Out of Asia / A New Framework for Dating the Spread of Agriculture in Europe

The switch from hunting and plant collecting to agriculture is one of the most significant economic and social changes in human history. Agriculture was introduced to Europe from west Asia 10,000 years ago and, during the next 3000 years, spread through Europe by two principal routes, one following the Danube and Rhine valleys through central Europe and into northern Europe (including Britain), and the second taking a coastal route through Italy and Spain to northwest Europe. The extent to which agriculture spread because of human migration, or through the transmission of ideas between existing hunter-gatherer groups, is a debate which can potentially be resolved by comparing the rate of spread in different geographic areas with variation in environmental and cultural factors. A widely held view of agricultural dispersal is that it spread rapidly into Greece and the Balkans, but slowed when it reached central Hungary before again moving rapidly through central Europe. Further delays in the adoption of agriculture have been suggested for the north European plain and Alpine foreland. On the other hand, the spread along the Mediterranean coast was apparently rapid throughout. Recently, however, doubt has been cast on the supposed delays in agricultural spread, which has left researchers debating the causes for variation in the rate of spread while the existence of delays is itself debated. Evidence for the rate of spread is based on radiocarbon dating but most of the dates are based on wood charcoal collected from sites classified as 'Neolithic' on the basis of objects found during archaeological excavation, rather than on cultivated cereals which are the products of agriculture. This project will directly date the earliest cereals (wheat and barley) found on Neolithic sites across Europe. The statistical tools used to analyse radiocarbon dates are also underdeveloped and assume that they provide a single date or 'time-stamp'. However, since radiocarbon dates are in fact an estimate of the actual date, they can be used as time-stamps only when the uncertainty on the date estimates is ignored, which is inappropriate. Another source of uncertainty arises due to fluctuations in the amount of radioactive carbon in the earth's atmosphere: dates based on radiocarbon do not exactly match actual calendar age. When radiocarbon dates are 'translated' into actual dates there may be a choice of more than one highly plausible date. We will develop a statistical model which takes account of the uncertainty of radiocarbon dates when mapping the spread of cultivated cereals across Europe. The new chronological framework, incorporating a large number of radiocarbon dates (both old and new) for cereals, will provide a powerful tool for testing hypotheses concerning agricultural spread.