Their argument is based on an analysis of charred grains and plant phytoliths (silicon articles that have a unique form specific to each plant) from numerous sites excavated in China over the last decade: the recovery of plant remains through soil flotation is now routine and has resulted in a critical mass of data from which it is possible to challenge the valley-based model of agricultural beginnings that has long dominated archaeological literature.
None of the sites from which foxtail millet, broomcorn millet and rice have been recovered in quantity lies in the valleys of such major rivers as the Changjiang (Yangtze) and the Huanghe (Yellow) rivers. Instead, they are located along tributary rivers, on the flanks of the hills oriented north to south between the floodplain and rising ground. Typically these are areas with a more complex ecology than the valley bottom.
The authors also note that there is a marked geographical interface between the two types of millet, a pattern that is consistent with the modern distributions of broomcorn and foxtail farms in China, with foxtail cultivation focussed in the wetter south Hebei and north Henan provinces, and broomcorn in the hillier and drier areas of Inner Mongolia and Shandong province. Further study is needed to work out what combination of geological, climatic and biological factors might account for the difference — but answering such questions offers fruitful scope for identifying the location pattern and phasing of early farming sites. •
This article is an extract from the full article published in World Archaeology Issue 35. Click here to subscribe