Contact-Induced Change in Southern Bougainville

Abstract

1. INTRODUCTION 1.1 ABERRANT AUSTRONESIAN LANGUAGES OF MELANESIA. The Austronesian languages of the Melanesian region have long been viewed as different iiom those of Micronesia and Polynesia in ways that raise questions regarding their historical development. Grace (1981, 1992, 1996) describes this difference as a continuum from more to more exemplary languages, noting the presence of many more aberrant Austronesian languages in Melanesia than in the other two regions of Oceania. By aberrant, Grace (1992, 1996) means a language with which the application of the Comparative Method of historical linguistics yields results that are complex and difficult to interpret. (For discussion of the comparative method, see Rankin 2003 and Harrison 2003.) Grace lists four ways in which languages can be aberrant: (i) they show relatively few cognates with other Austronesian languages; (ii) they are structurally atypical of the Austronesian languages of their subgroup (for example, Oceanic); (iii) they have large phoneme inventories and phonetic features that are atypical of Austronesian languages; and/or (iv) they do not show unambiguous regular sound correspondences, but rather often display multiple sets of contrasting correspondences. Languages with one or more of these features raise the question of what kinds of historical explanations account for their atypical nature. There is continuing debate regarding the histories of these languages in Melanesia (see Pawley 2006). (1) Ross (1988) demonstrated that the application of the Comparative Method to the Austronesian languages of northwestern Melanesia, namely Papua New Guinea and the northwestern Solomon Islands, does in fact yield regular sound correspondences and patterns of shared innovations that indicate phylogenetic relationships among the languages of the region. However, Ross (1988) also notes that these languages are characterized by a high degree of linguistic diversity or, in Grace's (1981) terms, are aberrant with respect to features (i) and (ii). For example, in terms of lexicon, languages of the Northwest Solomonic (henceforth NWS) subgroup are among the most innovative within the Austronesian family. That is, these languages retain proportionally fewer Austronesian etyma than other languages in the family (see Blust 2000, Pawley 2009, Ross 2010). Also, certain languages within this subgroup, in particular the Mono-Uruavan ones, show highly atypical grammatical structures in comparison with other Austronesian languages of the region. There is agreement in the literature that in this northwest region of Melanesia the linguistic diversity reflects histories of contact-induced change, resulting from widespread bilingualism among speakers of Austronesian and Papuan languages (Lynch 1981, Ross 1988, Pawley 2006). Thurston (1982, 1987, 1994) and Ross (1996, 2008) investigate the atypical or aberrant grammatical structures in Austronesian Lusi (West New Britain, Papua New Guinea) and Takia (Karkar Island, Papua New Guinea), respectively. They show that these structures are similar to ones found in neighboring Papuan languages, and that they reflect contact-induced change. Comparable aberrant structures in other Austronesian languages of northwestern Melanesia, including the three Mono-Uruavan languages of the southern Bougainville region, have been presumed to reflect similar processes of contact-induced change. 1.2 LANGUAGES OF THE SOUTHERN BOUGAINVILLE REGION. All of the Austronesian languages of Melanesia east of about 138[degrees] East are part of the large Oceanic branch of the family (see Lynch, Ross, and Crowley 2002). The Oceanic languages of Bougainville, Buka, and Nissan islands in Papua New Guinea, and of the Shortland Islands, Choiseul, Santa Isabel, (2) and New Georgia in the Solomon Islands form the NWS subgroup within Oceanic (Ross 1988). Within NWS, it is the Mono-Uruavan subgroup (Ross 1988) comprising three languages--Mono, Uruava and Torau--that display highly atypical grammatical structures. …