Pathological Effects of Local Disjunction

Abstract

When working in Optimality Theory (Prince & Smolensky, 1993/2004; McCarthy & Prince, 1995), a balance must be found when positing constraints between those that are not powerful enough to explain attested forms of opacity, and those that are too powerful and predict problematic and pathological languages. Constraint connectives, which create new constraints whose violations are defined relative to those of two simple constraints, can be a simple and intuitive way to extend the power of OT to capture opacity. Local Conjunction (Smolensky, 1993; Moreton & Smolensky, 2002; Itô & Mester, 2003) is the best known connective, but since its inception considerable restraint has been placed upon it to prevent it from predicting pathological effects (Itô & Mester, 1996; Łubowicz, 2005; Pater, 2009). Yet, as explored in Wolf (2007), Local Conjunction is just one of 16 possible constraint connectives. Along with and, some have posited or, as we do in natural language and logic. I call this connective Local Disjunction.1 Local Disjunction was first posited by Hewitt & Crowhurst (1996), and explored further in later work by Crowhurst & Hewitt (1997); Crowhurst (2011); Downing (2000, 1998). In this paper, I argue that Local Disjunction creates pathological effects, and even restriction like that which we have applied to Local Conjunction cannot save it.