’s follow closely what Hawks has to say:
“I have always favored the hypothesis that most of the ancestry of today’s people traces to Africa before 100,000 years ago. And I have consistently been a vocal multiregionalist. I see no contradiction here.”
And then he suggests that the genetic evidence marshaled for out-of-Africa in fact admits other interpretations:
“The strongest argument for an out-of-Africa bottleneck was the low global effective size. That argument has now completely evaporated with the recognition that genetic variation today includes Neandertal and Denisovan variants. A secondary argument was that non-African variation for most loci was a subset of African variation. We now can observe that argument is generally false when considering large enough samples, and has no bearing on the question of origins. A tertiary argument was the lack of mtDNA introgression from Neandertals, which is clearly now irrelevant.”
And he even made a small contribution to establishing this fact:
“Since the 1990′s, it has been clear that this pattern does not require a recent replacement, or even a large-scale migration out of Africa, since it is consistent with larger long-term population size or greater population structure within Africa. There is quite a long literature on this topic beginning with Harpending, Relethford and Takahata, and to which I have made some small contribution.”
Hawks doesn’t seem to realize that he contradicts himself. More importantly, he doesn’t realize that, while we now have evidence of “archaic admixture” in living Africans just like we have evidence of “archaic admixture” outside of Africa, we don’t have genetic evidence that humans descended from an archaic African population. And he admits that we don’t have genetic evidence for an out-of-Africa migration. We don’t have evidence that modern humans actually exist. Hawks won the battle but lost a war. Dienekes celebrates this nonsense with declarations that humans are not “special” and that they “didn’t originate anywhere in particular.”
So, while our databases have grown and our infrastructure to share the data has expanded, our abilities to interpret the data have plummeted. Humans are the only surviving member of the genus Homo, we have a unique system of adaptation, most sophisticated technologies, the ever-growing population size, we live everywhere, including the North Pole and the outer space, we exert dramatic impact on our environment, we teach chimps how to talk, we bicker about our origins in a virtual environment, but science just cannot pin down where it comes from, how all of this started and when the seeds of these behaviors were first planted.
The whole anthropological dimension of human origins is being thrown by the wayside. At least Klein was asking the right question.
This bickering at the top is disappointing. To me, the whole situation is like a R.T.A. with different opinions on what caused the crash, made worse through lack of evidence.
Admittedly, finding bones is very convincing, when they can be found. The lack of the like is frustrating but has to be progressed. D.N.A on the other hand can be very accurate if the source is not polluted. So, more evidence from all parties without throwing all findings into disrepute should be the way foreward.