SIXTH INTERNATIONAL CONFERENCE ON OCEANIC LINGUISTICS

 

4-9 July,

University of the South Pacific – Emalus Campus

Port Vila, Vanuatu

 

 

 

ABSTRACTS

 

 

 

Relative clauses where the head noun is coreferential to a possessor in Hawaiian and Hawai’i Creole English

 

Jason D. Cabral

University of Hawai’i at Hilo

 

Relative clauses where a head noun is coreferential with a possessor in the modifying clause have been overlooked in Hawaiian. Examples of the forms which such relative clauses take in texts will be presented along with their similarity to parallel structures in Hawai’i Creole English. The dates at which the Hawaiian texts were produced indicate that the Hawaiian structure existed before the Hawai’i Creole English structure developed, providing the possibility of Hawaiian substrata influence in this instance.

 

 

Extensions of the term for ‘family house’ in a non-house society

 

Ann Chowning

University of Auckland

 

In recent years a number of scholars, following Blust (1980), have postulated that AN-speakers often extended the term for ‘family house’ to a social group, and suggested that such speakers possessed “house societies”. Although Pawley and Green (1998) state that “the evidence for such figurative uses of *Rumaq seem to be lacking in Oceania”, Kirch suggested in 1997 that “for early colonizers of the Bismarcks, the *Rumaq was probably not just a physical dwelling, but also the social unit with which they affiliated” (see also Kirch and Green 2001). It seems worthwhile to describe the extensions of luma ‘family house’ in Lakalai of New Britain, which do not point towards a “house society”.

 

On the unity (or not) of North and Central Vanuatu

 

Ross Clark

University of Auckland

 

Is there a subgroup of Oceanic consisting of just the non-Polynesian languages of north and central Vanuatu, from Efate to Hiw? This paper will review a century of surprisingly diverse answers to this question, from Ray to Lynch, and present some evidence in support of unity, with qualifications.

 

 

 

Protovariability and parallel development: Reduced grammatical forms in Melanesian Pidgin

 

Terry Crowley

University of Waikato

 

The three major national varieties of Melanesian Pidgin (Tok Pisin, Solomons Pijin, Bislama)  all have pervasive transitive marking on verbs by means of a suffix that is generally described in published descriptions as being of the basic shape -Vm. In Bislama, however, the final vowel of the suffix is fairly commonly lost in a variety of structural environments, resulting in patterns such as Kare i kam! ‘Bring it here’ in alternation with Karem i kam! The frequency of such reductions in the speech of predominantly younger speakers and their rarity in the speech of older people bears all the hallmarks of a recent innovation. However, very similar patterns of reduction are also taking place in Tok Pisin. This paper examines the question of the extent to which these might represent independent parallel developments, and the extent to which these developments might point to a pattern of protovariability at a much earlier stage in the history of Melanesian Pidgin.

 

 

 

Reconstructing the dispersal of Papuan Languages in Island Melanesia

 

Michael Dunn

Max Planck Institute for Psycholinguistics

 

The relationships between the non-Austronesian languages of Island Melanesia are apparently intractable to the comparative method. This study approaches these relationships using cladistics on typological features. While linguistic typology is subject to borrowing, the extent to which this borrowing obscures ancient relationships is an empirical one. Two hypotheses about the dispersal of the IMP languages will be tested against a similarity tree produced using parsimony methods. These hypotheses are:

1) Wurm et al. (1975) proposal, involving a dispersal of Papuan languages into the Bismarcks/Solomons from the Louisiade Archipelago, with a later intrusion of TNG languages into Bougainville

2) A naive geographical model of dispersal, in which languages spread into the Bismarcks, Bougainville, then the Solomons following the simplest possible progression along the archipelago from the NG mainland.

Cladistic methods show that the naive geographical model is very unlikely to be correct, and produce some support for aspects of the Wurm proposal.

 

Language as fun and secret code: Some play varieties in the Pacific

 

Robert Early

University of the South Pacific

 

The extensive multilingual repertoires of Pacific individuals and speech communities are often noted. There is also a recognition that within any particular Pacific language, there may be significant geographical dialectal variation, and some Pacific languages also display highly developed systems of social dialect specialisation.

