Publishing Partner: Cambridge University Press CUP Extra Publisher Login
amazon logo
More Info

New from Oxford University Press!


Cognitive Literary Science

Edited by Michael Burke and Emily T. Troscianko

Cognitive Literary Science "Brings together researchers in cognitive-scientific fields and with literary backgrounds for a comprehensive look at cognition and literature."

New from Cambridge University Press!


Intonation and Prosodic Structure

By Caroline Féry

Intonation and Prosodic Structure "provides a state-of-the-art survey of intonation and prosodic structure."

Review of  Complex Predicates in Oceanic Languages

Reviewer: Kalyanamalini Sahoo
Book Title: Complex Predicates in Oceanic Languages
Book Author: Isabelle Bril Francoise Ozanne-Rivierre
Publisher: De Gruyter Mouton
Linguistic Field(s): Applied Linguistics
Historical Linguistics
Cognitive Science
Issue Number: 15.3139

Discuss this Review
Help on Posting

Date: Mon, 8 Nov 2004 03:36:43 -0800 (PST)
From: Kalyanamalini Sahoo
Subject: Complex Predicates in Oceanic Languages

EDITORS: Bril, Isabelle; Ozanne-Rivierre, Françoise
TITLE: Complex Predicates in Oceanic Languages
SUBTITLE: Studies in the Dynamics of Binding and Boundness
SERIES: Empirical Approaches to Language Typology 29
PUBLISHER: Mouton de Gruyter
YEAR: 2004

Kalyanamalini Sahoo, unaffiliated scholar


This book contains a collection of 14 articles. Along with a short
abstract, each article typically starts with an introduction and with
proper sections and subsections, ends with a conclusion/summary, a
list of orthographic conventions and abbreviations of the linguistic
terminologies used in the article, and notes. At the end of the volume,
there is a common bibliography for all the articles and an index of
languages referred to in the articles.

The volume is a contribution to the topic of complex predicates in
Oceanic languages. After a short introduction by the editors regarding
the organization and objectives of the book, the volume starts with
Isabelle Bril's article, which is an introduction to the volume. It
synthesizes the main data and findings of this volume from a
theoretical and typological perspective. It is followed by Gunter Senft's
broad overview of the research history and data on the topic of serial
verbs in various Austronesian and non-Austronesian Papuan
languages. Then come several detailed case-studies of complex
predicates in various Oceanic languages focusing both on synchronic
and diachronic factors.

SUMMARY Isabelle Bril's article provides an excellent introduction to
typological study of complex predicates in Oceanic languages. It
addresses a broad spectrum of phenomena on complex predicates
irrespective of the great variety of structural patterns and language
specific-parameters and syntactic constraints. It presents the facts on
Complex predicates discussed in the articles of this volume, uses data
from numerous different languages to support the claims, and
provides cross-linguistic comparisons.

Gunter Senft's article presents different types of Serial Verb
Constructions (SVCs) found in Austronesian and Papuan languages
and gives a general account of verb serialization in terms of features,
types, grammatical functions, syntactic descriptions, semantic roles,
order of verbs etc.. It discusses event conceptualization in SVCs and
argues for a cognitive approach rather than a purely syntactic
approach to the analysis of verb serialization.

Nuclear-layer serialization and core-layer serialization, out of the two
types of serialization found in the Oceanic languages, nuclear-layer
serialization is well attested in Saliba. However, Anna Margets claims
that core-layer serialization is also found in Saliba, although restricted
to a few constructions, like constructions expressing manner of
motion, followed by path-encoding motion. These constructions differ
from a free sequence of verbs by a number of formal criteria including
word order and formation of the sequence by a closed class of verbs.

Considering serial and complex verb constructions in Teop, Jessika
Reinig claims that the two-verb sequence SVCs typically operate on
the nuclear layer. The three types of SVCs found in this language are:
(i)[an in/transitive verb that implies a kind of motion + an intransitive
direction verb], the sequence specifies the direction of the first verb. (ii)
a sequence of two verbs that expresses either two states of affairs
that happen simultaneously, or a purposive action. In such sequences,
only the last verb can be transitive and the transitivity of the complex
is determined by the transitivity of this verb. (iii) a directional proclitic
that only appears in combination with another verb and the transitivity
in such sequences is determined by the transitivity of the second verb.
Such type of sequences seem to be cases of specialization and
grammaticalization varying with position.

In Mwotlap language, most VPs consisting of two or more verb roots
chained together , function like a single verb. In such
sequences, V1 functions as the head of the sequence and V2 as the
modifier of the head. The sequence can refer only to a single action,
hence, the valencies of both the verbs merge into that of the whole
macro-verb. Alexandre François investigates such sequences and
comes to the conclusion that calling such sequences 'serial verbs'
would mislead linguistic analysis as the V2 in such sequences is really
an adjunct. And if this adjunct is a binary predicate, it influences the
syntactic behaviour of the macro verb more than that of an unary
predicate. It is interesting to note that some verbs, along with other
word classes, take part in this mechanism and the adjunct has a major
syntactic function in the clause.

