According to the listing at the end of this book there are 466 Oceanic languages, making this the largest well-defined subgroup in the Austronesian (AN) language family short of Malayo-Polynesian. Well-known earlier surveys of the Oceanic branch of AN include Codrington (1885) and Ray (1926), both of which were restricted to central and southern Melanesia (Solomons, Santa Cruz, Vanuatu, New Caledonia and the Loyalties), and a few extraneous languages (Rotuman, Duke of York in Codrington). The present volume (TOL) is the first attempt to survey the entire Oceanic subgroup, including areas covered in earlier surveys together with western Melanesia, Micronesia, and Polynesia. It differs from Lynch (1998), for example, in aiming at massive documentation for areal specialists rather than an introduction for the general reader.
Following a preface and lists of abbreviations and illustrations, the volume is divided into five chapters: 1. The Oceanic languages, 2. Sociolinguistic background, 3. Typological overview, 4. Proto Oceanic, 5. Internal subgrouping. These are followed in turn by 43 language sketches that range from nine pages (Sakao) to 33 pages (Takia), a listing of all known Oceanic languages by subgroup, a list of references, and an index to chapters 1-5. While the authorship of the sketches is indicated, that of chapters 1-5 is not. Authorship of the sketches is as follows: written by Ross (Mussau, Takia, Bali-Vitu, Siar, Taiof, Sisiqa), adapted by Ross (Kele, Kairiru, Jabêm, 'Ala'ala, Kaulong), adapted by Lynch and Ross (Banoni), by Joyce Sterner and Ross (Sobei), by Mike Anderson and Ross (Sudest), written by Crowley (Gela, Mwotlap, Vinmavis, Sye), adapted by Crowley (Sakao, Port Sandwich, SE Ambrym), abstracted by Crowley (Raga), written by Lynch (), abstracted by Lynch (Cèmuhî, Xârâcùù, Iaai, Ulithian, Puluwatese, Marquesan), adapted by Lynch and Rex Horoi (Arosi). Assuming equal contributions to chapters 1-5, Ross appears to have contributed approximately 280 pages, Lynch about 158, and Crowley about 148. Twelve other sketches were written or adapted by other contributors, making this volume partly coauthored and partly coedited.
TOL is no ordinary book. It clearly will take its place as the successor to Codrington and Ray, and in every respect it reflects the tremendous advances in factual knowledge and theory that have taken place in the more than three-quarters of a century since the last of these appeared. Chapter 1 (1-22) surveys the AN language family, the geography and demography of Oceania, language contact, the history of research on Oceanic languages, and language names. This is a useful introduction, marred only by a higher than expected number of typos in a volume that generally has been carefully proofed: p. 6, 'Mao 1.3' for 'Map 1.3', 'Polynesian Trangle' for 'Polynesian Triangle', and p. 19, Spanish koopwure 'corrugated iron', Trukese cobre 'copper', where the source and borrowing languages have been reversed. [End Page 544]
Chapter 2 (23-33) discusses some of the major social parameters of language use in Oceania. One idiosyncratic feature that appears here and later in the book derives from Crowley's (1998) preference for "copying" over the traditional term "borrowing." Given the embedding of the borrowing concept in such terms as "borrowing," "loanword," "loanblend," "loanshift," "loan translation," and "lending language," and the longstanding use of the term "borrowing" in cultural anthropology for cultural "copying," the insistence on terminological novelty in this instance probably will strike some readers as quirky. This chapter concludes with a curious observation: "those areas where Melanesian vernaculars appear to be under greatest threat are those [that] have non-Austronesian speaking populations." It is suggested that this difference is due to size of language community: non-AN speaking groups tend to be smaller, and smaller communities tend to lose their linguistic and cultural integrity more easily than larger communities. However, unless there is evidence that such small communities have remained roughly the same size for generations (and hence that size alone is not the crucial determinant in language survival) such a statement...