Austronesian and theoretical linguistics (review)

Theoretical phonologists have always used a considerable amount of Austronesian data, particularly in connection with work on nasality and reduplication. During the first several decades of the generative era, however, the formal syntax community paid far less attention to Austronesian than the size, typological variety, and geographical extent of the family would seem to warrant. Two significant events helped to amend this situation. One was the publication of Guilfoyle, Hung, & Travis 1992, the first credible attempt within the principles-and-parameters framework to deal with nondemoting, Philippine-type voice systems. The other was the formation, in 1994, of the Austronesian Formal Linguistics Association (AFLA).

The papers in this volume represent a selection of those presented at the 14th annual AFLA conference, held in May of 2007 at McGill University. From among the thirty papers and eleven posters presented at the conference, the editors chose five phonology and twelve syntax papers for publication. These papers offer a representative sample of the issues of current interest within Austronesian formal linguistics. The sample of languages is fairly well balanced among Oceanic, Formosan, and Western Malayo-Polynesian; Malagasy and Tagalog are represented, but do not dominate the line-up as they have in some previous AFLA conferences. This is a handsome and well-edited volume, and the papers are generally well written and full of interesting data. But in my mind, what makes these papers so valuable is that they are written by formalists who are also language experts. The discussions of theoretical issues are based on detailed knowledge of the languages being discussed.

One recurring topic of discussion in recent years has been how to reconcile the surface word-order properties of various Austronesian languages with the universal underlying structure posited by Kayne’s (1994) antisymmetry hypothesis, and with Cinque’s (1999, 2005) proposals concerning the universal ordering of adverbial elements and adjectives. The challenge is particularly acute with regard to certain verb-initial members of the family.¹ Diane Massam, in ‘Deriving inverse order: The issue of arguments’ (271–96), addresses both the order of arguments (relative to the verb and to each other) and the order of adverbial elements in Niuean. She summarizes her earlier analysis of VSO order as resulting from remnant VP-movement. She then considers the ordering of adverbial particles, which is (approximately) the reverse of the underlying order posited by Cinque. It has been claimed that inverse ordering of adverbials correlates with verb-initial order, and that this correlation can be accounted for by assuming that adverbs arrive in their surface position by ‘roll-up movement’. Massam supports this view, but points out that in Niuean, this analysis runs into technical difficulties under normal assumptions about the underlying (merged) positions of internal arguments. She argues that in Niuean, all arguments must merge outside the VP, as specifiers of functional heads.

Similar issues are addressed by Elizabeth Pearce, in ‘Possession syntax in Unua DPs’ (141–62). The inventory of possessive constructions in Unua is not as complex as in some other Oceanic languages, but there is a syntactically significant contrast between direct and indirect possession, correlating roughly with alienable versus inalienable possession but not fully predictable on semantic grounds. Pearce argues for the very plausible claim that direct possessor nouns subcategorize for a possessor DP, which is a complement of N, whereas possessors of indirect possessor nouns are adjuncts and occupy a higher syntactic position. She also argues for iterative phrasal movement (‘roll-up’) as a way of accounting for the mirror-image ordering of adjectives and other elements within the DP, relative to Cinque’s model. But for Unua, this analysis works best under the assumption that all complements actually merge in specifier positions.

Mara Frascarelli, in ‘Intonation, information structure and the derivation of inverse VO languages’ (81–102), analyzes the prosodic and information structures of recorded sentences in Tagalog and Malagasy, and argues that the results obtained support the vP...