176LANGUAGE, VOLUME 68, NUMBER 1 (1992) Melanesian Pidgin and the Oceanic substrate. By Roger M. Keesing. Stanford: Stanford University Press, 1988. Pp. xii, 265. $37.50. Reviewed by John Victor Singler, New York University The 'Melanesian Pidgin' (MP) ofK's title refers to Tok Pisin (spoken in Papua New Guinea), Solomons Pidgin (the Solomon Islands), and Bislama (Vanuatu). K's scenario for the development of MP departs dramatically from previous accounts. He moves the stabilization of the pidgin back in time and vastly expands its site. This chronologic and geographic shift necessitates a change in the roster of potential substrate languages for MP. K argues for the dominance of Eastern Oceanic languages, not just those found in the Southeast Solomonic and North and Central New Hebrides subgroups of Eastern Oceanic but, crucially, those spoken hundreds of miles north and east of Melanesia in Pohnpei and Kosrae and the Gilbert Islands, as well as those in Rotuman and Fiji.1 The book is divided into three sections. After an introductory chapter, K presents his version of the history of MP (Chs. 2-5), drawing extensively on nineteenth-century sources. Next he makes the case for the Oceanic substrate, especially Eastern Oceanic (Chs. 6-10). Finally, he examines the close relationship between Solomons Pidgin and Southeast Solomonic languages, particularly Kwaio (Chs. 11-13). To the final section he adds an appendix that juxtaposes excerpts from narratives in Solomons Pidgin with excerpts from equivalent narratives in Kwaio. K sees MP as a descendant of the 'worldwide' nautical jargon that had reached the Pacific in the late 1700s. In K's view the critical stage for the stabilization and localization, i.e. the transformation from jargon to extended pidgin, began 'in the 1840s as whaling and trading ships began to frequent the islands ofthe central Pacific' (15). This is the significance ofthe non-Melanesian islands cited above, as trading posts, as havens for European beachcombers and deserters, and especially as a source of crew members for European and American ships. Indeed, K argues that 'crews of mixed-island origin' (4) were the principal actors in the Pacific pidginizing process. Knowledge of the pidgin was turned to economic advantage by Pacific Islanders in the next phase: as the sandalwood trade developed in Melanesia, former crewmen became interpreters and middlemen. In particular, K singles out the men of the Loyalty Islands. Next, when plantations opened in Samoa and Queensland in the era of the Labor Trade, the middlemen became involved in the recruitment and supervision of plantation laborers. In turn, as new recruits were brought in, 'finish times' (veteran plantation laborers) took over the supervisory roles. The plantation setting proved favorable for a broad dissemination of the pidgin, the laborers learning it from their supervisors, fellow Islanders. When the laborers returned home, they took the pidgin with them. In presenting this scenario for the development of MP, K is primarily arguing ' Jackson (1986) departs from K and the linguists whom K cites (Pawley, Blust, et al.) by arguing that, at present, insufficient evidence exists for the inclusion within Eastern Oceanic ofMicronesian languages, crucially the languages of the Gilberts, Pohnpei, and Kosrae. REVIEWS177 against Mühlhäusler's (1978, 1985a) advocacy ofa 'substantially separate origin for the New Guinea Pidgin lineage' (51). K points out that 'virtually all the syntactic patterns of Tok Pisin ... are found as well in the Bislama of Vanuatu and in Solomons Pidgin' (52). He acknowledges that Tok Pisin had a separate history from 1890 on, but he argues that '[m]ost of the essential syntactic and semantic/lexical patterns of Melanesian Pidgin are represented in the texts from the 1870s and 1880s (and the earlier texts we have seen)' (48). Specifically, he enumerates sixteen MP characteristics that are attested prior to 1890. These include most of MP's best-known features, e.g. 'the standardized use of a transitive suffix -em or -im\ 'use of the grammatical element -fela in a set of quite regular syntactic slots', 'the regularized use of the so-called "predicate marker" /' after noun subjects', and 'the regularization of long as an all-purpose locative particle' (48-50). Following the presentation of MP's history, K then sets...