Disjunctive application is a type of interaction between phonological mappings that has received special attention since the inception of generative phonology (Chomsky & Halle 1968) and has significantly impacted research in other subfields in linguistics. The principle commonly held to be responsible for disjunctive application, the Elsewhere Condition (Kiparsky 1973), is argued in this book to be little more than a collection of necessary stipulations within Chomsky & Halle's original SPE framework. By contrast, disjunctive application is shown to follow automatically from the most basic assumptions of Optimality Theory (Prince & Smolensky 1993), with no added stipulations necessary.
The stage for these demonstrations is set with in-depth discussion of the history and analysis of blocking interactions, of which disjunctive application is a special case. The distinguishing feature of disjunctive application is shown to be complementarity. The analyses of two types of complementarity (allophonic or ‘unbounded’ complementary distribution as opposed to ‘bounded’ complementary distribution) in both SPE and Optimality Theory are discussed in detail, and it is shown that both have been analyzed very differently in SPE but very similarly in Optimality Theory.
The various stipulated components of the Elsewhere Condition are then discussed and contrasted with the lack of any such stipulation in Optimality Theory. This is followed by a proof of two theorems within Optimality Theory that solidify the result that two mappings in a particular formal relation to each other are bound to apply in complementary fashion.