Feature Spreading 2.0: A Unified Theory of Assimilation
Peter Jurgec
June 2011

This thesis presents a universal and restrictive theory of assimilation as feature spreading. The main contribution is the idea that association lines between a feature and a root node are of different kinds, which is formalized in terms of maximally binary, headed and recursive feature domains. According to this approach, all targets can be associated with a spreading feature, but only some can be heads of a feature. When a particular target cannot be a head, it terminates spreading. Such patterns are attested. The representational differences between different kinds of segments are referred to by OT constraints. An advantage of the proposal is that it requires only six well-established constraint families, but their definitions are significantly modified. Feature spreading is driven solely by categorical alignment constraints that penalize triplets of a spreading feature, a targeted structure and a domain. Faithfulness, agreement, feature co-occurrence constraints, and constraints on heads all inhibit spreading rather than promote it. This framework allows for a unified analysis of vowel, nasal, consonant harmonies, local consonant assimilation, vowel-consonant interactions, and tonal spreading.
Format: [ pdf ]
Reference: lingbuzz/001281
(please use that when you cite this article)
Published in: PhD Thesis, University of Tromsø
keywords: assimilation, optimality theory, autosegmental phonology, vowel harmony, nasal harmony, consonant harmony, voicing assimilation, parasitic vowel harmony, rounding harmony, umlaut, icelandic, dissimilation, categorical alignment, positional faithfulness, agreement by correspondence, licensing, recursion, binarity, heads, prosody, hierarchy, phonology
Downloaded:2547 times


[ edit this article | back to article list ]