Issues in round harmony: Grounding, identity and their interaction.
Committee ChairArchangeli, Diana B.
MetadataShow full item record
PublisherThe University of Arizona.
RightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
AbstractThis dissertation examines the cross-linguistic patterns of "height-constrained" Round Harmony (RH), from the perspective of Grounding Condition Theory proposed by Archangeli and Pulleyblank (in press). The fundamental claim advanced here is that the attested patterns of RH are due to two conditions, (i) the grounded path condition R(oun)D/HI(gh), "if (+round) then (+high) " or "if (+round) then not (-high)", and (ii) the Phonological Identity Condition (PIC). The grounded condition RD/HI, motivated from the physiological correlation between lip rounding and tongue height (mediated by jaw opening), holds of the targets of rules, thereby restricting the application of RH to high vowel targets. The PIC, on the other hand, is proposed to characterize the contextual identity requirement for rules. Two important properties of the PIC are noted: (i) that it is a non-representational condition, and (ii) that identity referred to by the PIC is implemented as logical equivalence. The non-representational PIC is motivated from the cases of "nonadjacent" identity (i.e. the cases where contextual identity can be defined across a transparent element, e.g. Khalkha Mongolian RH), where a representational analysis based on OCP-induced fusion fails because it results in either illformed "gapped" representations or violation of the No-Crossing Line Convention. Further, the equivalence interpretation of identity is proposed to break away from otherwise mandatory rules of redundant feature insertion for cases like Yawelmani RH, where identity refers to both values of a feature. Drawing on these two conditions, this dissertation demonstrates that the observed patterns of height-constrained RH are correctly predicted by recognizing two factors, the rule types characterized by these conditions, and the way that these rule types are invoked by languages.