Studying Anejom, the Southern Vanuatu subgroup of Oceanic
languages, John Lynch claims that verb serialization has disappeared
from Anejom because of three factors:

1) the development of an 'echo-subject' construction, in which the
Proto Oceanic conjunction /*ma/ 'and' has developed as a preverbal
clitic (possibly only with directional verbs) indicating that the subject of
the clause to which it is cliticized is the same as the subject of the
previous clause.

2) Phonological reduction, which converted two serialized (non-
directional) verbs into one compound verb.

3) The development of directional suffixes from /m/-marked directional

He concludes that once these processes got firmly established in the
language, serialization disappeared and both echo-subject marking
and compounding were used extensively.

Considering various types of serial and complex verbs in Nêlêmwa,
Isabelle Bril shows that complex verbs are mainly of three types
involving various types of verb and different hierarchies: (i) time-
iconic, co-ranking active verbs; (ii) asymmetric verbs expressing
adverbial modification or aspectual and modal specification; (iii) semi-
grammaticalized verb strings with argument-expanding or conjunctive
functions. Contrary to Foley & Olson's (1985) generalization
that "nuclear layer" serialization is mostly found in SOV languages,
while SVO languages have predominant core layer serialization, Bril
shows that "nuclear layer" serial constructions [(S) V V(O)] are also
found in Nêlêmwa. She discusses various syntactic functions of SVCs
in Nêlêmwa, assesses the syntactic, semantic and discourse
differences between complex verbs and syndetic or asyndetic
coordinate or subordinate VPs and clauses.

Studying complex predicates in East Uvean language, Claire Moyse-
Faurie comes to the conclusion that the strictly defined
SVCs 'denoting one event and including simultaneous actions' are
very rare in this language. The two main reasons behind this are: (i) in
a sequence of two verbs, the V1 usually tends to be either a modal or
aspectual marker, while the V2 tends to be an adjunct of manner. So,
such constructions stand somewhere between modifying serialization,
or grammaticalization, or lexicalization. (ii) coordinate structures
referring to quasi simultaneous events and complementation
structures referring to successive actions of a single event are
preferred over SVCs.

Considering complex verbal sequences in Pileni, Åshild Næss shows
that because of argument ellipsis and lack of subordination markers,
in many instances it is difficult to distinguish SVCs from some other
type of complex or conjoined clauses. However, in the case of SVCs,
serialization occurs both at the level of the nucleus and of the core.
Core-layer SVCs must have the same temporal frame but may differ in
aspect, while nuclear-layer SVCs must share both tense and aspect.
At the core-layer serialization, the difference between "contiguous"
and "non-contiguous" serialization is exploited to achieve two different
functions with the same V1.

Investigating Tahitian two-lexeme strings 'X Y' referring to a single
process, Mirose Paia and Jacques Vernaudon claim that it is a
particular case of qualitative modification, following the order 'modified-
modifier'. This procedure is used both in predicate and argument

Ulrike Mosel has shown that complex predicates in Samoan consist of
two or more contiguous lexical words. Interestingly, in this language,
lexical categories like nouns, adjectives and verbs cannot be
distinguished by morphological or distributional criteria of
classification; that is, noun phrases and verb complexes cannot be
defined as projections of nouns and verbs. Consequently,
constructions of juxtaposed content words are not defined by morpho-
syntactic characteristics, rather they are established by the lexical
features of the content words involved. Beside the predicative
behaviour of the juxtapositional constructions, they can function as
either nucleus or modifier of noun phrases and verb phrases, as
components of superordinate juxtapositional constructions. Such
multifunctionality of juxtapositional constructions within noun phrases
and verb complexes shows that they form independent syntactic units
below phrase level.

Studying modern Oceanic languages, Malcolm Ross has shown that
directional verbs in SVCs have been grammaticized in three different
ways: as directionals, as pre-verbal clitics, and as relators /
prepositions, and these grammaticization paths have to certain extent
been constrained by the structure of the SVCs. A pre-verbal clitic
arises only from a sequential SVC, but a directional SVC may give rise
to either a directional or a relator/preposition (although this latter
choice is not conditioned by the SVC). As directional SVCs are
frequently used in discourse, they are more grammaticized than the
sequential SVCs. Not only transitives, but also intransitive verbs may
be grammaticized as prepositions.

