3 - Implicational Universals in Greenberg's Data and the Expanded Sample

Abstract

This chapter presents a set of reformulated implicational universals for much of Joseph Greenberg's word order data. It presents a discussion of the general properties of adequate implicational universals of word order. The implicational universals to be proposed in the chapter have the following properties: they are almost all nonstatistical, that is, exceptionless rather than statistical, relative to the data base; they are all unilateral rather than bilateral; and a significant number are multitermed rather than bitermed. Unilateral implications define three-way typologies, whereas bilateral implications define two-way typologies. And multitermed statements are on many occasions able to capture regularities of co-occurrence that bitermed statements cannot, by, in effect, raising the antecedent of the implication from one property to two. The co-occurrence of a third can then often be predicted in a straightforward manner