Another domain of language use which may entail the development of special varieties is that of play language, where the phonology of the standard form is manipulated to produce varieties which can be quite unintelligible to speakers of the standard language. These languages can therefore serve as markers of a special group identity, and as an effective means of keeping in-group communication exclusive and hidden from non-member participation. However, the way that these language varieties are formed does mean that the scrambling rule or code used to form them can often be broken quite readily by other speakers.

This paper looks at some data taken from two play languages, one based on Bislama and one based on Fijian, and analyses how they are formed. Speakers who develop these special languages do so by responding to underlying perceptions about segments and syllables, and so analysis of the manipulations they undertake can show up what some of these underlying perceptions must be.

 

 

 

Reconstructing botanical taxonomies for Proto Oceanic

 

Bethwyn Evans

Australian National Univesity

 

It seems to be a human universal to classify flora and fauna into hierarchies of labelled taxa .The classification of plants by Arosi speakers includes the primary taxa ‘ai to denote plants that are typically labelled “trees” in English, in contrast to rari “herbs and shrubs”, warawaro “vines” and kaaringa “mushrooms” .There are about 200 named ‘ai-type plants in Arosi, some of which are further classified into different subtaxa. Studies of such folk taxonomies cross-linguistically have shown certain universal tendencies in the types of semantic categories distinguished and the ways in which the categories are labelled (see Berlin 1992, Brown 1984) .This paper looks at botanical taxonomies in Oceanic languages with the aim of reconstructing semantic categories and their associated labels for higher order taxa of Proto Oceanic.

References:

Berlin, Brent. 1992. Ethnobiological classification. Principles of categorization of plants and animals in traditional societies. Princeton, New Jersey: Princeton University Press

Brown, Cecil H. 1984. Language and living things. Uniformities in folk classification and naming. New Brunswick, New Jersey: Rutgers University Press.

 

 

 

The cruel destiny of vowels in fourteen Banks languages 

 

Alexandre François

LACITO-CNRS, Paris

 

Data collected in the fourteen languages still spoken in the Banks Islands, north Vanuatu, reveal a striking case of historical convergence: the enrichment of their vowel systems. Except for Mota, which still perpetuates the five vowels of Proto-Oceanic, all languages in this area have increased their vowel systems up to 7, 8, 9, 10, 12, or even 14 or 16 members. The aim of this paper is to reconstruct the historical processes which led to the present-day phonemic exuberance. First, the combination of a stress-induced syllable reduction and of a general umlaut phenomenon resulted in the phonemicisation of new vowel qualities. Second, in two languages, the loss of certain consonants has resulted in compensatory lengthening, and the phonologisation of vowel length. Despite being typologically familiar, these examples of phonemic developments are remarkable here due to the diversity of the resulting vowel systems in such a small geographic area.

 

 

Subgrouping hypotheses in North Vanuatu 

 

Alexandre François

LACITO-CNRS, Paris

 

First-hand data collected in the Banks and Torres, combined with the documention already available on north Vanuatu languages, make it possible to propose a few hypotheses on the internal organisation of this area in genetic and historical terms. The comparative method proves its efficiency, provided it is applied not only to phonetic and lexical data, but also to features of the morphology and syntax. The result is a tentative cladogram classifying the twenty odd languages of the Torres Is, the Banks Is, Maewo and Ambae; it is hoped to later extend the tree further west and south, but the difficulty in accessing fine-grained data is sometimes an obstacle. Finally, conclusive remarks will be drawn on the scale of a possible "Southern Oceanic" higher-order subgroup (Lynch), highlighting the problems of discriminating between shared and parallel innovations.

 

 

Polynesian loans in the Solomon Islands 

 

Paul Gerarghty

University of the South Pacific, Suva, Fiji

 

In my 1994 paper on Linguistic Evidence for the Tongan Empire, I noted numerous Polynesian loans in Melanesia and Micronesia, but none from the Solomon Islands, thus implying that there was no prehistoric contact between Polynesia and the Solomon Islands. I now wish to propose that there may indeed be linguistic evidence for prehistoric Polynesian contact with the Solomon Islands. As in my previous study, I will present two types of data: a 1777 list of islands known to the Tongans compiled by Anderson, who accompanied Cook, and a study of Solomon Island words that appear, mainly on phonological grounds, to be of Polynesian origin.

