On these arguments, it would not be surprising if stone artefacts were found in contexts both in and outside Africa before 2 myr ago: first, because there is no reason why hominids were not in extensive areas of both continents before 3 myr ago (Dennell 1995); second, because they are likely to have been tool makers; and third, because primates other than hominids can make such artefacts. Finally, the imperfections and unevenness of the fossil hominid and early archaeological record need to be recognized more explicitly. Given the rate at which our views on early human evolution change in even the best known areas, new data from poorly known areas such as southern Asia are likely to be even less predictable. In particular, palaeoanthropologists cannot afford to be dogmatic about human origins when a wholly unexplored area the size of Europe, or the European Community and East Africa combined, lies between our two best data sets of Europe and East Africa. The most exciting developments in palaeoanthropology may yet occur in areas of former grasslands such as those in Africa outside the Rift Valley, or those of southwest Asia and the Siwaliks.