Prosody Drives the Syntax: O'odham Rhythm

In this paper, I examine two restrictions on the syntax of Tohono O'odham. I propose a reanalysis of these restrictions as following from prosodic considerations, rather than syntactic ones. When these restrictions are viewed as derived by the prosody, we unify the two cases of distributional facts. In contrast, a syntactic treatment of the restrictions shows them to be arbitrary and unrelated. Specifically, I will argue that the left edge of utterances in Tohono O'odham prefers an initial trochaic foot. To create this optimal prosodic structure, the prosody manipulates the syntax in two ways: 1) The g determiner does not appear initially in an utterance to prevent the sentence from beginning with a nontrochee, and 2) The auxiliary appears in second position to avoid beginning an utterance with a nontrochee. The analysis I propose will make use of the insights offered by Optimality Theory (McCarthy and Prince 1993; Prince and Smolensky 1993). I show that a theory with ranked, violable constraints, such as Optimality Theory, shows us how prosodic and syntactic constraints interact. The paper is organized as follows. First, I describe the syntactic restrictions. A brief overview of word-level stress in Tohono O'odham follows this section. The third section synthesizes the syntactic and phonological background with a look at the data from the prosodic perspective. I then provide an analysis for the data, after giving background on Optimality Theory. The paper concludes with a discussion of the further repercussions of the analysis presented here.