First Position and the Syntax/Prosody Interface: Old Irish Preverbs

In recent work a tentative suggestion has emerged: certain processes which appear to be core cases of head movement should be analysed instead as not involving the computational system at all, but rather as arising because of the way that the phonological systems interact with the structures delivered to the interface by the syntax (see, for example, Chomsky 1995, 4.7.3). In this paper, I examine a case of verb placement which suggests very strongly that at least some aspects of this proposal are on the right track, because the phenomenon in question eludes purely syntactic analysis without incurring unacceptable theoretical costs. I show that if proper attention is paid to the way that the syntactic representations delivered by the computational system interface with the system that is responsible for imposing prosodic structure, the analysis of this particular case of verb placement is much simplified. The phenomenon in question is the position of the verb and various preverbal particles in Old Irish (OI). The data are from the Eighth and Ninth centuries, mainly from explanatory glosses and comentaries on Latin texts. I am not aiming to place this discussion in a diachronic context. My aim is rather to show that we can place the surface positions of the various elements of the Old Irish verbal complex with some accuracy, but doing so brings up the question of how these elements got to these positions. Syntactic solutions to this question turn out to be not wholely satisfactory.