By Helen De Cruz
Ancient DNA (aDNA), extracted from deceased organisms, has recently let to exciting discoveries. In the field of paleoanthropology, there was a long-standing debate on whether or not anatomically modern humans and Neanderthals interbred. Thanks to aDNA, we know there is some introgression of Neanderthal DNA in the genepool of humans outside of Africa (Asians, Europeans, Oceanians) - as is shown on the figure. This seems to show that humans indeed interbred with Neanderthals to a limited extent.
Ancient DNA also led to the discovery of a previously unknown hominin, based on genes sequenced from a tooth and a pinky bone of a girl found in Denisova cave (Siberia), dated to about 40,000 years ago. The teeth and pinky bone were quite robust, outside of human variation. They were not from Neanderthals (as researchers expected), but from a totally unknown new species of hominin (provisionally termed "Denisovan"), which is more closely related to Neanderthals than to Homo sapiens. It turns out people in Oceania, and to a lesser extent in South Asia, have some Denisovan DNA.
Sequencing the mitochondrial aDNA of a femur from a hominin from the Spanish site of Sima de los Huesos of about 400,000 years old, researchers were surprised to learn that these hominins were not a link between Homo heidelbergensis and Neanderthals (Homo heidelbergensis is the supposed common ancestor of humans and Neanderthals, a subpopulation of which is hypothesized to have left Africa around 500,000 years ago to Asia and Europe, giving rise to the Neanderthals, and evolving locally in Africa into Homo sapiens). Instead, the Sima de los Huesos hominin is more closely related to the Denisovans. This discovery overturned previous ideas about human evolution in Europe, "baffling experts", as the Nature News commentary piece puts it. The trust in aDNA is high. The archaeologist Clive Finlayson, renowned Neanderthal expert, calls the finding "sobering and refreshing", and adds "The genetics, to me, don’t lie".
Should we welcome aDNA as the source of evidence that can settle disputes in paleoanthropology, even able to resurrect species of which we have, for all we know, no physical remains except for a pinky bone and two teeth? Or should we be more cautious? As Ann Horsburgh notes, "there has been a reversal in the place of genetic data in that it is now privileged over other sources of data. This kind of molecular chauvinism leads to overreach in interpretation and is no less likely to hamper our progress. Moving forward we would do best be judicial in the use of genetic data alongside other independent archaeological evidence in reconstructing the past."
To decide about the evidential value of aDNA in our total pool of evidence (which also contains fossils, climatological data derived from pollen and deep-sea core drillings, traces of human activity in tools, hearths and other archaeological finds), it's important to note that recovering and sequencing aDNA is no picnic. It involves recent and unstandardized methods of extracting and amplifying genetic materials. It should also cause some concern that most aDNA studies come from one lab, the Department of Evolutionary Genetics at Max Planck led by Svante Pääbo. Although it is beyond dispute that Pääbo and colleagues are very competent and scrupulous, the lack of standardized methodology and the early stages of this field would make it seem better to have many labs involved, given the benefits of epistemological pluralism.
Less than a decade ago, the idea that we'd be able to make full genome sequences of extinct species such as Neanderthals seemed pure science fiction. The sequencing methods at that time were expensive, time-consuming, and required large amounts of fossil material. Now, the procedure is less expensive and more fine-grained. Still, ancient DNA extraction is tricky. Once an animal dies, the DNA disintegrates rapidly, and even under favorable conditions, typically only a few percent can be recovered (see figure which shows percentage of DNA that is estimated to have been recovered in different specimens). There's an age limit although it's unsure what it might be - in spite of early excited reports of "antediluvian DNA", we can't extract DNA of dinosaurs and other specimens that are millions of years old. The oldest DNA was extracted from a horse-like animal, about 700,000 years old. DNA preserves best in cold and dry conditions. Homo floresiensis has not yielded any DNA, in spite of repeated attempts and a relatively recent date (about 18,000 years ago), because of the hot and humid conditions of Liang Bua cave. Even under ideal circumstances, it inevitably gets contaminated with DNA of living organisms (such as archaeologists), and it's a laborious and non-standard process to get rid of the contamination.
But if we for a moment bracket the nuts and bolts (and as a philosopher I am unfortunately not in a position to evaluate just how large the problems of contamination and degradation are, and how much they influence the interpretation of the results), is there any reason to privilege aDNA over other data that paleoanthropologists have at their disposal? Perhaps they do, and privileging such data isn't molecular chauvinism.
