Although commentators have discussed myriad themes presented in the rich and extensive oeuvre of Childe, one of the topics that has been, in my view, seriously neglected is the topic of settlement types. In this article, I seek to make good this omission, starting from a consideration of The Danube in Prehistory. The basis of Childe's ideas on settlement types in the Neolithic and Copper Age of eastern Europe was a binary classification into ‘tells’ and ‘flat sites’ that, in turn, reflected a division between permanent and shifting cultivation and greater and lesser cultural complexity. However, the introduction into this debate of questions of trade, surplus production, and Neolithic ‘self-sufficiency’, as well as metallurgy and ritual, meant that the initial binary classification left a series of contradictions that Childe struggled to transcend in the last decade of his life.