 

 

‘We-person’: What does it mean, why does it mean it, and where does it come from?

 

David Gil

Max Planck Institute for Evolutionary Anthropology, Leipzig

 

From Vanuatu, up through the Solomons and New Guinea, across the Indonesian archipelago, and up to the Southeast coast of mainland China, a noun meaning ‘person’ may be used in construction with one or more pronouns in order to derive complex pronouns. Addition of ‘person’ generally has the effect of assigning a plural interpretation to a pronoun that is otherwise either singular or unmarked for number. Moreover, in the case of first person pronouns, addition of ‘person’ sometimes but not always has the further effect of forcing an exclusive interpretation onto a pronoun that is otherwise either inclusive or unmarked for the inclusive/exclusive distinction.

The first part of this paper provides a description of the pronoun-‘person’ construction, focusing on Bislama and Tok Pisin, varieties of Malay/Indonesian, and dialects of Southern Min Chinese. The second part attempts to provide an explanation for why the pronoun-‘person’ construction is invariably plural and sometimes exclusive. The third part offers a somewhat speculative account of the history of the construction, suggesting that it originated in trade languages of Southeast Asia, from where it spread out into the Pacific.

 

 

A freely accessible computerised database of Austronesian basic vocabulary

 

Simon Greenhill, R. Blust, & R.D. Gray

(Greenhill: University of Auckland)

 

The Austronesian lexicon is of great interest to both academics and the public. However, most of this information is contained in books and journals, which can be difficult to access quickly and efficiently. Robert Blust at the University of Hawaii at Manoa has collected a large number of Swadesh 200 word lists from Austronesian languages. We have built these lists into a large dynamic database and have it placed on the internet. The purpose of the database is to make this data freely available for any interested parties, allow easy searching within and between lexical items and facilitate comparison between languages. The database currently contains word lists from over 220 languages, and over 42,000 lexical items. This talk will describe the database system, the technologies underlying it (PHP, MySQL, Unicode) and the use of this database. Any comments, additions or suggestions for the database will be gratefully received.

 

 

An evidentiality contrast in Pingilapese auxiliary verbs

 

Ryoko Hattori

University of Hawai'i at Manoa

East-West Center

 

The Pingilapese (Austronesian, Micronesian subgroup) canonical morpheme order is subject e/aeverbobject. Good and Welley (1989), the only known study of Pingilapese, reports that the particles e (/e/) and ae (/E/) mark an NP as subject, and that the two forms are in free variation. Through the analysis of data elicited from native speakers of Pingilapese, however, I claim that e and ae are not subject markers in the traditional sense. Rather, I propose, they are auxiliary verbs that convey a contrast involving evidentiality the speaker’s relative certainty about the truth of the proposition expressed (Chafe 1986, Mushin 2001, and Payne 1997). While e encodes a high degree of evidentiality, ae encodes a low degree of evidentiality. This is the first time that a grammatically expressed evidentiality contrast has been reported for a Micronesian language (Kenneth Rehg, personal communication).

 

 

 

Oral versus written stylistics in Hawaiian

 

Emily Hawkins

University of Hawai’i at Manoa

 

This paper discusses some specific differences between written and spoken Hawaiian, particularly looking at sentence types and the verb phrase. Simplification of tense/aspect features is coupled with greater detail in the postposed verbal particles along with intonation features. The integration of new and old observations based both on oral samples and reported speech reveals significant differences between written and spoken Hawaiian that heretofore have been overlooked in teaching the language.

 

 

Eye, 'heart' and place: idioms in Maskelynes

 

David Healey

SIL, Vanuatu

 

This paper describes the three Maskelynes nouns most commonly used to form idioms. Besides their primary physical meanings of eye, inside(s), and place, they occur far more commonly in idioms with extended meanings. Eye idioms are often associated with round shapes that sometimes include a centre point. Inside idioms are usually associated with emotions and the inner being or 'heart'. Place idioms are usually associated with weather and ambient conditions. The structure and function of these idioms are described in this paper.

