The prosody of Cayuga particles

Abstract

My thesis presents an analysis of the prosodic system governing the realization of particles in Cayuga (Northern Iroquoian). Traditionally words in Cayuga are divided into three categories: nouns, verbs, and particles. Nouns and verbs are typically longer than one syllable and are always accented; most particles are only one syllable and are not always accented in context. Particles can form prosodic groups with other particles or with nouns and verbs, and their accentuation depends on the words around them. My work is specifically concerned with identifying the principles behind the way particles are grouped and accented. Working within the framework of Prosodic Phonology (Nespor & Vogel 1986; Selkirk 1984, 2011) and Optimality Theory (Prince & Smolensky 1993/2004), specifically Match Theory (Selkirk 2009, 2011), I assume that syntactic and prosodic structure are required to be identical by Match constraints, but markedness constraints that require phonological well-formedness lead to mismatches between syntactic and prosodic structure. This results in, for example, a particle group that is comprised of several syntactic words, but is accented like one prosodic ‘word’. As the basis for my analysis, I mark prosody-related phonological processes (shortening, loss of final segments, pitch-accent assignment, etc) and use statistical analysis of acoustic data (pitch and duration) to form an objective description of the prosodic properties of particles. I then use objectively-determined criteria to determine which particle forms are prosodically strong or weak in context. I use the resulting transcription to demonstrate how the interplay of Match and markedness constraints can account for the peculiar prosodic behaviour of particles. My work adds to the linguistic literature on Cayuga and Cayuga phonology, as well as theories of the syntax-phonology interface. It also provides speakers and learners of Cayuga with a descriptive account of the prosodic behaviour of particles.