Cambridge University Press, 2000. — 322 p.
This book analyzes some differences among English, Scottish and American accents of English, and shows how they developed and why they have their current form. Although the revised version of lexical phonology presented here is intended to describe present-day patterns, it can also show how historical sound changes gave rise to these patterns.
This book has two main goals: the re-establishment of a rule-based phonology as a viable alternative to current non-derivational models, and the rehabilitation of historical evidence as a focus of phonological theory. Although Lexical Phonology includes several constraints, such as the Derived Environment Condition and Structure Preservation, intended to reduce abstractness, previous versions have not typically exploited these fully. The model of Lexical Phonology presented here
imposes the Derived Environment Condition strictly; introduces a new constraint on the shape of underlying representations; excludes under-specification; and suggests an integration of Lexical Phonology with articulatory phonology. Together, these innovations ensure a substantially more concrete phonology. The constrained model is tested against a number of well-known processes of English, Scottish and American accents, including the Vowel Shift Rule, the Scottish Vowel Length
Rule, and [r]-Insertion, and draws interesting distinctions between what is derivable by rule and what is not. Not only can this Lexical Phonology model the development of low-level variation to phonological rules, and ultimately to dialect differentiation in the underlying representations; but a knowledge of history also makes apparently arbitrary synchronic processes quite natural. In short the phonological past and present explain one another.