Licensing by the Initial Foot

In this paper I present a range of evidence to support the claim that a prosodic constituent — such as a syllable or foot — can serve as the licenser of marked phonological or morphological structure (cf. Zoll 1996, Beckman 1998). In particular, such a constituent can be strong, and therefore a valid licenser, by virtue of either its inherent prominence, such as stress, or its positional prominence, such as peripheral position. The case I focus on here is the licensing of marked structures by a foot which is prominent due to its location in word-initial position, independent of whether it correlates with phonetic stress. In the main body of the paper I show that a variety of non-metrical processes in Kashaya are sensitive to the location of the first foot of the word. That is, the foot structure which determines stress and iambic lengthening — and which is dependent on whether or not initial-syllable extrametricality is present — is also crucial for the proper analysis of phenomena which have no inherent relation to feet. The initial foot licenses three different marked structures. These are the outcome of vowels in hiatus, the choice of suffixal allomorphs, and the distribution of the glottal segments called laryngeal increments. Since the first foot does not always bear stress, the licenser cannot be reduced to the category " stressed foot ". Adopting the framework of Zoll (1996), I propose an analysis within Optimality Theory which generalizes alignment theory to the notion of " coincidence " of some element with another — in this case, the first foot of the word. The function Edgemost(P, Q), which here picks out the leftmost foot (P) of the word (Q), crucially defines the prosodic category which licenses the marked structure. I begin in §1 by discussing three cases from the literature where certain featural constrasts are restricted to prominent, especially initial, positions. In §2 I present data from Nootka which shows that the initial (but perhaps only) foot in the word determines the distribution of long vowels and also restricts the realization of glottal stop. Starting in §3 I turn to Kashaya, first sketching the pattern of both stress and foot-based vowel lengthening, and initial-syllable extra-metricality which displaces the initial foot. I then turn to the non-metrical phenomena in Kashaya that depend on the location of that foot. In §4 I show that whether elision of two adjacent …