Conflict between constraints in Optimality Theory (OT; Prince & Smolensky 1993) is always and only resolved by ranking. By contrast, conflict between rules in work in the SPE tradition (Chomsky & Halle 1968) is assumed to be resolved in two distinct ways: disjunctively when the rules describe sets of strings that are in a proper inclusion relationship, and conjunctively otherwise (i.e., with straight-up ordering when the sets of strings described by the rules merely overlap). This distinction is enshrined in the form of the Elsewhere Condition (EC; Kiparsky 1973, 1982), a well-established and -regarded principle within phonological theory and formal linguistics more generally.

The two questions I address in this talk are: (1) whence the distinction between proper inclusion and overlap in SPE work? and (2) is such a distinction necessary? Given certain assumptions about rule construction, some but not all rule conflicts do appear to require disjunctive resolution; proper inclusion was a first (though not entirely arbitrary) attempt to carve out the necessarily disjunctive cases. But I argue that the distinction between proper inclusion and overlap is a spurious one, buttressed only by the genuine distinction between disjunctive and conjunctive application. Conflict resolution in SPE is inherently conjunctive, which is why the EC is necessary to ensure disjunctive resolution of some kinds of conflicts. Conflict resolution in OT is inherently disjunctive, which is why additional assumptions are necessary to account for some of the kinds of conflict resolutions that conjunctive application appears to be better-equipped to describe (e.g., opaque rule interactions).