| This study examines certain issues of French morphophonology under Optimality Theory, and addresses the question of how language-specific processes can be incorporated into a constraint-based analysis. Essential to my analysis is the use of output-to-output faithfulness. Although it is argued in the literature that an account based on the distinction between the root and the word is not available for all languages, this distinction is shown to be necessary to account for cyclic effects in French. In addition, I adopt local conjunction as a means of expressing language-specific constraints.; Chapter One describes the basic principles of OT, its variations, and the controversy over multiple levels.; In Chapter Two Gliding is reanalyzed with Rosenthall's mora-based approach as a foundation. I show that the distribution of high vowels and glides can be captured by the interaction of syllable structure, faithfulness constraints, and alignment-based constraints.; In Chapter Three four phenomena associated with nasal vowels and vowel nasalization are examined: three different patterns displayed by certain prefixes; reduplication; prosodization of latent nasal stops; lowering effects of nasalization. It is shown that the different affixal patterns can be captured by different dialectal rankings of an alignment constraint in conjunction with correspondence-based constraints.; The disparity between stem and reduplicated vowels, among other things, follows from the pattern described as a metaconstraint. Further, I briefly show how clitics and prenominal adjectives are prosodized. Finally, I argue that lowering effects of nasalization can be captured by the interaction of faithfulness and markedness constraints.; In Chapter Four I assume that schwa is represented as [-cons] to account for basic schwa realization patterns, and adopt local conjunction as a means of expressing language-specific constraints peculiar to schwa realization. Sequences of verb plus adverb, and encliticized imperative forms are also examined. It is argued that while sequences of verb plus adverb in finite sentences form intonational phrases, encliticized verbs form prosodic words. Finally, it is assumed that Closed Syllable Adjustment (CSA) cannot be, handled by constraint rankings. As a solution I adopt the view that CSA is the result of allomorphy. |
