Chapter 14 Quantitative tests of implicational verb hierarchies

Abstract

This chapter will begin by discussing the implicational verb hierarchy of Tsunoda (1985) as a convenient starting point for looking at what happens when are latively large dataset and a principled, quantitative approach to their analysis are brought to bear on a linguistic typological hypothesis. After introducing new methods for assessing the validity of an implicational hierarchy, I go on to inquire into the presence of implicational hierarchies governing the distribution of 5 different alternation types across 87 verb meanings and 22 languages (Ainu, Balinese, Bezhta, Bora, Chintang, Eastern Armenian, Even, German, Hokkaido Japanese, Hoocąk, Icelandic, Italian, Ket, Mandarin Chinese [henceforth ‘Mandarin’], Mandinka, Mapudungun, Mitsukaido Japanese, Modern Standard Arabic [henceforth ‘Arabic’], Russian, Yaqui, Yucatec Maya, and Zenzontepec Chatino).1The data used are from the database of the Leipzig Valency Classes Project(Hartmann et al. 2013) in the state it was in as of July 17, 2012, although the names used to designate different alternations have been updated.