Instapundit found a BBC report today, but the story actually originated last Fall.
The London Times reported last September 23rd:
It was an unremarkable Roman cemetery, containing the bodies of ordinary people. They lived and died on the banks of the Thames, making a living in the poorer and dirtier districts of Roman Londinium.
When an analysis of the skeletons came through, no one expected a result that could change our view of the history of Europe and Asia. But that is what they seem to have found, because two of the skeletons, dated to between the 2nd and 4th century AD, were Chinese.
Here at the most westerly point of the known world, in the cultural backwater of ancient Britain, lived people who came from its easternmost extremity. How did they get there?
To the Romans, the Chinese were a mysterious civilisation: technologically advanced, disquietingly powerful, and purveyors of, according to Seneca, obscene garments that corrupted the empireâ€™s womanhood. To the Chinese, the Romans were a moderately intriguing civilisation with, according to one account, weak and pliant rulers.
It would be 1,000 years before the travels of Marco Polo would help properly to bring the culture of the east to the people of the west. Now, though, that history has to be revised.
After the excavation of a cemetery in Southwark, new skull analysis techniques identified a multicultural community containing four people who were ethnically African and two Asian, probably Chinese.
The find is spectacular but it is also mysterious, according to Rebecca Redfern from the Museum of London. She has no idea how they had ended up lying in this cemetery, so far from home.
Several British news outlets today ran a story with headlines about Chinese people in Roman Britain. While there is no doubt that the Roman Empire was cosmopolitan, and it is entirely likely that people of East Asian ancestry will be found in all parts of the Empire, we need to take a step back from the hype and look at the data.
The new study in question is by Rebecca Redfern and colleagues, out in the October issue of the Journal of Archaeological Science. The researchers looked at 22 skeletons from the Lant Street cemetery in the London borough of Southwark, dating to the 2nd-4th century AD. In order to figure out where people might be from, they examined oxygen isotopes from the teeth, carbon and nitrogen isotopes from the bones, and the shape of the skull, correlating those data where possible with burial evidence.
The data that Redfern and colleagues produced are really quite interesting. The oxygen isotope values, which were isolated from 19 of these people, range widely — too much to be explained by local variation in water sources. This means that many of them came to London from elsewhere, some time after childhood. Far fewer of the individuals produced good data for carbon and nitrogen analysis of diet — just half of the sample was testable, and those data reveal a diet similar to what was eaten in selected other parts of the Empire. (They did not compare the data to Rome itself, for example, only to Portus Romae, south-coastal Velia, and Herculaneum in Italy and to Leptiminus in Tunisia.)
But the new method that Redfern and colleagues use to figure out ancestry is not ancient DNA analysis, but a statistical modeling of variations in the skulls and teeth that could be linked to ancestral differences. In short, they employ a method similar to what forensic anthropologists use to figure out if an unknown skeleton is of Asian, African or European ancestry.
The shortcomings of this method, however, are considerable and are outlined by Redfern and colleagues in their article. For example:
The fact that many of the samples were fragmented means that 41% of the sample had only two traits to score. As the researchers write, “This degree of missing data can affect classification accuracies, particularly among the sample having two or less (sic) traits.”
“We recognise that this is a subjective approach… and that many of the individuals used to generate these methods derive from modern populations outside of the territories that formed the Roman Empire. […] The population affiliation divisions used here may disguise or fail to find many affiliations because they are subjective, and morphology varies between individuals and over time,” they further note. This is problematic because bioarchaeologists cannot be sure how much the skull and tooth shapes have changed over 2,000 years. Comparing an ancient population with a modern one may not yield accurate results. (For example, when I put metric data from skeletons from Rome into FORDISC, a software program that compares metric data from skulls, the program happily classifies them into Asian samples.)
“The method development was particularly lacking in north African and southern Mediterranean populations, whose DNA shows a greater degree of genetic diversity compared to sub-Saharan and more northern ones. Therefore, the results must be understood in their temporal and spatial context, and the biases introduced by the methods acknowledged.” With few comparative samples from contemporary Africa and the southern Mediterranean, which are much more likely to be the origin of Roman Britons than is Asia, this means there may be bias introduced into the interpretation of the skull and tooth shapes.
This article is a remarkable attempt to correlate three different isotopes and skeletal morphology to answer questions about the diversity of Roman Britain in the later Empire, and it succeeds in showcasing that diversity even in this small sample. But it does not show, as the tabloids have been crowing, that there were Chinese in Roman London. The statistical results are intriguing, but the oxygen data from the two so-called Asians seem to be within the range of others in the sample, and only one produced dietary isotope data. For a slam-dunk, they need DNA. If and when they produce this, though, establishing a solid correlation between DNA from the Roman era and the results of the statistical method on the Roman skulls and teeth has the potential to help other bioarchaeologists assess ancestry without doing expensive destructive analysis.