This article was read at the Neurolinguistics Seminar organised by Professor M. Paradis at the Canadian Linguistic Institute, McGill University, Montreal, Summer 1977. I would like to take this opportunity to thank Prof. Paradis and Professor С Douglas Ellis and other members of the Linguistics Department, McGill University for their hospitality and many fruitful discussions on various topics during my two months stay on the McGill campus. I would like to thank Lyn Doole, member of the Institute, for reading the early draft of the present paper. (1) Our symbols N, N* and R correspond generally to the S, 0 and V used . in traditional schematic presentation of word orders . For more information about our hypothesis Cf. YAU, monograph 1977a : "Constraints on basic sign order and word order universals".