225 THE PROTO-SAPIENS PROHIBITIVE/NEGATIVE PARTICLE *MA PIERRE J. BANCEL, ALAIN MATTHEY DE L’ETANG ASSOCIATION FOR THE STUDY OF LANGUAGE IN PREHISTORY (CAMBRIDGE, MA); ASSOCIATION D’ETUDES LINGUISTIQUES ET ANTHROPOLOGIQUES PREHISTORIQUES (PARIS, FRANCE); AND JOHN D. BENGTSON ASSOCIATION FOR THE STUDY OF LANGUAGE IN PREHISTORY (CAMBRIDGE, MA); EVOLUTION OF HUMAN LANGUAGE PROJECT (SANTA FE INSTITUTE, NM) We report here on a lexical root, very widespread in diverse languages worldwide, includ- ing more than 50 ancient languages, long-isolated languages, and proto-languages. Most of these rely on uncontroversial reconstructions, while others, from Proto-Nilo-Saharan to Proto-Trans-New Guinea through Proto-Austric and Proto-Amerind, go back to far more than 10,000 years ago and cover all continents. We argue that this lexical root may only have been part of the ancestral language common to all modern humans. 1. INTRODUCTION We will document here an ancestral word root, which is found in such a huge number of language families across all continents that it can only be a common inheritance from the original lexicon of our remote Sapiens ancestors. Following the common linguistic custom of naming the ancestral language of a family by the name of this family with the prefix Proto- (Proto-Germanic, Proto- Algonquian, Proto-Bantu, etc.), we call the ancestral language of our species Proto-Sapiens. Proto-Sapiens is not a newcomer in historical linguistics: building upon the pioneering work of Trombetti (1905), about three dozen Proto-Sapiens words have recently been identified (Bengtson & Ruhlen 1994), making use of the massive linguistic materials and comparative works that have accumulated during the 20 th century. However, many historical linguists deny the validity of Proto-Sapiens etymologies, a subject which deserves a brief preliminary discussion. Their rejection basically results from an orthodoxy which has held for more than a century that languages evolve so fast that, after 5,000 to 8,000 years of evolution, nothing significant remains of an ancestor language in its descendants. This orthodoxy is easily demonstrated to be false. We can illustrate this point with an example taken from the Indo-European family, to which most modern European languages belong and which is, for this (unscientific) reason, by far the best studied of all language families. Its ancestor language, Proto-Indo-European, is estimated to have been spoken some 5,000 to 8,000 years ago. This is close to the limit beyond which any trace of it should have vanished in modern languages;