Publishing Partner: Cambridge University Press CUP Extra Publisher Login

New from Cambridge University Press!


Revitalizing Endangered Languages

Edited by Justyna Olko & Julia Sallabank

Revitalizing Endangered Languages "This guidebook provides ideas and strategies, as well as some background, to help with the effective revitalization of endangered languages. It covers a broad scope of themes including effective planning, benefits, wellbeing, economic aspects, attitudes and ideologies."

New from Wiley!


We Have a New Site!

With the help of your donations we have been making good progress on designing and launching our new website! Check it out at!
***We are still in our beta stages for the new site--if you have any feedback, be sure to let us know at***

Review of  Language and Text in the Austronesian World

Reviewer: Malcolm D. Ross
Book Title: Language and Text in the Austronesian World
Book Author: Yury Lander Alexander K. Ogloblin
Publisher: Lincom GmbH
Linguistic Field(s): General Linguistics
Historical Linguistics
Genetic Classification
Issue Number: 20.1943

Discuss this Review
Help on Posting
EDITORS: Lander, Yury A.; Ogloblin, Alexander K.
TITLE: Language and Text in the Austronesian World
SUBTITLE: Studies in Honour of Ülo Sirk
SERIES: LINCOM Studies in Austronesian Linguistics 06
PUBLISHER: Lincom Europa
YEAR: 2008

Malcolm Ross, Department of Linguistics, Research School of Pacific and Asian
Studies, The Australian National University

This book is a collection of papers written in honor of the Estonian linguist Dr
Ülo Sirk. It is divided into three parts, titled respectively Diachrony,
Synchrony and Text Studies. It is prefaced by a short paper by the editors ''On
Ülo Sirk and his work'' (1–4) and a bibliography of Sirk's works (4-7). Sirk's
academic interests form a set of concentric circles. In the center is Bugis
language and literature (Bugis is spoken in south Sulawesi). Sirk has produced
two grammars of literary Bugis, one in 1975, and the second in 1996 after he was
first able to visit Sulawesi. His interest in the languages of Indonesia forms a
somewhat larger circle, with work on Indonesian, on the grammar of western
Indonesian languages, on the typology of eastern Indonesian languages, and on
the historical phonology of the languages of Ambon. The largest circle of all is
the comparative and historical linguistics of the Austronesian language family.
reflected in a number of publications and, the bibliography tells us, a
forthcoming historical-comparative introduction to the Austronesian language
family written in Russian.

The twenty papers in the volume reflect these interests. Two concern Bugis
literature and five more have to do with Sulawesi languages. Of the other
thirteen nine are concerned with Indonesia, one with Malagasy, one with
Philippine languages and two with reconstructing the history of the Austronesian

Below I provide a short summary of the contents of each paper, in most cases
followed by an evaluation. At the end, I provide a short evaluation of the book
as a whole.

Part 1. Diachrony.
In ''On the classifiability of Malayic'' (pp 11-22) Alexander Adelaar provides a
survey of the evidence concerning the internal subgrouping of the Malayic
family, which includes the many varieties of Malay as well as a number of
languages located in Sumatra and Kalimantan. The paper is a critique of a 2004
paper in which the author of this review proposed that Malayic has two branches:
(what Adelaar calls) the Kanayatn subgroup and a group containing all other
Malayic varieties. I also suggested that inscriptional Old Malay did not belong
to Malayic proper. Adelaar presents evidence that both these claims are wrong,
as the features which I claimed to be Kanayatn innovations are largely
retentions from an earlier interstage which are also reflected in various other
Malay varieties and in Old Malay.

Adelaar's paper substantially updates the discussion of Malayic subgrouping in
Adelaar (1992) and complements his 2004 paper on the history of the movements of
Malayic speakers. His criticism of Ross (2004) is entirely justified. The latter
was written in the mid-1990s, when there was frustratingly little activity in
the historical linguistics of western Indo-Malaysia, but publication of the
volume in which it appeared was delayed until 2004. It was a mistake on my part
to let it be published, and its only virtue is that it has sparked Adelaar's
detailed and lucid update, which goes some way towards clearing muddied waters.
An important implication of Adelaar's paper is that there remains a great deal
of work to be done on the history of the Malayic languages (and for that matter
of many of the languages of western Indo-Malaysia).

