Tonio Green has just posted a complete draft of his interesting book Phonology Limited on the Rutgers Optimality Archive. For those of you who do not regularly visit the Archive or subscribe to the Optimal List — the latter having become pretty much just an announcement list for new postings on the Archive — the abstract for the book is copied below.
Phonological theory was revolutionized by the 1993 appearance of Prince and Smolensky’s manuscript ‘Optimality theory: Constraint interaction in generative grammar’, finally published eleven years later. The idea that all phonological processes in all languages could be reduced to a language-specific hierarchical ranking of universal constraints proved highly attractive to theoretical linguists around the world, but it remains of course very controversial as well, with many preeminent linguists either unconvinced of its accuracy or convinced of its inaccuracy. In ‘Phonology Limited’ I intend to address some of the concerns that opponents of optimality theory (OT) have brought up, and to argue that many problems encountered in OT are solved by allowing not only universal constraints on phonological well-formedness (markedness constraints) into the theory, but language-specific constraints on morphemes as well. Previous approaches that have assumed only universal markedness constraints have run into problems with phonological opacity, allomorphy, static irregularities, and lexical exceptions, and have often required very powerful mechanisms to deal with those problems. In this book, on the other hand, I argue that any process commonly held to be ‘morphophonological’ is actually part of the morphology and is to be analyzed in terms of language-specific morphological constraints. As a result, phonology is based on the interaction of faithfulness constraints with truly universal markedness constraints; everything language-specific is allocated to the lexicon.