Rewriting rules, derivations and underlying representations is an enduring characteristic of generative phonology. In this book, John Coleman argues that this is unnecessary. The expressive resources of context-free Unification grammars are sufficient to characterize phonological structures and alternations. According to this view, all phonological forms and constraints are partial descriptions of surface representations. This framework, now called Declarative Phonology, is based on a detailed examination of the formalisms of feature-theory, syllable theory and the leading varieties of nonlinear phonology. Dr Coleman illustrates this with two extensive analyses of the phonological structure of words in English and Japanese. As Declarative Phonology is surface-based and highly restrictive, it is consistent with cognitive psychology and amenable to straightforward computational implementation.