Word-prosodic systems of Raja Ampat languages (review)

This book deals primarily with the tonal system of Ma'ya and its historical origins, and secondarily with a description of the tone system of Matbat, Both are languages of the South Halmahera branch of South Halmahera—West new Guinea (SHWNG), a subgroup of Austronesian. Remijsen presents evidence showing the South Halmahera membership of Ma'ya and Matbat, a subgrouping consistent with that of Blust (1978). Beyond the primary focus on Ma'ya and the secondary focus on Matbat (chap. 5), there is a wealth of short treatments of various other Raja Ampat archipelago languages, some in the initial chapter and some in the rich appendixes. Brief wordlists, grammar notes, and commentary on the social context and patterns of multiligualism are also given for Matbat, Misool Ma'ya, Butleh, Salawati Ma'ya, Langayan Ma'ya, Wauyai, Kawe, and Ambel. Despite their brevity, these are valuable, given the lack of data on many of these languages.

The book is dedicated to Lex van der Leeden, and appropriately the core of the book is R's insightful reanalysis of van der Leedens's analysis of the Ma'ya tone system, particularly of van der Leeden 1983, 1993, 1997, ms. 1, and ms .2. The reanalysis in R's book makes a contribution to the small but growing instrumental literature on stress, particularly on stress in tone languages (Potisuk, Gandour, and Harper 1996). R takes van der Leeden's four-tone system and reanalyzes it as a three-tone system plus contrastive stress, an otherwise typologically unattested system. The analysis is convincing. Unlike the other three tones, which do not differ significantly along these parameters, van der Leeden's fourth tone has unique characteristics typically associated in the literature with stress, not tone—namely, duration differences, vowel quality distinctions, and selective intensity > perceived loudness (cf. Beckman 1998). In this regard, it should be pointed out that the Ma'ya tone system differs from many Southeast Asian tone systems in that Southeast Asian tones are often differentiated not just by pitch, but also by duration and vowel quality (Henderson 1967), making the factoring out of stress more difficult. This difference between Ma'ya and Southeast Asian systems makes Ma'ya a much better candidate for sorting out the interaction between stress and tone. The findings are straightforward: the duration, vowel quality, and selective intensity distinctions occurring with the so-called fourth tone indicate that it actually marks stress, not tone. The instrumental findings use the appropriate statistical analyses. A minor shortcoming acknowledged by the author is the lack of some sort of perception study. However, even though such perceptual studies might reveal more about what the speakers themselves use as cues, the absence of such a study does not undermine the basic findings.

The fourth chapter compares and contrasts the tone systems of three of the five dialects of Ma'ya: Salawati, Misool, and Laganyan. In part, the motivation is straight comparison and in part R is interested in challenging Donohue's 1997 claim that Ma'ya is a pitch-accent language. What emerges is a rice, albeit brief, study of the strikingly different manifestations of a three-tone system in what are only marginally different dialects.

Not unexpectedly, given the complexity of the data and the language situation, the diachronic chapter, Chapter 6, answers only a few of the crucial questions. It is evident that the tone system is a secondary development, as it is obviously not inherited from Austronesian, nor even from the SHWNG subgroup. The older pattern of stress placement, in contrast, is inherited. Here, R contributes one piece to the ultimate solution by using two processes identified by Blust (1978): a syncope pattern and an apocope pattern, which occur in all the languages of what R terms the Raja Ampat-South Halmahera branch of SHWNG and in several languages of the West New Guinea branch to suggest the earlier placement of stress, as it can be assumed...