Sergey Kullanda, ''Old Javanese Kinship Terminology: Some Historical-Typological
Implications'' (pp22-30), lists the terms for consanguineal kin that are found in
Old Javanese texts, observes that they are a formal set in that they are all
prefixed with /ra-/, and examines their usage, in particular their extended
senses, to draw a conclusion about Old Javanese social structure. This
conclusion, which depends to some extent on comparisons with other (not
Austronesian speaking) societies, are that Old Javanese society was viewed by
its members in terms of stratification by age. Both men and women were divided
into senior and junior groups on the basis of whether or not their children had
passed through initiation. Uninitiated children in their turn were referred to
without a gender distinction. Kullanda takes this conclusion as an argument for
the hypothesis that classificatory and individual kinship patterns are derived
from this stratification, rather than vice versa.

Kullanda does Austronesianists the favor of presenting Old Javanese kin terms
together in one place and surveying their usage in texts. A number of steps in
his arguments seem less than fully justified, however. Comparative arguments
based on other societies do not sit well with the reconstructive practices of
historical linguists, and his proposal that classificatory systems are derived
from a system of social stratification, rather than vice versa, is scarcely
supported by Austronesian kinship data. Cognates of the Old Javanese terms
appear in classificatory systems in many societies that lack the kind of
stratification Kullanda describes for Old Javanese. In this respect Old Javanese
is the exception rather than the rule. I was also troubled by Kullanda's
apparent assumption that Old Javanese immediately reflects Proto Austronesian.
On the widely accepted theory that Proto Austronesian was spoken in Taiwan, Old
Javanese appears many nodes below it in the family tree. Furthermore, it was
spoken by a society that had undergone a major transformation under Indic influence.

In his paper ''In Search of Middle Javanese'' (pp31-45) Alexander K. Ogloblin asks
whether the term 'Middle Javanese' denotes a language that is distinct from Old
and Modern Javanese. He concludes tentatively that it does but that the boundary
between Middle and Modern Javanese is fuzzy. As a rule of thumb Ogloblin applies
the term Middle Javanese to works written between the 14th and the late 17th
century, equating the rise of Modern Javanese with the appearance of Islamic and
Arabic influence on texts. He uses a corpus of texts from Old, (allegedly)
Middle and Modern Javanese to determine the frequency of certain phonological
and morphological features in the three literary languages and to decide whether
there is a distinct feature set that characterizes Middle Javanese. It is
difficult to arrive at a definitive answer because the various texts and text
genres display significant differences in frequencies, reflecting the fact that
writers have consciously modeled their styles on more archaic forms.

Ogloblin's quest for Middle Javanese is interesting, as it highlights the fact
that the languages of the texts are literary languages and that more recent
writers have often sought to imitate earlier styles. Because of this, the Old,
Middle and Modern Javanese periods overlap, with texts written in the language
of one period appearing at the same time as texts written in the previous one.
This not only makes the task of distinguishing between the stages (and
especially between Middle and Modern) difficult, but it also makes the paper
rather difficult to follow for the reader who has no knowledge of Javanese
literature. One thing is clear, however: the historical linguist who goes to
Javanese texts hoping to quickly find out something about the development of
spoken Javanese will be disappointed, as the relationship between dialects of
the spoken language and the various forms of the written language must have been
very complex indeed.

Andrew Pawley's ''Where and When Was Proto Oceanic Spoken? Linguistic and
Archaeological Evidence'' (pp47-71) is one of the longer papers in the volume and
answers the question in the title. Proto Oceanic was the ancestor of the large
subgroup of 500 or more Austronesian languages spoken in Melanesia, Polynesia
and Micronesia, i.e. in most of the islands of the Pacific from New Guinea
eastward. Pawley provides an up-to-date and quite detailed summary of the
current state of Oceanic studies. Linguistic matters he discusses include the
position of Oceanic in the Austronesian family tree, the boundary between
Oceanic and non-Oceanic Austronesian languages, the shared innovations which are
evidence for the Oceanic subgroup, and previous attempts to place the Proto
Oceanic homeland based on the internal subgrouping of Oceanic. Archaeological
issues include the equation of Proto Oceanic with the languages of the Lapita
culture, the emergence of the Lapita culture in the Bismarck Archipelago (to the
east and north of New Guinea) and its eastward spread to the Reefs/Santa Cruz
Group, Vanuatu, New Caledonia, Fiji, Tonga and Samoa, and the preference of
Lapita communities for small offshore islands. The archaeological evidence
places the break-up of Proto Oceanic sometime between 3400 and 3100 years ago.
Pawley suggests that this was preceded by a Pre-Oceanic period of 2-3 centuries
during which speakers were located somewhere in the islands along the north
coast of New Guinea, through which they must have traveled to reach the
Bismarcks, or in the Bismarcks themselves, where the break-up of Proto Oceanic

