A grammar of Mantauran (Rukai) (review)

Mantauran is one of the six dialects of the Formosan language Rukai, spoken in the south-central region of Taiwan. It is spoken by only 250–300 people and is highly endangered, with only a few elderly speakers still fluent. This alone is reason enough to document the language that in a generation or so will probably no longer be spoken. Rukai is unique in that it apparently exhibits an accusative case-marking system, while most other Formosan languages are arguably ergative, and it does not exhibit the widespread “focus” system characteristic of the so-called “Philippine-type” languages of Taiwan, such as Amis, Kavalan, Bunun, Thao, and Atayal. While a substantial grammar exists of one of the other dialects, Tanan Rukai (Li 1973), until Elizabeth Zeitoun (henceforth EZ) began her research on the language, there was very little information available about the morpho-syntax of this dialect. With this grammar, we now have extensive coverage of two considerably different dialects of Rukai.

EZ's goals in writing the grammar in effect match the reasons given above. They were “to provide a description of the most salient characteristics of the grammar of Mantauran in order to reach a better understanding of this language and second, to make available enough empirical data to show in what respect Mantauran differs from the other Rukai dialects and other Formosan languages in general” (13). In these respects it is clear that EZ has succeeded admirably. The grammar is a model of clear prose and elegant argumentation supported by a multitude of carefully chosen examples drawn primarily from over 600 pages of transcribed narrative and folktale texts that the author collected from her primary language consultant.

Chapter 1 discusses the current status of Mantauran Rukai (henceforth MR), providing brief notes on some of the commonly used loanwords, most introduced during the Japanese colonial period, and it is to this language, according to the author, that speakers persistently resort. The possibility that Japanese has also influenced the morphology will be commented on below. This chapter also provides the theoretical orientation of the work, stating that the grammar is a functional and empirically based account of MR, utilizing principles laid out by Dixon (1997) and Noonan (2006). The Conclusion (463) also notes that the grammar is “not circumscribed in any formal theory.” Nevertheless, various syntactic operations used by formalists, such as “raising,” form part of her argumentation to support one analysis over another in later chapters. A succinct outline of the grammar closes the chapter.

Chapter 2 provides a standard structuralist account of the phonology and morphophonemics of MR, and a summary of the sound changes that have occurred in the language from Proto-Rukai as reconstructed by Li (1977). Of particular interest in this chapter is the evidence, rare in Austronesian languages, of a contrast between vowel-initial forms, and glottal stop-initial forms, such as ʔoɭipotso ‘unwrap’ vs. oɭipotso ‘wrap’. Of interest in the morphophonemics section are the various processes labeled by EZ as rightward and leftward “glottal hopping” associated with the addition of various affixes. Typically, for much of the grammar, EZ provides a clear summary of the data, drawing generalizations and providing explanations where possible, and acknowledging problems that she is unable to account for.

Chapter 3, Morphological units and morphological processes, begins as do most chapters with definitional statements and supporting general literature. EZ briefly outlines various approaches to the description of words, including IA and IP morpheme-based accounts, generative lexeme-based morphology, word and paradigm, and seamless morphology, and makes explicit her claim that, despite recognized difficulties with morpheme-based accounts, this model allows a transparent treatment of the structure of MR words, helps account for the distribution of some forms, and facilitates a comparison with other Rukai dialects and other Formosan languages (45). She concludes that typologically MR is a synthetic-agglutinative language, with a wide range of “dependent” morphemes, or...