Robinson Hall A, #447
November 21, 2014, 11:00 AM to 09:00 AM
This dissertation applies an Optimality Theory Candidate Chains (OT-CC: McCarthy 2007) analysis to transient opaque productions which appear during early L1-acquisition (Smith 1973; Velten 1943; Cho and Lee 2003; Ettlinger 2009; Barlow and Keare 2008) and argues that early child opacity differs in critical ways from the type of opacity that appears in mature grammars as well as from what is predicted to be possible by current theory. I show that these previously unpredicted patterns of opacity require the theoretical modifications of lifting the B » PREC(A, B) metaconstraint ranking and splitting the (a) and (b) PRECEDENCE clauses into individually rankable constraints (Wolf 2011). I further show that comparison of these early child opacity patterns to the temporary opacity which emerges during adult L2-acquisition (Eckman et al. 2003) and delayed child L1-acquisition (Dinnsen et al. 2000; Dinnsen et al. 2011) reveals that the early child patterns are distinct from a general language acquisition effect. I argue rather that this opacity is a consequence of simultaneous versus sequential phonological process application during early L1-acquisition and is driven by a revised initial state of MARKEDNESS » PRECEDENCE » FAITHFULNESS; thus, it is limited to stages of early L1-acquisition and disappears after key re-rankings irreversibly alter default structure.