I must confess a personal interest in Pawley's paper, as I had a hand in various
ways in the discussions that led up to its writing, and it cites various
publications of mine, some of which provide evidence or arguments that feed into
Pawley's findings with regard to the Oceanic homeland and one of which he nicely
dissects. I find his conclusions the best account of the evidence that I have
encountered, and his dissection thus fully justified. One problem which he
solves well is the question of the origins of Western Oceanic languages (spoken
on the margins of mainland New Guinea, on New Britain, New Ireland, Bougainville
and the northwest Solomons). Pawley proposes that since the archaeology implies
that Proto Oceanic was spoken on small offshore islands, the Western Oceanic
languages reflect the gradual spread of Oceanic speakers, after the break-up of
Proto Oceanic, onto the larger islands just listed. His scenario neatly solves
the problem of how the Western Oceanic linkage arose, which I had previously
attributed to a post-Proto Oceanic dispersal of Western Oceanic dialects that
led to the extinction of earlier Oceanic dialects.

Ilia Peiros's short paper ''Malayic, Chamic and Aceh: Some Lexicostatistical
Remarks'' (73-88) is followed by an appendix (78-87) containing the
lexicostatistical data on which his lexicostatistical analysis is based. The
paper is both a defense of the Russian school of lexicostatistics and a critique
of the classical comparative method of subgrouping by shared innovations. Peiros
sets out his view of the family tree as a representation of relationships within
a family tree. A tree, he says, is built from two kinds of relation, between
ancestor and descendant and between sisters. Both are based on similarity, but
sister relations exclude similarities that are due to contact. An objective
method of measuring similarities provides the basis for ''an automatic procedure
of genetic classification''. This method is lexicostatistics. Peiros objects to
the use of shared innovations on the grounds that shared innovations are an
insufficient reason for establishing a node (and thus an interstage language) in
a tree, as they may be the outcome of contact. Instead, an interstage language
must be reconstructed on the basis of shared similarities in the modern
languages (more on this under ''Evaluation'' below). Finally, Peiros uses a
lexicostatistical analysis to propose that Aceh is more closely related to the
Malayic subgroup (cf Adelaar's paper above) than to the Chamic subgroup to which
it is usually assigned.

Peiros' defence of lexicostatistics and his critique of the comparative method
of shared innovations is so brief and programmatic that I am not quite certain
that I have understood it correctly. His insistence, however, that relatedness
should be based on inherited similarities turns the comparative method on its
head, and I am hard put to understand how, for example, the raft of innovations
which characterize the Oceanic subfamily of Austronesian (see Pawley's paper)
can be attributed to contact and dismissed as a basis for subgrouping and as a
starting point for the reconstruction of Proto Oceanic (which of course is also
based on shared similarities of widely distributed daughter languages). I am
also unable to understand how lexicostatistics can be regarded as the
recognition of shared inherited similarities: Peiros offers no response to the
standard criticism that lexicostatistics fails to distinguish between innovation
and retention, presumably because he considers retentions, not innovations, to
have probative value. He offers his lexicostatistical analysis of the
relationships of Acehnese to Malayic and Chamic with no comment on the reasons
that others have advanced for subgrouping Acehnese with Chamic or on Sidwell's
(2005, 2006) recent publications on this matter.

In ''Notes on the Historical Phonology and Classification of Wolio'' (89-113) René
van den Berg investigates the history of Wolio, spoken on the island of Buton
southeast of Sulawesi. The author describes the social history of Wolio and
cites published research which shows that Wolio does not belong to the
Muna-Buton group, which occupies the rest of Buton. He suggests that instead it
belongs to the Kaili-Pamona group, one of the microgroups of Sulawesi proper.
His main grounds for this are the Wolio reflexes of Proto Malayo- Polynesian
phonemes, and in the course of presenting these he also provides a fairly full
overview of Wolio historical phonology. The paper concludes with a very brief
look at other features that may suggest a Kaili-Pamona origin for Wolio and with
notes on the interaction of Wolio and Muna-Buton languages.

Van den Berg's paper requires little comment other than to say that it is a very
workmanlike survey of Wolio historical phonology and a fine contribution to the
growing literature on Sulawesi linguistic history. It is hard to fault the
author's conclusion that Wolio is a Kaili-Pamona language, and I look forward to
seeing answers to some of the questions mentioned towards the end of the paper.

