Phonological representations: Their names, forms and powers By John Coleman (review)

206 LANGUAGE, VOLUME 76, NUMBER 1 (2000) want to be included, they should 'do something about it' (viii), apparently produce their own first person singular feminine volume. Readers can draw their own conclusions about such a position. These essays vary as much as the work and contributions of the authors, and all are interesting. Some describe an early introduction to language and linguistics in great detail, giving a clear sense of the ways in which people entered the field in the first halfofthe twentieth century, e.g. Catford in Scotland and England; Pulgram in Vienna; Pike in Arkansas, Mexico, and Michigan; Stockwell in Virginia. There are not many common features in these early experiences except for the study of Latin, mentioned by more than half of the writers (Ferguson, Francis, Kuöera, Lamb, Ney, Pulgram, and Stockwell). One cannot help but wonder whether this shared background has influenced the way these and other linguists look at language and whether future linguistics might be affected by the decline in the availability of Latin studies in our schools. Each of these linguists has followed a different path, only Stockwell ever a contributor to the development of generative grammar, several others expressing overt hostility to generative linguistics, even to structuralism. What attracted many was applied linguistics, broadly defined. Catford, with a lifetime of contributions to phonetics and language teaching, suggests 'that theory must always be subservient to application, and that no theory is "the best" for all applications' (28). Ferguson, founding director ofthe Center for Applied Linguistics (1959-66), saw himselfthroughout his career as 'operating with a professional tension between solving practical language problems and doing academic linguistics' (48). Lamb, creator of stratificational grammar, was at one time head of a machine translation project and later left academia briefly to work m electronics. Francis and Kucera, in addition to their contributions to English and Slavic linguistics, respectively, developed the Brown Corpus database of American English. Stockwell has a long and continuing record ofcontributions to English studies, especially in the history of the language. Pike has consulted for the Summer Institute of Linguistics on procedural methodology, field problems, and the linguistic analysis of more than a hundred languages. Samarin, too, has guided students through field work, his own specialization the languages ofcentral Africa. And Teeter continues his publications on native American languages, work that began with his dissertation on Wiyot under the direction of Mary Haas. Most of the essays conclude with extensive bibliographies and references to more complete lists of the authors' works. They present a sample, though not a representative one, of the greater diversity that has formed twentieth century American linguistics. [Julia S. Falk, Michigan State University.] Phonological representations: Their names, forms and powers. By John Coleman. Cambridge: Cambridge University Press, 1998. Pp. xvii, 345. This book is a prime example ofthe growing discipline ofcomputational phonology. It is highly recommended reading for both computational linguists and phonologists. Formal analysis is pursued to a level of explicitness reminiscent of the early era of generative syntactic theory. However, it gives relatively more weight to the presentation of language data than was the practice in early formal language theory research. The first half of the book is primarily formal while the second half is more empirical. As the name implies, the focus of this book is on representation and not 'derivations' (e.g. transformations , metrical grid construction, autosegmental linking or delinking). It is Coleman's pnmary goal to develop the theory of declarative phonology (DP), which is based on unification grammar. DP is entirely representational, with no derivational component. In DP, generalizations that might be encoded by rules in other theories are encoded by constraints, represented as partially specified structural descriptions. A secondary theme of the book is to highlight the difference between phonological representation and phonological notation in order to sort out differences between theories that are substantive from those that are merely notational. C asserts that this can only be accomplished by considering phonological representations in a formally rigorous way. The first half of the book introduces the notation and formal machinery that is the basis for the analysis to follow (1 8-45) and reviews and criticizes transformational segmental phonology (46...