Leo tuai: A comparative lexical study of North and Central Vanuatu languages (review)

The northern and central islands of the Republic of Vanuatu are home to slightly less that one hundred languages. The core of this book, chapter 5, is the reconstructed lexicon of the putative common ancestor of all but two of these, Proto– North-Central Vanuatu (PNCV).¹ Hence the title: if the speakers of PNCV knew about protolanguages, Leo tuai would mean 'protolanguage', being the PNCV reflexes of Proto-Oceanic (POC) *leqo- 'voice, language, etc.' and *tuqaRi 'long ago, old'.

Ross Clark has been conducting linguistic research in Vanuatu since the mid-1970s. His main research focus in those days was the Polynesian Outlier Ifira-Mele, but since that language has been heavily lexically influenced by the neighboring non-Polynesian languages of Efate, he began to make comparative notes on those languages. Like Topsy, the project grew, especially after the publication of Tryon (1976), which made available basic wordlists for nearly all Vanuatu languages, and then the subsequent appearance, over the next three decades, of dictionaries and longer wordlists for quite a number of these languages. The reconstructions themselves have been "around," in various versions and various formats, for many years, and have been used and cited by a range of linguists.² The eventual appearance of a published version is thus most welcome to those conducting research on languages of the eastern part of Oceania especially.

The first four chapters provide background of various kinds to the reconstructions. Chapter 1 (1–9), "The North and Central Vanuatu languages," provides an overview of various attempts to classify these languages, from Ray in 1893, through Grace and Capell in the 1950s and 1960s and Pawley and Tryon in the 1970s, to Lynch in 2001. There is no attempt to provide the evidence on which various subgrouping hypotheses were based: that is not the purpose of the chapter, or the book. Rather, Clark states that there seems to be a general consensus that there is a North and Central Vanuatu (NCV) subgroup of Oceanic, with its two major divisions being North and Central.

Important to the lexical reconstructions is Clark's concept of areas and local groups. Clark (1985) put the NCV languages into 22 "local groups," based on the cognation figures in Tryon (1976). These 22 groups themselves are placed into five "areas": two making up Northern Vanuatu, and three Central Vanuatu. No lexical item could be reconstructed for PNCV unless it had cognates in at least two of these five areal groups.

Chapter 2 (10–18), "Outline of comparative phonology," details the PNCV sound system and its development from POC. First, though, a word on orthography. Although Clark basically uses the same symbols as have become standard for POC (though see the discussion relating to PNCV *z and *v below), he makes three orthographic departures: the prenasalized voiced velar stop (POC *g) is written as *q in PNCV, the velar nasal (POC *d ) is written as PNCV *g, and the postvelar/glottal stop (POC *q) is written as PNCV *ʔ

The major developments from POC to PNCV, according to Clark, are:

1. 1. merger of the POC palatals *j, *c, and *s, and emergence of a new contrast of PNCV *z and *s;

2. 2. complex attrition of POc *R: in some cases, *R is lost throughout; in some others, it merges with *r throughout, as *r; in yet others, it is retained in the far north, as *r, and lost elsewhere, and in these cases Clark reconstructs PNCV *R;

3. 3. some loss of POc *y;

4. 4. lenition of POC *p as *v; and

5. 5. regular loss of POC final consonants (though see below).

The consonant inventory possibly errs on the side of conservatism (by which I mean the opposite of "risk-taking" rather than "aberrancy"), though Clark himself admits that this is the case. His consonant inventory shows a merger of POC *n and *ñ as PNCV...