John U. Wolff's ''The Reconstruction of the Proto-Austronesian Phoneme *g''
(115-128) is one of a series of papers the author has published at intervals
since 1974 (Wolff 1974, 1988, 1991, 1997) presenting his ongoing analysis of the
Proto Austronesian phoneme system. I write ''ongoing'', since his 1982 paper
denied the existence of *g. The reconstruction of Proto Austronesian has its
roots in the work of Dyen in a series of papers beginning in the late 1940s
(e.g. Dyen 1953) and of Dahl (1973), and has since bifurcated into the analyses
of Blust (refining Dahl and Dyen) and Wolff (a radical break with previous
analyses). Wolff's *g corresponds to initial *g and medial and final *j as
reconstructed by his predecessors and by Blust. According to Wolff, *g- and *-j-
are complementary and so he treats them as a single phoneme. Wolff first
discusses, then rejects, Sagart's recent (2004) suggestion that *j was a palatal
nasal. He presents the reflexes of (his) *g in a sample of Austronesian
languages and discusses the puzzle whereby *g- is reflected as /k-/ in the vast
majority, concluding that this change must already have occurred in some
dialects of Proto Austronesian. The reflexes of *-g- and *-g are many and
varied. Few languages retain /g/, and /d/, /z/ or a liquid are far more common.
Whatever the phonetics of *-g-/*-g, Wolff concludes that it must have undergone
change independently on many occasions.

Wolff's reconstruction of Proto Austronesian *g is hard to evaluate, as a proper
evaluation would have to consider the reconstruction of the whole system of
phonemes to which *g belongs, and this reconstruction is as yet unpublished,
although Wolff provides us with an inventory. The observation that *g-, *-j- and
*-j are in largely complementary distribution was made by the writer of this
review in 1992. The point made there, however, was that their complementarity
dated back to a pre-Proto Austronesian date and was imperfect in Proto
Austronesian itself. Wolff appears to overlook this observation. Wolff is right,
though, that the modern reflexes of his *g are puzzling. But his suggestion that
''some of the dialects which became MP [Malayo- Polynesian] languages still
retained the contrast [between *g and *k] and others had lost it'' is odd, as it
says that Proto Malayo- Polynesian was not a unitary language: I am not sure how
he would then account for the phonological innovations which are usually
considered to have occurred in a unitary Proto Malayo-Polynesian ancestral to
all Austronesian languages spoken outside Taiwan. I do agree with him that the
medial and final reflexes of his *g (others' *j) are something of a riddle.

Part 2. Synchrony
In their ''Word Order of Prepositional Phrases in Aralle-Tabulahan and Moronene''
(131-140) T. David Andersen & Robin McKenzie present data supporting a curious
construction in member languages of two Sulawesi microgroups. The languages are
Aralle-Tabulahan and Moronene, and the construction is one in which the object
of a locative preposition is modified by a demonstrative. In both languages the
demonstrative *precedes* the preposition, giving the sequence 'that in house',
i.e. the constituent 'that house' is discontinuous.

Andersen & McKenzie's analysis of 'that in house' is curious in that it offers
no discussion of the origins of this constituent-breaking construction. Their
own data offer a possible source. Aralle-Tabulahan locative and directional
phrases must begin with a locative or directional word. This is apparently the
head of the phrase, and the following (omissible) prepositional phrase is
apparently its dependent. Such constructions occur in both Austronesian and
Papuan languages of Maluku (Holton 2003) and the Bird's Head (Dol 2004) (and in
southern New Ireland: see Mosel 1982 on Tolai), and it seems possible that the
'that in house' construction reflects a reanalysis of the locative head as a
deictic modifying the noun governed by the preposition.

Two papers, one by Mikhail A. Chlenov and Svetlana F. Chlenova, the other by the
second author, describe Damar Batumerah (West Damar), a previously undescribed
language from a small rather isolated island in southwest Maluku. The first,
''The Damar Batumerah (West Damar Language) of South-Eastern Indonesia'' (141-162)
is a wordlist of 513 entries with a two-page introduction locating the language
and pointing out that although West Damar is typologically similar to its
neighbors, members of the SW Maluku subgroup, its reflexes of Proto Austronesian
phonemes show that it does not belong to this subgroup. Chlenova's ''Preliminary
Grammatical Notes on Damar Batumerah or West Damar, a Language of Southwest
Maluku'' (163-177) is a 9-page sketch with a 5-page appendix of additional examples.

These two papers on Damar Batumerah offer the briefest account of the language
but for their brevity are remarkably rich in data. Given that the languages of
Maluku are only now receiving detailed attention and that their history is still
little understood, these papers are a welcome addition to the literature and
fill one more gap in our knowledge.

