« Social Mobility: more data | Gene Expression Front Page | Culture vultures scavenging on biology »
April 27, 2005

Dragon's battles

The narrative high point of Dragon Bone Hill: An Ice-Age Saga of Homo Erectus is on page 15:


...One of his jobs was to obtain cadavers for the medical students' dissection. The Beijing police were only too happy to oblige and one day sent over to the anatomy department a number of headless corpses of executed criminals...Black visited the politic and explained that he needed intact bodies for the medical school...Some days later a line of shackeled prisoners arrived at Black's office from the police station with a note saying "kill them any way you like."

Don't blame this on Communism, this was the 1920s. Much of Noel Boaz and Russell Ciochan's book covers the history and controversy related to "Peking Man," but in the latter half of the book they propose two hypotheses that are of interest from a purely paleoanthropological perspective.

First, they suggest that the extremely thick skulls of H. erectus was an adaptive response to intraspecies competition. They analyze the pattern of depressions, possibly due to injuries, in ancient fossils and use analogies to modern peoples to make their case. In particular the Australian Aboriginals serve as a model, not only do they have the thickest crania among H. sapiens populations, they practiced one-to-one "ultimate fighting" and intertribal melee as a way to resolve disputes.1 From what I can recall H. erectus exhibited more sexual dimorphism than modern humans, so I can buy into the hypothesis that male-to-male physical competition played an important role in sexual selection.

Their second contention is less specific and more controversial: they want to rework the conventional paradigm battles between "Multiregionalism" and "Recent-Out-of-Africa." Traditional Multiregionalism posited worldwide genetic exchanges between homonid demes maintaining species cohesion (or at least populational interfertility). Additionally there would be periodic selective sweeps. All of this in concurrence with deep time continuity in local regional phenotypes. In other words, a significant portion of the ancestry of modern Chinese and Europeans might derive from Asiatic H. erectus and H. sapiens neandertalis, or at least one can discern a phenotypic continuity (one does not always imply the other). In the case of ancient Homo erectus in Asia and modern East Asians the "shovel shaped incisor" is the canonical example. Traditional Recent-Out-of-Africa usually implied total replacement and an almost blitzkrieg expansion out of an African ur-heimat through the rest of the world. This expansion was often explained in terms of not just cultural superiority, but biological capacities which other ancient homonid populations lacked. In the domain of paleoanthropology it seems Europe is the strongest case of discontinuity, as an expanding wave of "moderns" seems to have swept aside a culturally and phenotypically, and likely genotypically, distinct population of "Others." Over the past generation the paleoanthropological case, inferred from fossils, has been supplemented by an avalanche of genetic data, usually focused on the mitochondrial DNA and the Y chromosome (neither seem subject to recombination and are transmitted down an unbroken unisex line, at least to a very high order of aproximation).

So in comes "mitochondrial Eve," the African "Mother of us all." Here you have a scientific hypothesis which had such a cultural resonance that many rational considerations are swept aside. Twenty years ago when Rebecca Caan, Alan Wilson and Mark Stoneking argued for "mitochondrial Eve" in Nature they were greeted by a NOVA special and the cover of Time magazine. No doubt if they had argued for a "mitochondrial she-wolf" they would not have received the same response. Not only does the hypothesis offer a modernist interpretation of the ancient Biblical story (think of how the Big Bang Theory has also appealed to the philosophical and religious biases of Westerners-no matter that it is correct), it also serves as a way to argue that "we are all related" in a more secular humanist liberal paradigm.2 Nevermind that the most recent common ancestor of all human beings likely lived around 3,000 years ago, the archetype of the Mother of us All walking the plains of East Africa ~100,000-200,000 years ago has a certain romantic appeal. But there is a problem with this narrative: even in the most extreme Recent-Out-of-Africa replacement theories where one single band serves as the font of modern humanity, there is more than one female within that band. Mitochondrial Eve is actually simply the lineage of a the mitochondrial DNA, not the history of our whole species (sampling error, random genetic drift, implies that most lineages will become extinct-the same process caused problems in the perpetuation of unbroken patrilineages during the reign of Augustus in Imperial Rome). The data from the Y chromosome has been used to support the Recent-Out-of-Africa hypothesis as well, though it is far less clear from what I can gather. But, new evidence that other loci might have different histories that are difficult to shoehorn into the Recent-Out-of-Africa narrative, as well as a earlier analyses, suggest that no matter what the genetic history of the mitochondria might tell us, the story is not finished.

The authors of Dragon Bone Hill support what they term "Clinal Replacement," offering that they are drawing upon population genetical ideas which are traditionally outside the realms of both paleoanthropology and molecular genetics, the two disciplines that have framed the Multiregionalist vs. Recent-Out-of-Africanist debate. They make repeated references to Theodosius Dobzhansky, who was one who generally promoted and favored Sewall Wright's adaptive landscape metaphor, and I think the hand of that concept lurks in the background.3 The authors argue that though total genetic replacement is plausible in some regions, with the advancing population being clearly a distinct and distantly related species in Java and possibly Europe, in other cases it was more an instance of closely related homonid groups giving way to each other, so the genetic difference in any given locale might have been minor.4 Making an explicit reference to Alan Templeton, they seem to be positing an Out-of-Africa-Again-and-Again hypothesis, supplemented by their own palaeontological insights.

