Talking rhythm stressing tone attempts to accomplish two equally significant aims. First, it attempts an exposition of a theory of prominence that provides a coherent synthesis of diverse views on phonological prominence. Second, it attempts an explanation of the structure of some prosodic features of Anglo-West African creoles using this theory. The proposals of the theory of prominence provide fresh insights into issues such as moracity, tone and stress assignment, and syntax-tone interaction. The theory proposes that well-known issues such as the association of stress and tone with moras or positions within a domain in languages result from phonological prominence effects. This is the first time an integrated approach, in the form of prominence theory, has been proposed to look at these issues.

The exposition of the theory of prominence is presented in chaps. 1, 3 and 4. Devonish proposes two major prominence types, positional prominence and lexical prominence. Positional prominence, a predictive prominence, assigns focus and places emphasis on a unit within a phonological domain by rule. Positional prominence may be additive or subtractive. The phonetic manifestations of positional prominence are segmental and tonal features – segmental in terms of duration, and tonal in terms of pitch. Lexical prominence, an unpredictive prominence, assigns tone and intonation to units in a phonological domain that already has lexical specifications of the same phonetic features, duration and pitch. The two types of prominence do interact in certain cases. Devonish illustrates elements of the theory by drawing from several prosodic analyses of language data from previous publications as instances of prominence effects. For example, prosodic data were drawn from English, the African languages Twi, Yoruba, and KiKongo, and an Amerindian language. What seems to be a serious problem here is that Devonish uses loanwords in the African language data only, while in the other languages he uses indigenous items to illustrate his points. I believe readers would see this as an imbalance in this work. Also, there is no critical analysis of the theoretical positions that the data taken from other publications were selected to support. This may be a serious weakness in mainstream phonological literature, but it saves us the mental torture of having to spend grueling hours going through complicated and sometimes irrelevant critiques of other theories. Devonish's theory is a complementary and not a competitive prosodic theory.

Chap. 2 addresses the history of Anglo-West African pidgins and creoles, beginning with a justification of using the term “Anglo-West African” in the work. He states that earlier terms are inadequate on the grounds that they give “the unfortunate impression that one is dealing with basically English linguistic systems which have been Africanised” (p. 22). He proposes that “Anglo-West African” avoids this negative connotation. In my opinion, the term is not transparent, and Devonish's criticism of earlier terms can be applied here, as well. An alternative is the term “Caribbean English-lexifier Creoles,” used by Faraclas (ms.). This term simply implies a strong English lexical base for the creoles that developed in this region. Further, in this chapter and throughout the work, Devonish implies a major distinction between Guyana and the Caribbean. This may create an unjustified distinction in the reader's mind. Linguistically, sociohistorically, and culturally, Guyana and the rest of the Caribbean have always been known and called “the Caribbean.” Geographically, this distinction may be acceptable, but it is not one that is necessary in a work of this nature.

Chaps. 5 and 7 address the historical evolution of rhythm and tone in English lexifier creole varieties in the Americas. Devonish uses modern creoles such as Saramaccan, Djuka, Kwinti, Paramaccan, and Boni to trace the phonological history of earlier varieties of Caribbean and Guyanese creoles. He proposes that these varieties with major African traits provide the best source for the reconstruction of rhythm and tone in these creoles with the strategies applied in loan adaptations found in the major source African languages (Twi, Yoruba, and KiKongo) as the bases for the prominence effects on the Caribbean creoles. A survey of the vowel systems of the creoles provides the necessary background for understanding the evolution of tone and stress in the English lexifier creoles in the Caribbean. The problem in these chapters is the highly speculative nature of the discussion, with very little primary data to illustrate the evolutionary processes in the creole varieties discussed.

Devonish continues the analysis of Guyanese prosody in chap. 6. He incorporates the well-known phenomenon of extrametricality to account for the behavior of certain deviant prominence patterns. Subtractive prominence is major factor in Guyanese Creole. Although the analysis of Guyanese is quite insightful, nevertheless the acoustic data derived from a single speaker of Guyanese Creole is not sufficient evidence to illustrate prominence.

Chap. 8 considers prosodic structures of creoles in West Africa (Krio, Cameroon/CP, and Nigerian Pidgin English/NP). (The use of this term suggests that NPE is a variety of English, but Nigerian Creole scholars now use the more neutral term Nigerian Pidgin/NP.) These Creoles have prominence manifested in similar ways: penultimate prominence, which also attracts segmental prominence. Evidence in support of the similarities between Krio and other West African creoles is not conclusive. Devonish relies on the prosodic structure of English loan words. Krio, CP, and NP are spoken in an environment surrounded by hundreds of African languages. The choice of English loan words provides a partial picture of prominence effects in these Creoles. A complete picture of prominence in these and other West African creoles can be accomplished only through a study of the effects of indigenous African loans, English loans, and items unique to these creoles.

In chap. 9, Devonish compares the position of H tone in many English-derived bisyllabic words in the Caribbean creoles and those of West Africa. He notes that tone-shifted items differ significantly between the West African varieties and their counterparts in the Americas. The difference is found on the first syllable: HL versus L. In both cases, however, the HL surfaces on the second syllable. Devonish proposes three possible sources for this difference, two of which are flawed. The third is argued to be the plausible source: that tone shift was an innovation in Barbadian Creole that subsequently spread to other creoles in the Americas and across the Atlantic to West Africa. The source for H-tone shift is a very interesting issue that is arguably open to further research.

The crucial issue raised in chap. 10 is prosodic relexification in Anglo-West African creoles. At the suprasegmental level, relexification involves adapting the citation HL melody of the English forms and applying it across the board in the entire lexicon. A very strong point in this chapter is Devonish's use of a step-by-step derivational process of items of KiKongo origin in Saramaccan (191). Although very little prosodic analysis of African loans is carried out, the proposal that the application of HL citation melody across the board in the lexicon of these creoles is suspect.

Like many theories, linguistic and nonlinguistic, the work draws heavily from data taken from published sources. This is an unavoidable weakness. Another problem is the book's organization. Related discussions are split between noncontiguous chapters. For example, chaps. 1 and 3 discuss the exposition of the theory. Chap. 2 should have been the first chapter. Chaps. 5 and 7 could either be combined into one or run sequentially. Further, a graphical representation would have greatly enhanced the exposition of the theory.

Despite the weaknesses pointed out in this review, Devonish's work is highly commendable in that it provides a cohesive view of an integrated theory of phonological prominence. It does this elegantly with language that even a non-phonologist can readily comprehend, a characteristic lacking in many a theory of phonology. The addition of an appendix containing creole language data makes this work even more attractive to readers and phonologists in particular.

In the scheme of things, creole language data have had little attraction or impact in the development of mainstream phonological theories. This work serves as a major departure from that trend. This should be seen as a major contribution to linguistic scholarship. It further demonstrates the potential contributions creole language data can make to any overall theory of phonology. Although the analysis in the study presents the theory as a theory of prominence in loan phonology, I see the potential for the application of the theory to other issues addressed in the prosodic phonologies of creoles and other languages.