Mark Donohue's ''Obligatory Incorporation and ''Have'' in Tukang Besi'' (179-197) is
both a description of a feature of this SE Sulawesi language (supplementing his
1999 grammar) and a typological excursion. Donohue shows that the Tukang Besi
'have' construction entails noun (more strictly, N') incorporation and is a
counterexample to Mithun's (1984) generalization that languages which have noun
incorporation will also have an alternative non-incorporating (transitive)
construction. It is, however, as Donohue observes, a counterexample only if we
take it that every noun-incorporating verb in a language will also occur in the
non-incorporating construction. Tukang Besi itself has both an incorporating and
a non-incorporating construction, but the verb 'have' is exceptional in not
allowing the non-incorporating construction. Donohue associates this with the
fact that 'have' is the only bivalent stative verb in the language.

Donohue's paper is a fine addition to the literature on Tukang Besi. One matter
which the author does not comment on is that across languages 'have' clauses
often entail a construction that is found nowhere else in the language or, if
'have' is a verb, this verb differs in its constructional behaviour from the
other verbs of the language. The Tukang Besi verb 'have' is just such a case.
Unfortunately the chapter is marred by a number of editorial glitches. Footnote
numbering starts at 20 instead of at 1. Italics are missing from some vernacular
morphemes (bottom of p180). In the phonetic transcription of (2) [na] occurs
instead of [te]. In (51) 'stone' occurs where 'knife' is expected.

Barbara and Timothy Friberg have worked on Konjo (a language of the South
Sulawesi group) for many years, and their paper on ''/-ka/, a Marginalized
Grammatical Morpheme in Konjo'' (199-207) is one of those mopping up exercises
that occurs when one has described the major structures of the language but is
inevitably left with a collection of oddments that fall outside one's core
description. In order to clear the way for an account of this /-ka/, the authors
list and briefly describe some (near-)homophones. They then list the occurrences
of enclitic /-ka/, which is attached obligatorily to some roots, optionally to
others. Words of which it is an obligatory part mean 'but regrettably', 'or' and
'lest'. Words to which it is optionally attached include the meanings 'anyway',
'possibly', 'but' and 'perhaps'. The Fribergs suggest that it has an aspectual
function, which they equate with contingency or possibility.

One wonders whether the Fribergs' goal of pinning down the function of the Konjo
enclitic /-ka/ is achievable. Of the eight uses they list, three are obligatory,
i.e. the combination of the root + /-ka/ has become fully lexicalized and /-ka/
has become part of the word in which it occurs. Of the five optional uses, some
also show signs of lexicalization in that the root + /-ka/ combination has a
distribution and function all of its own. This paper is an excellent example of
fine-grained description, but I suspect that any general conclusion about the
function of /-ka/ must be a conclusion about its meaning in the past ('contrary
to expectation'?) rather than in the present. The characterization of
'possibility' as aspectual, incidentally, seems odd. Its meaning falls under the
rubric of mood rather than of aspect.

In ''Functions of the Mori Bawah Indefinite Particle /ba/: Towards a Comparative
Study'' (209–232) David Mead examines uses of this particle and shows that its
uses include adverbial subordination (condition, recurrent contingency, reason,
purpose), conjunction ('or'), reactivated topic ('as for'), preceding the
wh-word in a question to give senses such as 'whatever?' or 'what ... anyway?',
an element in an indefinite quantifier construction, a complementiser, and a
deferential request marker. Mead also examines the uses of forms in related
Sulawesi languages and concludes that its original meaning was 'perhaps, maybe'
and that other uses reflect grammaticisation.

Mead's paper is an interesting comparative study. I agree entirely with the
conclusion that studies using careful definitions are essential to accurate
reconstructions of morphosyntax and its functions. However, I was a little taken
aback by some of the terminological choices in this paper. /ba/ is described in
the title as an 'indefinite particle'. Definiteness is conventionally associated
with noun phrases, but the uses of /ba/ have to do with the speaker's certainty
about a predication. 'Discourse connective' is used where 'subordinating
conjunction' would seem more appropriate. I am left with the impression that
Mead wishes to remain agnostic on matters that entail syntactic analysis.

