McCullough Kerry. The phonological and orthographic consequences of Irish initial consonant mutation

University of Arizona, 2022. — 227 p.This dissertation investigates a typologically rare linguistic phenomenon found in Irish from three different perspectives: how it challenges phonological theory, how it is used by contemporary speakers, and how its written representation affects its acquisition. Initial consonant mutation (ICM), as it appears in the Celtic languages, is known to integrate numerous components of grammar including phonology, morphology, and syntax. At its core, initial ICM displays a regular phonological alternationin a phonologically unpredictable environment. The simultaneous cooperation and dichotomy of this process has challenged linguists, language users, and language learners, alike. The three main chapters of this dissertation examine the challenges each of these populations face in understanding (and using) ICM; their heterogeneous methodologies and bootstrapped research questions tell a more compelling and comprehensive story about this unique language phenomenon. Chapter 2 provides a novel analysis of the Irish mutation system within the Emergent Grammar model (Archangeli and Pulleyblank, 2014, 2015, 2016, 2017, 2022). By allowing the phonology and morphology to share the responsibilities of the alternations, this framework accommodates subtle details of the mutation system that modular models of grammar disregard. Moreover, the mutation system’s weak points pertaining to morph memorization, feature abstractness, and pattern generalizability fall out logically in an Emergent Grammar framework. How the mutation system is changing and what is so challenging are predictable. In fact, the remaining chapters are an extension of the empirical studies used to support the Emergent Grammar analysis in Chapter 2. Chapter 3 reports the findings from fieldwork conducted in Cork City, Ireland that documents urban speakers’ current use of ICM. Of particular interest are the phonological restrictions on consonant mutation, which are known to be shifting inrelated Scottish Gaelic (Hammond et al., 2017). Employing traditional elicitation methods, data was collected from over 30 regular Irish speakers that contained expected environments for the cluster restriction, the coronal restriction, and the [s] ∼ [t] alternation. Results indicate that the patterns of non-alternation and exceptional alternation are in flux with an overall tendency for no mutation if such a pattern is grammatical. This chapter also features discussions of dialectal overlap and new speakerhood before reconnecting to the Emergent Grammar model. Chapter 4 presents the results from a novel word learning study which explores the learnability of mutation-like morphophonological processes with differing orthographic conventions. In this study, the real-world contrast between Irish andManx writing systems forms the basis of the investigation: