The Cambridge handbook of phonology (review)

This is an excellent book. It provides an informative, up-to-date (2005) synopsis of phonological research in the generative paradigm by recognized leaders as well as some rising stars. While de Lacy correctly observes that optimality theory (OT; Prince & Smolensky 2004 [1993]) is the most significant recent development in the field and most contributors work within this paradigm, the volume is not focused on technical developments in this approach. Rather, its chapters present a balanced overview, which in the best cases illustrate OT's insights into the results from earlier periods as well as the challenges these results continue to pose. Only a few chapters stray into the polemical or self-referential. There are some ninety pages of references. The book is beautifully produced with clear typeface, tables and drawings, comfortable spacing between lines, and ample margins. Separate subject and language indices as well as an IPA chart enhance its usefulness. While substantial (697 pages), it does not look or feel bulky.

De Lacy's introduction provides a useful overview of the volume, a review of the basic concepts, architecture, and notation of the OT model, and a synopsis of several themes that run through the various chapters, revolving around the role of markedness and representations in phonology, or more generally, formal vs. functional explanations. The volume is organized into five parts. Part 1, 'Conceptual issues', contains remarks on theory and methodology (ALAN PRINCE), functionalism (MATTHEW GORDON), markedness (KEREN RICE), derivations and opacity (JOHN J. MCCARTHY), representations (JOHN HARRIS), and contrast (DONCA STERIADE). Part 2, 'Prosody', features chapters on the syllable (DRAGA ZEC), metrical stress (RENE KAGER), tone (MOIRA YIP), intonation (CARLOS GUSSENHOVEN), and sonority (PAUL DE LACY). Part 3, 'Segmental phenomena', examines features (TRACY HALL), assimilation (ERIC BAKOVIC), harmony (DIANA ARCHANGELI and DOUGLAS PULLEYBLANK), and dissimilation (JOHN D. ALDERETE and STEFAN A. FRISCH). Part 4, 'Internal interfaces', discusses phonological interfaces with phonetics (JOHN KINGSTON) and syntax (HUBERT TRUCKENBRODT), along with morpheme position (ADAM USSISHKIN) and reduplication (SUZANNE URBANCZYK), while Part 5, 'External interfaces', includes chapters on diachrony (RICARDO BERMUDEZ-OTERO), variation (ARTO ANTTILA), acquisition (PAULA FIKKERT), learnability (BRUCE TESAR), and phonological impairment (BARBARA BERNHARDT and JOSEPH PAUL STEMBERGER). The chapters are of roughly comparable size (twenty to thirty pages each).

A major theme running through the volume, and the focus of this review, is the proper grammatical relation between phonetics and phonology. Under the standard 'modular feed-forward' architecture (Pierrehumbert 2002, Bermudez-Otero in this volume) that has characterized generative phonology since its inception, the only interface is the distinctive features, which provide the categories for lexical contrasts and phonological computations that are cashed out as the articulatory and acoustic aspects of the speech signal under grammatical control. OT's constraint-based format over 'outputs' has made it possible for phonetic properties to be woven directly into phonological computations, raising a host of questions that occupy the attention of much of today's research. Hale and Reiss (2000, 2008) and others marching under the banner of 'substance-free' phonology see the phonetics-phonology divide as a mind-body problem (how can the plus-minus, on-off phonology be affected by the phonetically continuous?), echoing Bloomfield's (1926:157) dictum that 'such a thing as a "small difference in sound" does not exist in a language'. Other researchers contend that phonological categories emerge when the threshold in a phonetic gradient is crossed or as prototypes in a vast storage space.

A well-known example where this issue comes to the fore, reviewed in Gordon's chapter, concerns the interpretation of Steriade's (1999) important finding of a hierarchy in the typology of contexts for laryngeal neutralization: preobstruent > word-final > presonorant. The...