Carol Genetti

The tapestry of Dolakha Newar: Chaining, embedding, and the complexity of sentences

Abstract In Dolakha Newar the boundaries of syntactic sentences are clearly demarcated. However, each tightly-bounded sentence has the potential for infinite expansion and structural complexity. This results from the recursive interaction of two basic combinatorial structures: chaining and embedding. While these structures are basic to many of the worlds languages, in Dolakha Newar speakers combine them freely and frequently to spontaneously create sentences of remarkable intricacy. Additional structural nuance is provided by the chaining of constituents at levels below the clause and the sharing of core arguments. The result is a syntactic fabric of depth and complexity.

Segmental alternations in the Sunwari verb stem: a case for the feature [front]

In the Sunwari verb stem there are a number of segmental alternations that specifically involve the interaction of front vowels and coronal consonants. This paper discusses two possible analyses of these alternations, one using a traditional feature inventory in which the front/back distinction is marked by the feature [back] positioned under the dorsal node, and one using an inventory that replaces [back] with the coronal feature [front]. It is shown that the analysis with the traditional feature inventory fails in a number of respects, most conspicuously in requiring a simultaneous delinking of features from two unrelated branches of the hierarchy in order to properly derive high front vowels from coronal stops and nasals. The alternative analysis with [front] is by contrast simple and insightful. Use of the feature [front] also provides for a perspicuous characterization of constraints that hold throughout the phonology of Sunwari. In the recent phonological literature there has been a considerable amount of interest in the theory of distinctive features, both in terms of how features are organized and in terms of the nature and degree of feature specifications. Clements (1985) proposes that all distinctive features should be represented on hierarchically organized autosegmental tiers. The hierarchical organization of the tiers classifies features in recognition of the fact that certain subsets of features exhibit similar phonological behavior. The grouping of similar features under a single node in the hierarchy allows phonological rules to apply to all features dependent on that node simultaneously, rather than to an otherwise arbitrary set of unordered features. The details of the hierarchical structure, such as how many nodes are needed and which features belong under which node, has been a topic of considerable discussion (Sagey 1986; Levin 1987; McCarthy 1988; Avery Linguistics 30 (1992), 319-35J8 0024-3949/92/0030-0319

From postposition to subordinator in Newari

2.00