Bernd Nothofer's ''E-mel sebagai bahan pengajaran'' (233-242), written in Standard
Indonesian, lists and exemplifies the major characteristics of emails written in
that language. These include phatic particles and phrases, interjections,
topicalisations, the use of non-standard elements (e.g. absence of verbal
prefixes), and code-switching between Indonesian and a regional language, the
choice of language often depending on the subject matter. Indonesian is used
when talking about technological, academic and religious issues, the regional
language when talking about emotions and personal or family issues. The author's
interest in this topic arises from the use of e-mails as teaching material, as
they provide a kind of written conversational style. Since e-mail is
increasingly replacing conventional letters, its incorporation into language
teaching is relevant. More than this, however, e-mails mirror spoken Indonesian
and thus represent a written language which provides insight into the spoken
language and brings students into contact with facets of the language that are
otherwise neglected in teaching materials.

Maria Polinsky's analysis of ''The Existential Construction in Malagasy''
(243-261) is conducted within a mainstream generative framework. The Malagasy
existential construction consists of the verb /-isy/ 'be' (appearing as /m-isy/
and /n-isy/ in the present and past respectively), followed immediately by the
obligatory (indefinite) existential theme, then an optional coda, e.g. 'are' +
'three rooms' + 'this house' ('there are three rooms in this house' or 'this
house has three rooms'), or 'were' + 'children' + 'at school' ('there were
children at school'). Polinsky argues (i) that the existential theme is the
object, since indefinite objects form a constituent with the verb in Malagasy
and cannot be separated from it (definite objects can be separated, however) and
(ii) that the coda may consist of the external argument (i.e. subject) and/or an
adjunct. Thus in 'are' + 'three rooms' + 'this house', the subject is 'this
house', and 'are three rooms' is the predicate, and the existential construction
here and elsewhere in Malagasy encodes a part-whole relationship.

If Mead appears to eschew syntax, Polinsky embraces it wholeheartedly. It is in
the nature of the generative framework that one assumes certain properties of
syntactic structure and uses them in the analysis of a given language. Thus
Polinsky assumes that the existential theme of the existential construction will
either be its subject or its object, and she uses morphosyntactic properties to
argue for the latter. Interestingly, the main property of the existential theme
is that it is inseparable from the verb. This is a property only of indefinite
objects in Malagasy (not of definite objects), and one is reminded of indefinite
oblique-marked undergoer argument of actor-voice verbs in certain Formosan
languages (e.g. Puyuma) and of object incorporation in Tukang Besi (Donohue,
this volume) and in many Oceanic languages. In each of these cases the verb is
usually analyzed as intransitive and therefore as lacking an object in any
surface-syntactic sense. One wonders whether the Malagasy verb + indefinite
object (and thence the existential construction) is amenable to a similar analysis.

In her ''Some Aspects of Relations between Deixis and Syntax in Philippine
Languages'' (263-276) Lina Shkarban examines the hypothesis that the much
discussed difficulty of distinguishing between nouns and verbs in Philippine
languages is related to the tendency for information structure to be mapped more
directly onto syntax than in many other languages. She shows that deictics
(which for her include both demonstratives and case-marking determiners) play an
important role in differentiating predicate and subject. In addition to well
known Tagalog facts she cites data from Pangasinan, in which different deictics
are used in subjects and predicates, and from Cebuano in which adverbial
demonstratives encode tense and participate in certain information structure
functions. The functional mismatches whereby verbs occur in subject noun phrases
and nouns in predicates are a direct result of the fact that the topic is
typically encoded as subject and the comment as predicate, so that if the topic
happens to be an event, a verbal subject occurs, and if the comment is a class
or a known entity a noun phrase predicate occurs.

Shkarban's paper gives a somewhat different perspective on Tagalog syntax from
the accustomed discussions of Philippine grammatical relations. I found the
connection between this mapping and the very interesting deictic patterns she
describes a little difficult to make, but her point appears to be that deictics
play (among other things) an important role in encoding predicate-subject
relations which would otherwise be potentially ambiguous because of the
occurrence of verbs in subjects and noun phrases as predicates. Although her
point about mapping is well taken, it seems a little overstated, as Tagalog
certainly uses other constructions as well as subject-predicate to encode
information structure (Kaufman 2005).

Hein Steinhauer examines ''Synchronic Metathesis and Apocope in Three
Austronesian Languages of the Timor Area'' (277-296), namely Helong, Dawanese and
Letinese. These synchronic phenomena may be viewed as phonologically conditioned
variants of a change which occurs when a word is not phrase-final, i.e. it marks
phrase-internal juncture. The variants include loss of final glottal stop, loss
of final vowel, and metathesis entailing the word-final open syllable. In Helong
and Dawanese this syllable undergoes metathesis such that the onset of the
syllable becomes its coda. Letinese has two metathesising variants. One is the
same as in Helong and Dawanese. In the other, the word- final vowel and the
initial consonant of the following word undergo metathesis, e.g. /vatu masa/
becomes /vat muasa/. The paper concludes with a brief discussion of whether the
change reflects shared inheritance or is an areal feature, a question which
remains unresolved.

