Phonetically based phonology (review)

The front matter describes Phonetically based phonology (PBP) as ‘centered around the hypothesis that phonologies are determined by phonetic principles’, with ‘phonetic patterns … expressed linguistically as grammatical constraints’. Indeed, all but two articles adopt an optimality-theoretic (OT) model (Prince & Smolensky 2004) whose constraints instantiate functional factors. So PBP is not a general overview of approaches to the phonology-phonetics interface, but this focused perspective has advantages: it underscores the consequences of implementational differences among related models, and it explores how well similar formalisms model diverse empirical phenomena. Contributors address sonority and perceptibility (Richard Wright), place assimilation (Jongho Jun), vowel harmony (Abigail R. Kaun), metathesis (Juliette Blevins and Andrew Garrett), contour tones (Jie Zhang), vowel reduction (katherine M. Crosswhite), perceptual distinctiveness (Edward Flemming), syllable weight (Matthew Gordon), lenition (Robert Kirchner), and the obligatory contour principle (Stefan A. frisch). An introduction (Bruce Hayes and Donca Steriade) places the articles in the context of typological and OT-based approaches to markedness, with discussion of obstruent voicing.

Many articles (Jun, Kaun, Zhang, Crosswhite, Flemming, and Gordon) have a similar structure: they discuss asymmetries in the crosslinguistic typology of some phonological phenomenon; review relevant phonetics research, sometimes contributing original findings; use phonetic factors to motivate OT constraints; and demonstrate that the permutations of these constraints generate the observed range of patterns, thereby linking the phonetics of the phenomenon to the typology of its phonological manifestation. This argumentation seems generally convincing, although sometimes phonetic studies of a single language are cited as the basis for typological generalizations. Of the remaining articles, Kirchner’s replaces the review of experimental results with predictions concerning biomechanical effort for different segment classes. Both Wright and Frisch present more detailed phonetic discussions, but less explicit phonological proposals. Blevins and Garrett relate phonetics to diachronic change rather than to a synchronic phonological model.

PBP is an important addition to the literature on the phonetics-phonology interface. Admittedly, some of the research dates back several years; all contributors except Blevins, Frisch, and Garrett completed dissertations at UCLA between 1995 and 2001, and most chapters represent summaries or extensions of these dissertation projects. Still, by collecting article-length explications of this influential work, PBP makes a useful overview or reference volume. The chapters are of a length and scope appropriate for graduate-level course readings, so the editors’ decision to include a separate list of references at the end of each article is particularly helpful. Another valuable aspect of this book is that contributors’ conceptions of the phonology-phonetics relationship are far from uniform. Some of their major points of difference reflect two fundamental debates: To what extent can phonology be directly equated with phonetics? And, does phonetics influence phonology by restricting synchronic grammars, or by shaping diachronic change?

The first of these questions receives more attention in PBP. Among the articles that assume phonetically motivated OT constraints, two approaches to the formal implementation of such constraints can be distinguished. The direct phonetics (DP) approach (Wright, Jun, Zhang, Flemming, Kirchner) allows constraints to be stated in terms of gradient phonetic properties such as a percentage of the possible perceptual cues for place contrasts (Jun), or a weighted sum representing the sonority and duration of a syllable rime (Zhang). Another view, which for convenience can be called the structurally mediated grounding (SMG) approach (Kaun, Crosswhite, Gordon), holds that constraints are phonetically motivated, but are nevertheless stated in terms of more traditional, symbolic phonological categories. Except for Gordon, SMG-oriented contributors in PBP do not defend their use of phonological structures as explicitly as DP proponents do for fine-grained phonetics, perhaps because SMG is closer to traditional views of the phonology-phonetics interface (though see also §5 of Hayes and Steriade’s introduction for discussion supporting aspects of both SMG and DP).

Proponents of the DP approach in PBP make...