| ROA: | 121 |
|---|---|
| Title: | Prosodic Levels and Constraints in Banawa and Suruwaha |
| Authors: | Dan Everett |
| Comment: | |
| Length: | 65 |
| Abstract: | Prosodic Levels and Constraints in Banawa and Suruwaha ROA-121 65 pp. proslev.ps, --.rtf Daniel L. Everett University of Pittsburgh dever@verb.linguist.pitt.edu This article is a drastically revised version of an earlier paper that circulated via this list in 1995. It offers a detailed description of two languages of the little-studied Arawan language family of Amazonas, Brazil. It argues that a constraint-based approach to prosody in these languages is superior to a derivational approach, especially in its ability to handle edge effects that would be accounted for by Extrametricality in a rule-based analysis, as well as Word Minimality effects - words in these languages are minimally bimoraic. The paper argues that the relationship between prosodic levels is not as direct as phonological theory has previously assumed. Phonological models generally assume some notion of hierarchy (e.g. Pike 1967, Selkirk 1980, Ito and Mester 1995) in which a given level on the hierarchy is 'manifested by' units of the next level down. So, for example, words are manifested by feet and feet by syllables. This paper argues that feet are constructed on moras in Banawa, but that phonotactics (in which a family of Enhancement constraints is argued to play a crucial role), exceptional stress patterns, and a language- specific phenomenon of Name Truncation (shorten names to two syllables, as opposed to two moras) argue strongly for an active role for the syllable in Banawa phonology. However, foot boundaries may fall within syllables, violating a Syllable Integrity constraint. This raises problems for models which assume that foot boundaries can never fall within syllables. The paper also demonstrates the importance of a constraint of Word Binarity in Suruwaha - words like to be two well-formed feet in size. These languages also show that the Minimal Word cannot be derived from the Minimal Foot. This is because the languages have degenerate feet (which cannot be reanalyzed in terms of either Lengthening (Hayes 1995) or Catalexis (Kiparsky 1991). They also show that Foot Binarity must be broken down into Foot Minimality and Foot Maximality. Finally, they lead to refinements understanding of diphthongs, suggesting two new constraints Rising Diphthong and Falling Diphthong. Most importantly, perhaps, is the fact that this paper presents new data of potential interest to phonologists of all theoretical persuasions. :::::::::::::::::::::::::::::::::::::::::::: |
| Type: | Paper/tech report |
| Area/Keywords: | |
| Article: | Version 1 |