| Summary: |
This thesis presents a new approach to the behavior of coronal segments. It examines seven aspects of coronal uniqueness; (i) the interaction of coronal consonants with front vowels, (ii) the confinement of liquids to coronal Place, (iii) the preference of "weak" syllabic sites for coronal Place, the processes of (iv) palatalization and (v) coronalization of coronal and non-coronal consonants by palatal glides and front vowels, (vi) the confinement of consonant harmony processes to consonants of coronal Place, and finally, (vii) the frequency and subplace richness of the coronal Place. It is argued that this range of behavior can be given a unified analysis if coronality is represented by the Government Phonological element [I]. Further, this element is argued to be the head of a Resonance Phrase in an element-geometric tree which is divided into a Resonance, Manner and Laryngeal Phrase. The headship of [I] gives this element greater powers to license other (Place, Manner and Laryngeal) elements, so deriving the behavior noted. This is contrasted with approaches which underspecify coronal Place, or try to capture coronal anomalies by recourse to phonetic context. The headship or dependency of elements drives element combinations, and thus derives the structure of phonemic inventories. This is traced to functional underpinnings, drawing on phonetic theories which argue for the optimality of segments based on the acoustically integrative effects of the articulations by which they are executed. The interpretation of [I] is thus investigated in some detail. At the level of segment generation, therefore, it is argued that there are formal and functional constraints operating. Finally, the distribution of coronal segments in the word is looked at in a broad range of typologically diverse languages. This is modeled using the above tools, in conjunction with a Government Phonology approach to syllabic structure and licensing. |
