Vowel phonology of Old and Middle English: An optimality-theoretic account

This dissertation presents the application to sound change in English of a constraint-based approach to phonology. I employ Optimality Theory (Prince & Smolensky 1993) in the analysis of diphthong-related sound changes in Old and Middle English. I pay special attention to Old English Diphthongization, a variety of Early ME developments following OE Breaking and Middle English Diphthongization. I investigate the phonetic and phonological mechanisms involved and provide explicit formal accounts for these sound changes. With a view of the historical phonology of English as a set of recurrent processes, I argue that historical sound changes can be localized to a few partial rankings within OT.; Under the assumption that consonants as a breaking conditioner weakened by the time of OE Breaking, I propose an OCP-based approach to OE Breaking and its related dialectal variation. I show how a syllable-based perspective and ambisyllabicity can overcome the difficulty in the account of breaking peculiarities that traditional segment-based approach cannot. Through detailed OT analyses of OE Breaking, I show the emerging partial rankings can characterize the breaking process as well as derive the asymmetric patterns in OE Breaking.; I argue that the particular gestural configuration of ‘V + /h, r, l/’ is responsible for the repetition of diphthongization and monophthongization throughout the early history of English. I also claim that sound change can be understood as fluctuation between different ways of parsing the raw phonetic substance of the specific segment sequence. Then, I present an OT analysis of each of EME developments after OE Breaking. Through an analysis of these EME developments within OT, I demonstrate how the attested EME dialects are matched by the actual typologies as spelled out for OE Breaking. In addition, I construct a partial grammar for the EME period, an invariant part of EME grammars. Then, I present an OT analysis of ME Breaking after EME monophthongization. I show that ME Breaking can be accounted for in the same way as OE Breaking, only with a different constraint ranking.