In lieu of an abstract, here is a brief excerpt of the content:

Reviewed by:
  • Balinese morphosyntax: A lexical-functional approach
  • Paul Kroeger
I Wayan Arka . 2003. Balinese morphosyntax: A lexical-functional approach. PL 547. Canberra: Pacific Linguistics, xvii + 270 pp. ISBN 0-85883-537-1. Aus$59.00, paper.


This is an important book. It is clearly indispensable for anyone interested in Balinese grammar, but beyond that, it makes major contributions to our understanding of Austronesian syntax and the typology of voice systems in general.

Arka's key claim, and the unifying theme of the book, is that the alternation between Agentive Voice (AV) vs. Objective Voice (OV) is "symmetrical" in the sense of Foley (1998) and Ross (2002). This term is used to indicate that, while the alternation does involve a change of grammatical subject, neither of the voice categories is derived via demotion of a core argument; that is, both AV and OV clauses are fully transitive.

The concept of "symmetrical voice" is not included in most typologies of voice systems, and poses a challenge for many formal syntactic frameworks. Foley coined the term in the context of his discussion of Philippine languages, Tagalog in particular; but as Ross points out (2002:24), its applicability to Tagalog (and other Philippine languages) remains controversial. Arka has provided the clearest, most detailed, and most convincing description to date of a symmetrical voice language, thus demonstrating that the pattern is a real one that both typologists and theoreticians need to account for.2

Arka has also (in this book and in other recent work, e.g., Arka 2000, 2003) given the concept a much firmer empirical basis by clarifying what syntactic "demotion" means. The central claim of the symmetrical voice analysis is that both AV and OV clauses contain two (or more) TERMS or CORE ARGUMENTS (i.e., "direct core arguments" in the terminology of Role and Reference Grammar). Terms are essentially the subject and object(s) of the clause; nonterms would include oblique arguments (such as PP recipients in English) and adjuncts. Familiar voice alternations work by demoting [End Page 306] a term argument to the status of an oblique argument or adjunct: passives by demoting the agent of a transitive clause, and antipassives by demoting the patient.

In addition to identifying a list of syntactic properties in Balinese that uniquely identify the subject of a clause, Arka identifies a second set of properties that distinguish terms from nonterms. These properties, which are discussed in greater detail below, provide a strong empirical basis for claims about termhood and demotion, a basis that has been lacking in much previous discussion of these issues. For example, under the "ergative" analysis of Philippine languages (Gerdts 1980, 1988; Payne 1982; Brainard 1994, 1996), the AV is identified as an "antipassive." The key issue here is the syntactic status of the AV patient: is it a term (as claimed under the symmetrical voice analysis) or an oblique argument (as claimed under the ergative analysis)? The question has not been decisively resolved, largely because very few termhood properties have been identified for these languages.3 While the relevant tests in Balinese cannot be adapted directly into (for example) Tagalog, Arka's work gives us a model for the kind of argument that needs to be made to decide such questions.

Arka describes his book as a slightly revised version of his 1998 PhD dissertation at the University of Sydney. That work was one of a trio of excellent dissertations by native speakers of Balinese studying at Australian universities, the other two being Artawa (1994) and Pastika (1999). Together with Clynes (1995), they provided a remarkable advance in our understanding of Balinese within a brief span of five years.

Arka adopts Lexical-Functional Grammar (LFG) as the framework for his study; but in one sense, that choice does not seem particularly crucial. Aside from the material in chapter 5 (see below), almost none of the discussion is focused on theory-internal issues or couched in theory-specific notations; linguists of any theoretical persuasion will find the discussion easy to follow, for the most part. What is crucial is that LFG, a nontransformational version of generative grammar, does not define grammatical relations like subject and object in terms of...


Additional Information

Print ISSN
pp. 306-313
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.