The Mon Tasiek dialect of Acehnese (Indonesia) was analysed by Lawler 1977 as having a passive in which the verb agrees with the initial, or underlying, subject. His argument was that the preverbal NP in Acehnese has the expected properties of a surface subject, but that verb agreement was controlled by the underlying subject—which, in ‘passive’ clauses, is not the preverbal NP. This analysis was based on limited and unrepresentative data. In the light of more extensive data, Lawler's claim that the preverbal NP in Acehnese acts as a final subject does not hold. The most obvious—and, for syntactic theory, the least problematic conclusion—is that the so-called ‘underlying’ subject is in fact a surface grammatical relation; Acehnese has no passive, and its verbal agreement is a surface phenomenon.

The paper of Durie 1988 raises interesting and useful questions about some work I did twelve years ago. Before I consider the substantive issues involved, however, some background is in order about the genesis of that work and the questions it raised.

This is as comprehensive a study of Italian syntax as I've seen in modern linguistics: it deals with verb classes, cliticization, impersonal si sentences, causatives, restructuring sentences (i.e. those which allow the phenomenon commonly known as ‘Clitic Climbing‘), perception verb complements, and the analysis of reflexive clitics, all in great detail. It also deals with innumerable other matters, sometimes rapidly, but always intelligently—including impersonal passives, there sentences, criticisms of VP analyses which allow a variety of projections of V, reconstruction in LF, theta role assignment, and Case assignment. The data, despite the title, are drawn from English as well as Italian, with interesting sections on Piemontese and French.