According to philosopher of science Carol Cleland, historical scientists, such as evolutionary biologists, paleoanthropologists etc, differ from other scientists in their formulation of hypotheses about token-events, rather than type-events. For example, paleoanthropologists tend to be interested in specific, unique events in human prehistory, such as when humans first colonized the Americas, when hominins first left Africa, etc. To test these hypotheses, historical scientists are on the lookout for what Cleland terms "smoking guns", "a trace(s) that unambiguously discriminates one hypothesis from among a set of currently available hypotheses as providing “the best explanation” of the traces thus far observed." Smoking guns do not falsify alternative hypotheses, it "(so-to-speak) cinches the case for a particular causal story". For example, there were many hypotheses to explain the mass extinction that occurred at the K-T boundary, wiping out the dinosaurs around 65 million years ago, but the discovery of an even layer of iridium around this age around the world (an element found in abundance in asteroids) privileged the hypothesis that dinosaurs became extinct as a result of a massive asteroid impact (see evocative picture).
Notice though, that the iridium was one element that, in combination with several others, cohered well into the asteroid hypothesis. In the current reconsideration of hominin evolution through aDNA, we have not reached this stage. For one thing, the data do not fit the dominant models. In the case of Neanderthal and Denisovan admixture, the leading hypotheses were out of Africa (humans evolved in Africa, around 200k years ago, and spread to other parts of the world without admixture with other hominins) and the multiregional hypothesis, which posited ancient hominin Homo erectus populations that had regular gene flow, evolving into the populations we see today. The current data suggest the latter did not happen, but the gene flow is not predicted under out of Africa. It would seem paleoanthropologists have to make a new theory, quite ad hoc, for each aDNA find they've got. Maybe this is not so problematic as it seems, as paleoanthropologists have been doing this for more traditional findings as well. And there is something recommendable to revising one's theories in the light of compelling evidence.
Maybe we should privilege aDNA for particular kinds of information we are looking for, where we can expect aDNA to be a better source than other traces. For example, aDNA and mitochondrial and nuclear DNA of extant people are the only traces we have of evolutionary relationships. Although skeletons also provide some indirect evidence of family relationship, it is rough grained. Darwin and Huxley knew on the basis of comparative anatomy that humans were more closely related to apes than to other animals, but based on the skeletal material alone they could not say whether the gorilla or chimpanzee was the most close relative. Thanks to molecular evidence, starting with Sarich & Wilson's 1967 paper which looked at similarities in immunological reactions of blood serums, we know chimpanzees are our closest relatives, and this finding has been confirmed again and again, most recently also with aDNA. I think that the close evolutionary relationship between humans and chimpanzees has attained fact-status, thanks to molecular analysis.
However, for other kinds of data, DNA and aDNA provide just one line of evidence, and other findings provide equally compelling, perhaps even better data. Take, for instance, the dating of evolutionary divergences between hominids and hominids lineages. Many of these datings are calibrated on the human-chimpanzee divergence, and there is no agreement on when this happened (somewhere between 4.5 and 5.5 million years ago). Some very early hominin-looking fossils in Africa, such as Orrorin tugenensis (about 6 million years ago) predate the molecular split between humans and chimpanzees, which might mean they aren't hominins that led to us, or that the calibration of the human-chimpanzee split is mistaken.
There are lots of things aDNA is relatively silent on. For example, Neanderthals and Denisovans share with humans mutations on the FOXP2 gene, which are implicated in the development of speech. This would suggest Neanderthals and Denisovans could speak. But there is a nucleotide variation present in most humans, and absent in Neanderthals and Denisovans, suggesting further selection took place. But what does this mean in practice? Did Neanderthals grunt or perhaps sing (as Steven Mithen suggested)? We have no idea. Archaeology does not place us in a better position, given that Neanderthals seemed to possess art and symbolic culture, but this does not give us direct evidence about how complex that language was, and how it differed from the languages spoken by Homo sapiens populations today. To give another example, claiming that Neanderthals were patrilocal based on the genomes of just six individuals seems a bit premature, to say the least.
In sum, given the paucity of data available to paleoanthropologists, aDNA is a welcome addition, and in some cases, I have argued it can function as a smoking gun to decide between competing hypothesis. Still, given that it is a recent technique concentrated mainly in one lab, caution is recommended especially for information that goes well beyond what genetic data currently provide us with.