SIL Electronic Book Reviews 2003-022
URL: http://www.sil.org/silebr/silebr2003-022
The syntax of verb initial languages
Oxford Studies in Comparative Syntax. New York: Oxford University Press, 2000. Pp. . cloth $55.00, paper $30.00. ISBN 019513222X (hardback), 0195132238 (paper).
Reviewed by Helga Schroeder
SIL International and University of Nairobi
The twelve articles in this book contribute to the ongoing research on how to explain VSO word order within the framework of Generative Grammar.
The editors’ introduction of the book presents a fairly detailed overview of nearly all the existing derivational approaches that have been proposed for verb-initial languages. This chapter is very valuable for those who want to get a thorough orientation on the subject.
The first three articles (Randall Hendrick, “Celtic Initials”; Andrew Carnie, Heidi Harley and Elizabeth Pyatt, “VSO Order as Raising out of the IP? Some Evidence from Old Irish”; and Eithne Guilfoyle, “Tense and N-features in Irish”) discuss verb raising in Irish. The first article shows that the verb raising is determined by the Celtic verb-initial particles which cause the verb to raise to inflection (INF) and not to the complement (CP)—a view of Irish word order held by other linguists as well. The controversy on verb raising either to INF or CP is discussed and explained as a development from Old Irish (second article), where conjunct particles suggest a filled complement (C). The third article extends the discussion to language typology. The placement of the subject, as initiators (movement to the Tense Projection (TP)) or noninitiators of events (movement to the agreement phrase (AGRSP)) manifests the difference between VSO or SVO languages. This approach, however needs more testing.
The fourth article, by Edit Doron, “VSO and Left-conjunct Agreement: Biblical Hebrew vs. Modern Hebrew,” suggests that in Modern Hebrew the subject never moves out of the VP, due to the left-most conjunct agreement. The rationale is simple: as only the left-most conjunct in coordinated male and female constituents is marked on the verb, the subject cannot be raised to TP in such constructions, and thus in Modern Hebrew tense (T) does not have any Extended Projection Principle (EPP) features, an approach that has also been tested for Modern Irish, Spanish, and Standard Arabic, and which leads to the analysis that the EPP must rather be a lexical feature.
Diane Massam's article “VSO and VOS: Aspects of Niuean Word Order,” tackles another controversial issue: the difference between verb raising versus VP raising. The existential VP that has a copula and a nominal is taken as a basis for VP raising. In case the VP has a complement object, the object first moves out of the VP so that the VP carries its trace to the front of the clause to an extra projection. The conventional projections like IP or C cannot be used, because negation moves between the C and the verb and thus blocks the moving to C. VP raising is a popular approach for verb-initial languages and can be supported by various language data from around the world.
Andea Rackowski and Lisa Travis, in their contribution “V-initial languages: X or XP Movement and Adverbial Placement” argue for VP movement, as adverbials occur in Malagasy and Niuean postverbally and force VP movement, as otherwise the verb complement relationship would be split.
Felicia Lee (“VP Remnant Movement and VSO in Quiviani Zapotec”) observes that in Quiviani Zapotec, a Mexican language, the VP shares the same distributional constraints as NPs. Lee therefore concludes that the VP must undergo the same syntactic processes as NPs and initiates VP movement rather than verb movement.
As interesting and valid as the VP-raising approach is, it does not necessarily justify the creation of a new parameter in language typology. For example, Andrea Rackowski, Lisa Travis and Diane Massam see the difference between SVO and VSO languages in the movement of the NP or the VP, respectively: in SVO languages the NP moves, whereas in VSO the predicate moves. Thus, according to Rackowski and Travis, SVO languages are argument-fronting while VSO languages are predicate-fronting.
On the other hand, two articles, “Locus Operandi” by Ray Freeze and Carol Georgopoulos, and “Animacy Hierarchies and Sentence Processing” by Seth Minkoff, treat the difference between VSO and SVO languages not as a matter of verb movement, but as a matter of locative phrases. For Freeze and Georgopoulos the various categories of locative phrases, using data from languages all over the world, serve as basis for a word order typology of SVO and VSO languages. Minkoff employs the concept of animacy as a criterion for word order differences, arguing that the subject of any transitive verb must be at least as animate as its object.
The last article, by Eloise Jelinek, “Predicate Raising in Lummi, Straits Salish,” is unique in that it tackles the word order problems in polysynthetic languages that do not have any VP. All polysynthetic languages have incorporated arguments; thus, the Phonetic Form (PF) and the Logical Form (LF) coalesce, and there is a direct link between argument and information structure.
The book contributes to the linguistic community at large, as it shows a variety of verb-initial languages and different approaches to the problem. It also contributes to the research of Universal Grammar as it tests and verifies different approaches to the derivation of verb-initial languages.