Steinhauer's work is intriguing, the more so as phonologically and functionally
similar metathesis occurs in an Oceanic Austronesian language, Rotuman (Schmidt
2003). Phonologically, but not functionally, similar metathesis is also found in
Ririo and in Kwara'ae, also Oceanic languages, but closely related neither to
each other nor to Rotuman. Oceanic metathesis has already spawned several papers
by theoretical phonologists, and as research by Steinhauer and his colleagues
continues on Timor area languages I assume that these will also become objects
of theoretical discussion.

Part 3. Text studies
In ''Form Criticism and Its Applicability to Bugis Historical Texts'' (299-310)
Ian Caldwell applies the form critical method, developed by Biblical scholars,
to the early part of a Bugis historical text. By analyzing its form, the form
critic seeks to identify each constituent part of a text and its literary
origins. Caldwell analyzes the first four pericopes (more or less independent
units) of the /Mula ritimpaqna Sidénréng/. The Bugis text of the pericopes and
their translation takes up just over four pages of the paper. These deal with
the pre-Islamic history of the kingdom of Sidénréng, i.e. the period before
1609, and on formal grounds each evidently has a separate origin in an oral
tradition which functioned to legitimize political and economic relationships.
The author shows how form criticism feeds into historical criticism: although
the pericopes are not historical narratives in the modern sense, they contain
historical assertions which no present-day Bugis would make and probably reflect
memories of the period after 1300.

Entitled ''Ktunu: Clues in the Quest of the Sailfish: Linguistic Insights in
Southwest Malukan Narratives (East-Indonesia)'' (311–325) Aone van Engelenhoven's
contribution is concerned with southwest Maluku oral traditions. These differ
from oral narratives elsewhere in Indonesia because they are prose rather than
epic poetry. This means that the mnemonic devices that help the teller to
recreate the story at each telling are not rhythmic patterns and lexical
parallelisms but /ktunu/, clichés or proverbial sayings which serve as mnemonic
triggers for each chunk of the narrative. A /ktunu/ may itself entail a lexical
parallelism, but it functions somewhat differently from poetic parallelism. Van
Engelenhoven shows that /ktunu/ composed of common nouns are governed by
phonological rules. Many /ktunu/, however, are place names or personal names,
each half composed of morphemes that in themselves recall a story for both
storyteller and audience. Their inclusion helps to pin down the story and its
extended meanings and contribute to a situation (common in oral narrative
traditions) whereby there is little room for the storyteller's individual input.

As Van Engelenhoven mentions, his paper takes up a theme introduced into the
anthropology of East Nusantara by Fox (1988), namely the use of lexical pairings
in stylized speech. Perhaps the most interesting feature that Van Engelenhoven
describes is the fact that in SW Maluku many of these pairings are proper names
which themselves tell a story, the audience's awareness of which is crucial to
their full understanding of the narrative and its implications.

In his ''/Sureq/ versus /lontaraq/: The Great Divide?'' (327–338) Sirtjo Koolhof
examines the now conventional use of these two terms to label Bugis written
genres, and finds that they reflect a division on a Western model into
'literature' (in the sense of /belles lettres/) and other writings which dates
from as recently as the 1970s. The division has no basis in Bugis tradition.
Koolhof instead identifies certain genres which do have a traditional basis: the
poetic /La Galigo/ texts, the /toloq/ (heroic poems with fictional or semi-
fictional content), and /attoriolong/, which ''are essentially an elaboration of
a genealogical core''. The /toloq/ potentially marked the beginnings of modern
authored literature, but, as Koolhof laments, this potential was not fulfilled,
and Bugis literature is now a museum piece.

The editors of almost any modern linguistic festschrift face a problem, namely
how to ensure sufficient thematic unity among contributions to satisfy a
publisher that the book will sell and yet to avoid excluding thereby too many
scholars who might otherwise have wished to contribute. The editors of this
volume have been strikingly successful in balancing these goals. It is a varied
and interesting collection of papers, yet, as I noted above, it displays a unity
centered on Ülo Sirk's interests. Every linguist with Indonesian interests will
want to have this book in their library, and I suspect that most
Austronesianists will too.

