Case distinctions, rich verb agreement, and word order type (comments on Hawkins’ paper

Hawkins discusses a variety of asymmetries in the area of word order and proposes a variety of explanations in terms of parsing or sentence processing. I am sympathetic overall to the kinds of explanations he proposes (cf. Dryer 1980, 1992). In some instances, however, the data supporting the generalizations he puts forward is rather limited. Because I have a typological database (Dryer 1989, 1991, 1992) containing a large amount of relevant data and because of space limitations, I will concentrate my discussion on empirical evidence from my database bearing on two of Hawkins ’ empirical claims. The first of these is his claim that the presence of a case distinction between subjects and objects is most frequent in SOV languages, next most frequent in SVO languages, and least frequent is V-initial languages. The second is his claim that the presence of rich verb agreement (agreement with both arguments in a transitive clause) is most frequent in V-initial languages and least frequent in SOV languages. 1. Case distinctions between subject and object Hawkins predicts that SOV languages will distinguish subjects and object by case marking more often than SVO languages and that SVO languages will do so more often than V-initia