Investigating the evolution of the verb 'take' in New Caledonian
languages, Françoise Ozanne-Rivierre has shown that both,
lexicalization and grammaticalization, are at work in this
language. 'take', which usually occurs in the V2 position in a sequence
of two verbs [V1 V2], has been fully lexicalized in all the languages of
the group, and in some languages, it is gradually assimilated in
transitive compound verbs. It has been grammaticalized as the object
case marker in serial constructions in several language families
including Mandarin. In Nyelâyu, this verb is further delexified and the
transitive suffix /-va/ in causatives is derived from the cliticized form of
the verb /Pha/ 'take'. Contrary to Lord's (1993:96) analysis of this
verb 'take' in Mandarin, where 'take' has been analyzed as a transitive
construction marker and has been grammaticalized as a syntactic
object marker; in New Caledonia, Ozanne-Rivierre claims that 'take'
has evolved into an applicative transitivizing morpheme, and instead
of using a reflex of the Proto Oceanic applicative transitivizing suffix,
New Caledonian languages use a semantically transparent transitive
verb 'take, carry' for associative case-marking.

Considering the [V V] sequences in New Caledonian languages,
Françoise Ozanne-Rivierre & Jean-Claude Rivierre show that these
serial verbs have evolved into verbal compounds. In the languages of
the Mainland of New Caledonia, the "classifying" verbal prefixes
derive from compound constructions in which the first verb is reduced
to its first syllable (or first mora). Unlike the languages of the north, the
proliferation of monomoraic lexical verbal prefixes is characteristic of
Southern languages which have open syllables and a strong tendency
to monosyllabism. Development of "classifiers" and co- lexicalization of
two (or more) verb-stems to create more complex verbal concepts
contribute to the lexical renewal in Southern languages. Thus, given
the highly variable productivity of verbal prefixes in Kanak verb
compounds, the term "lexical prefixes", used by Nojima (1996) for
Bunum, seems more appropriate for Kanak languages than that
of "classificatory prefixes" which is traditionally used in Oceanic


This is an excellent reference book. Anyone interested in serial verbs,
complex predicates, grammaticalization, language typology, Oceanic
languages, etc., would benefit from the huge collection of data and
cross-linguistic comparisons presented in the book. The data
presented in the book come from a broad range of languages
constituting various subgroups of Oceanic languages. It presents
syntactic, lexical, and to certain extent semantic and phonological
information on complex predicates, which has important theoretical
implications. There is a clear effort on the part of many authors to
discuss previous analyses. The presentation presumes no specialist
knowledge of Oceanic languages and is very friendly towards those
with no experience in these languages. In short, it is a great

The fourteen papers included in this volume constitute a valuable
contribution to research on complex predicates as well as on Oceanic
languages. As concerns the complex predicates, a great variety of
structural types (i.e.types such as nuclear and core-layer serialization,
having two main structural subtypes such as "symmetrical" (co-
ranking), and "asymmetrical" (implying head-modifier hierarchy)) have
been discussed. A number of issues concerning characteristics of
serial verbs (e.g. prosody, shared tense-aspect marker and polarity
markers, finiteness vs nonfiniteness), ordering principles (iconic
ordering vs parametric settings), different approaches (syntactic vs
lexicalist approach) are addressed. The volume explores several
evolutionary paths on different levels: degrees of grammaticalization
(into operators, case-markers, conjunctions) and morphologization at
the morphosyntactic level, and degrees of co-lexicalization and
compounding at word level.

However, there are certain shortcomings. Although the editors in
the 'introduction' (p-x)to the volume, state that the label 'complex
predicate' has been used as a cover-term for both complex predicate
and SVCs, Bril distinguishes serial constructions from Complex
predicates in terms of monoclausality (p-25), and other authors
(François, Reinig, Næss) of the volume distinguish between the two
terms as well. So, a different cover-term would have been a better
choice. Also, the subtitle of the book in terms of 'binding and
boundness' is very little justified, except for Bril's article.


Foley, William A. & Mike Olson (1985) "Clausehood and verb
serialization", in: Johanna Nichols & Anthony C.Woodbury (eds),
Grammar inside and outside the clause. 17-60. Cambridge:
Cambridge University Press.

Lord, Carol (1993) Historical Change in Serial Verb Constructions.
Amsterdam-Philadelphia: John Benjamins.

Nojima, Motoyasu (1996) "Lexical Prefixes of Bunum Verbs". Gengo
Kenkyu 110: 1-27.


Kalyanamalini Sahoo, PhD. in Linguistics, Norwegian University of
Science & Technology, Trondheim. She has extensively worked on
serial verbs and complex predicates, with a number of publications
including her Ph.D thesis on complex verb constructions in Oriya (an
Indo-Aryan language). Her research interests include syntax,
typology, and computational morphology.

Amazon Store: