An interdisciplinary team of researchers analyzing ancient pots and burnt food found that people at Zhanqi ate millet and likely dairy, revealing that a centuries-old Chinese narrative of meat-loving "barbarians" was wrong.

Early Chinese writings about the people who lived in the Loess Plateau 3,350 to 2,650 years ago, in what is now northern China, described them as nomadic pastoralists who ate very little grain and had bad dining manners. The Zhanqi community is dominated by Siwa material culture and did not leave a written record; thus, until the publication of this multifaceted study in PLOS ONE on April 29, there has been little to contradict that characterization.

"There's always a political motivation behind writing these," said Karine Taché, an assistant professor of archaeology at Laval University and a co–lead author of the paper with Yitzchak Jaffe. "I would insist on the importance of … this tendency to try to give more voice to these marginalized people of history — those who had, maybe, not the dominant position, those who had no written documents. This is one important role that archaeology has, is to verify and correct, if possible, these perspectives, these false views of people who are not part of these [cultures]."

Their study involved use-wear analysis of ma'an jars recovered from Siwa graves; isotopic and molecular analysis of burnt food recovered from those jars; and, for good measure, an experimental cooking session with some millet and an obliging chef.

"The strength of this paper is that we're using a lot of different approaches to complement and inform us; we're not just singling out one," Jaffe, an archaeologist at the University of Haifa, told The Academic Times. "People would look at these ceramic vessels, and they would say, 'These are ugly, simple vessels that simple, backwards nomadic pastoralists would make.' And then we study them, and they're doing really interesting things with them."

The excavation in China was conducted by the Gansu Institute of Archaeology before a dam was constructed in the area. "This is a collaborative effort working with local Chinese archaeologists," Jaffe said. The excavation found 66 human graves, and this study examined 10 of the ma'an jars from a cemetery, nine of which were clearly from graves. 

Although the jars seemingly had some ceremonial function in death, the amount of residue and use marks on the jars clearly indicated that they had been repeatedly used to cook food prior to their burial. Patterns in soot, oxidation and internal carbonization and food remains revealed that the jars had been exposed to intense heat with little oil or liquid involved in the cooking — the food may have been roasted or baked, or placed next to a fire with one side exposed to direct heat. The patterns left by the cooking process indicated that the jars had been used in many different positions throughout their cookware lives.

"Because they're these nomadic pastoral groups, we're 'expecting' to find a kind of inferior material culture — what that means is that their ceramics are not going to be as sophisticated, they're not gonna do things as nicely as the sedentary developed civilized agriculturalists," Jaffe said. "The use-wear analysis, just looking at the vessels, shows that this was something a little more complicated than that."

Taché led the chemical analysis of the 10 pieces of charred food. In an isotopic analysis that shows how plants process carbon during photosynthesis, Taché initially found a marker that indicated either millet or an animal that ate millet, or a grass or sedge with the same marker, was cooked in the jar. But in a gas chromatography–mass spectrometry analysis of lipids, she found eight samples with a plant biomarker, miliacin, that could only mean one thing: Millet was in that jar. Taché sent a late-night, exclamation-point-heavy email to Jaffe.

She believes that she likely found dairy indicated in an isotopic analysis as well, which would be the easternmost evidence of early dairy consumption in what is now China.

In an unusual series of experiments at New York University, Jaffe and Taché got a large ceramic vessel similar to the ma'an jars and worked with chef Raymond Childs to cook millet and meat in a way that might produce familiar use patterns. Jaffe met the chef when Childs was working at NYU, and Childs developed a dish with the ingredients people in Zhanqi would have had.

"It was really good," Taché said. "The mix of meat and millet was really good." 

The study, "What do 'barbarians' eat? Integrating ceramic use-wear and residue analysis in the study of food and society at the margins of Bronze Age China," published April 29 in PLOS One, was authored by Karine Taché, Université Laval; Yitzchak Jaffe, University of Haifa; Oliver E. Craig, Alexandre Lucquin and Edward Standall, University of York; Jing Zhou, Gansu Institute of Archaeology; Hui Wang, Fudan University; Shengpeng Jiang, Oxford University; and Rowan K. Flad, Harvard University.