Personally, I think they are probably correct, or, they are more likely to be correct than the "pure" Multiregionalists and Recent-Out-of-Africanists. But, in less than 200 pages, with over half the space devoted to historical anecdote and reconstruction on the topic of Peking Man, they couldn't elucidate the genetical issues to any great level of detail. I think a key problem is that the further you go back in time the common sense relationship between genetic lineages, ancestral lines and breeding populations becomes difficult. Humans percieve themselves a a unitary whole with one ancestral past, but in reality we are a conclave of discrete genetic bits which might have disparate origins. This is why Recent-Out-of-Africa in the replacement mode is so appealing to the public and popularizers, it offers a neat resolution of the problem by fixing the idea that one small population, a primordial "band," was the ancestor of us all, that all our genes and ancestral lines goes back to this one breeding population. Multiregionalism goes to the other extreme, basically positing one world breeding population via a dynamic system of gene flow mediated by population-to-population migration and intermarriage. In the Multiregionalist narrative this one worldwide breeding population attains the same cognitive phenotype, sapiency, together (via a selective sweep event?). The idea that one genetic locus might not map on to the sum pattern of our historical ancestry or that the relationship between phenotype and ancestry is not always clear in the long term is difficult for many people, and this sort of conceptual sloppiness is what inevitably creeps into the Clinal Replacement or Out-of-Africa-Again-and-Again hypotheses. In other words, some of our genes might be fixed from a population which is distinct from the "main stem" of our ancestry that emerged out of Africa, perhaps some Eurasian Homo with which there was a mild level of hybridization. Or, perhaps the mitochondrial data itself is the result of a selective sweep event, perhaps due to adaptational factors which have been neglected by population geneticists.5 In this case, we are descendents predominantly of peoples who left Africa millions, not hundreds of thousands, of years in the past who simply happen to have received a favorable mitochondrial allele from Africa.6

We live in exciting times, though I am sure evolutionary biologists who grapple with human history have more mixed feelings.

Addendum: One point of great irritation for me is that part of Boaz and Ciochan's argument about pushing population genetics into the mix with molecular genetics and paleontology is that population genetics/molecular genetics in the context of human origins is an artificial dichotomy. Geneticists who are concerned with the mechanisms of gene regulation and the processes that mediate expression on the molecular level are not the drivers of the study of human origins (most of them are studying "model animals" that can elucidate fine grained molecular processes with greater clarity), rather, evolutionarily oriented thinkers who use molecular methods to make inferrences about phylogenetic relationships have been in the driver's seat from the get-go. Many population genetics questions are now explored via molecular assays and some models start from first principles on the DNA base-pair level. Not all population geneticists are involved in pedigree analysis of fly phenotype (though that goes on too) or sampling fish in isolated streams.

To give a specific example about the difficulty of the dichotomy offered is that the most prominent contemporary apostle of Recent-Out-of-Africa is Spencer Wells. Wells' mentor at Harvard was Richard Lewontin. Lewontin's mentor was the very Theodosius Dobzhanksy that Boaz and Ciochan point to as an exemplar of an "alternative tradition" to the scientists that brought us the molecular revolution in human origins. Population genetics and its mathematical models may offer hypotheses to test, but the tests will often be methodologically oriented to the molecular level. This is the age of synergy.

Related: The authors use some data from John Hawks in their argument. "Bloggers" make a difference!

1 - In Albion's Seed David Hackett Fisher documents similar "boxing" traditions among some of the Appalachian Scotch-Irish.

2 - The Multiregionalists returned the salvos in kind in the PC-war. In Race and Human Evolution Milford Wolpoff and Rachel Caspari basically end up concluding that Recent-Out-of-Africa is simply a narrative of the first genocide, while the Multiregional Hypothesis offers the vision of worldwide peaceful genetic exchange between demes and concomitant transitions toward sapiency.

3 - The beauty of metaphors is that they are easily contorted into the conformation you might fight congenial.

4 - I found the authors' presentation a bit muddled, they argue for example that H. erectus and H. heidelbergensis were not interfertile, that they were separate species, but the latter was also far more closely cognate with the former than H. sapiens neandertalis and H sapiens sapiens.

5 - There is some evidence, to give just one example, that certain mitochondrial types confer upon their carriers greater longevity because the genome does not degrade as fast. Mitochondrial degradation is the result of the fact that the nuclear DNA repair mechanisms are not operative in the organelle.

6 - Right now the balance of the evidence suggests to me that our ancestry is mostly African in the short term even if there is evidence of exogenous* genetic material from other homonid populations through hybridization.

* I'm not sure if I'm using exogenous correctly, but what other scientifically pretentious term would work?

Posted by razib at 05:49 PM