The book is fairly free of typographical errors, apart from those mentioned
above in association with Donohue's paper, plus 'tact' for 'tack' and
'restricing' for 'restricting' in Mead's paper, and the shocking typo
'Conclsuions' in the large bold-printed title at the top of p. 125. A volume of
iv + 338 pages needs a good binding, and this one has a strong hardback binding.
My one complaint is that some readers will have trouble with the
smaller-than-usual size of the type.

Adelaar, K. Alexander. 1992. _Proto Malayic: the reconstruction of its phonology
and parts of its lexicon and morphology_. Canberra: Pacific Linguistics.

Adellar, K. Alexander. 2004. Where does Malay come from? Twenty years of
discussions about homeland, migrations and classifications. _Bijdragen tot de
Taal-, Land- en Volkenkunde_ 160: 1–30.

Dahl, Otto Christian. 1973. _Proto-Austronesian_. Oslo: Studentlitteratur.

Dol, Philomena. 2004. _A Grammar of Maybrat: A language of the Bird’s Head
Peninsula, Papua Province, Indonesia_. Canberra: Pacific Linguistics.

Donohue, Mark. 1999. _The Tukang Besi language_. Berlin: Mouton de Gruyter.

Dyen, Isidore. 1953. Dempwolffa’s R. _Language_ 29: 131–137.

Fox, James J., ed. 1988. _To speak in pairs: Essays on ritual languages of
Eastern Indonesia_. Cambridge: Cambridge University Press.

Kaufman, Daniel. 2005. Aspects of pragmatic focus in Tagalog. In Wayan Arka &
Malcolm Ross, eds, _The many faces of Austronesian voice systems: some new
empirical studies_. Canberra: Pacific Linguistics.

Mithun, Marianne. 1984. The evolution of noun incorporation. _Language_ 60: 847–894.

Mosel, Ulrike. 1982. Local deixis in Tolai. In Jürgen Weissenborn & Wolfgang
Klein, eds, _Here and there: crosslinguistic studies on deixis and
demonstration_, 111–132. Amsterdam: John Benjamins.

Ross, Malcolm. 1992. The sound of Proto-Austronesian: an outsider’s view of the
Formosan evidence. _Oceanic Linguistics_ 31: 23–64.

Ross, Malcolm. 2004. Notes on the prehistory and internal subgrouping of
Malayic. In _Papers in Austronesian subgrouping and dialectology_, 97–109.
Canberra: Pacific Linguistics.

Sagart, Laurent. 2004. The higher phylogeny of Austronesian and the position of
Tai-Kadai. _Oceanic Linguistics_ 43: 411–444.

Schmidt, Hans. 2003. Temathesis in Rotuman. In John Lynch, ed., _Issues in
Austronesian historical phonology_y, 175–207. Canberra: Pacific Linguistics.

Sidwell, Paul. 2005. Acehnese and the Aceh-Chamic Language Family. In Anthony
Grant & Paul Sidwell, eds, _Chamic and Beyond: Studies in mainland Austronesian
languages_. Canberra: Pacific Linguistics. (Pacific Linguistics 569).

Sidwell, Paul. 2006. Dating the separation of Acehnese and Chamic by
etymological analysis of the Aceh-Chamic lexicon. _Mon-Khmer Studies: A Journal
of Southeast Asian Languages and Cultures_ 36: 187–206.

Wolff, John U. 1974. Proto-Austronesian *R and *d. _Oceanic Linguistics_ 13: 77–122.

Wolff, John U. 1988. The PAN consonant system. In Richard McGinn & Richard
McGinn, eds, The PAN consonant system, 125–147. Athens, Ohio: Center for
Southeast Asian Studies, Ohio University Center for International Studies.

Wolff, John U. 1991. The Proto-Austronesian phoneme *t and the grouping of the
Austronesian languages. In Robert A. Blust, ed., _Currents in Pacific
Linguistics: papers on Austronesian languages and ethnolinguistics in honour of
George W. Grace_, 535–549. Canberra: Pacific Linguistics.

Wolff, John U. 1997. The Proto-Austronesian voiced apical and palatal stops *d
and *j. In Cecilia Odé & Wim Stokhof, eds, _Proceedings of the Seventh
International Conference on Austronesian Linguistics_, 581–601. Leiden: Dept of
Languages, Cultures of South East Asia, & Oceania, Leiden University.

Malcolm Ross is Emeritus Professor of Linguistics in the Research School of
Pacific and Asian Studies at the Australian National University. His research
interests include the history of Austronesian and Papuan languages and language