 

 

English loanword adaptions into Polynesian languages

 

Jonathon Herd

University of Toronto

 

This study considers patterns of English loanword adaption into several Polynesian languages, including Tongan, Hawaiian, NZ Maori, Rarotongan, Tahitian and Niuean, and focuses on nativization patterns for non-native consonants. While certain cross-linguistic adaption tendencies are universally maintained (e.g. devoicing of voiced plosives), other non-native segments - particularly fricatives - are subject to various repairs intralinguistically. I argue that both these generalized and language-particular patterns of adaption fall out naturally as a consequence of individual underlying phonological structure, which is subject to a constrained degree of cross-linguistic variation. Patterns of diachronic sound change are investigated and used as evidence for internal phonological organization within each host language, by implementing a phonological model employing contrastive features, and conceptualizing sound change as the result of either neutralisation or reinterpretation of underlying feature contrast. This analysis has implications both for our conception of nativization and our view on the internal organization of sound systems.

 

 

 

Genitive relative constructions in Polynesian languages

 

Jonathon Herd, Diane Massam, and Catherine MacDonald

University of Toronto

 

In this paper we examine aspects of the genitive relative construction in four Polynesian languages: Hawaiian, Maori, Niuean, and Tongan, with a focus on Maori and Niuean. In this construction, the subject of a relative clause, cleft, or question can appear as a genitive alongside the head. This construction is known within Polynesian studies (e.g. Clark, 1976, Hawkins, 2000, Baker 2002), but its grammatical properties have not been studied in detail. We study the properties of the construction and extend the discussion to related constructions in Maori (actor emphatic), Malagasy (bodyguard), Tongan (nominalizations), and English (‘It is her game to lose.’). A Niuean cleft example of the genitive relative construction appears below.

Ko e  tama fifine fulufuluola  haana         ne      lagomatai.

Ko    child   girl     beautiful      3ps.GEN     T/A    help

‘It is the beautiful girl that he helped.’

 

 

 

Ups and downs in Tokelau

Robin Hooper
University of Auckland

Tokelauan contains a morpho-syntactically defined set of directional adverbs which are reflexes of Proto Polynesian forms. The set comprises mai ‘towards speaker’/venitive, atu ‘away from speaker’/anditive, ake ‘upwards’ and ifo ‘downwards’. In addition to meanings directly related to the vertical axis, ake and ifo figure prominently in the linguistic expression of the system of spatial organisation. They also exhibit a number of semantic extensions to more abstract domains, and these form the subject of this paper. Meanings related to the widely attested metaphors up is more (better, higher status, older) and down is less (lower status etc.) are not unexpected, but other metaphorical uses are less transparent, for example those connected with thought, introspection and perception. Some aspectual functions can also be detected.

 

Aspects of Pileni phonology

 

Even Hovdhaugen

University of Oslo

 

The Polynesian language Pileni has compared to most other Polynesian languages a rather complicated phonology. The purpose of this paper is to look at Pileni phonology from three perspectives:

a) the relationship between Pileni and the other Polynesian languages,

b) recent/ongoing phonological changes in Pileni, and

c) the relationship to the neighbouring Äiwo language.

This is a report from a research project in progress and the paper will probably contain more questions than answers.

 

 

 

Bidialectal effects on reading: Word recognition in Hawai‘i Creole English

 

Aya Inoue

University of Hawai’i at Mānoa

 

In heterogeneous linguistic situations, the relation of the spoken to the written language is more complex for bilingual/bidialectal speakers than monolingual/monodialectal speakers.  This paper investigates the effects of different orthographic and phonological systems as factors in visual word recognition by dialectal speakers.  Hawai‘i Creole English (HCE), the English lexifier creole spoken in Hawai‘i, is phonologically different from Standard English (SE), but like many other creole languages it has no widely accepted orthography.  SE-HCE bidialectal speakers and monolingual SE speakers were tested with familiar visual forms (items in SE, loanwords from substrate languages) and unfamiliar visual forms (items in 2 HCE spelling systems, pronounceable nonwords).  The experimental results suggest the inhibitory effect of bidialectalism for the processing of unfamiliar visual forms.  Bidialectal speakers arguably have more complex orthography to phonology mappings from the dual phonological systems (HCE, SE) they command.

 

 

 

Cyrtosperma taro in Yap

 

Ritsuko Kikusawa

Tokyo University of Foreign Studies

 

It has been suggested that the introduction of Cyrtosperma cultivation into Yap was by Carolinians who migrated to the island to settle in the coastal area on the island (Intoh, pers. comm.). In this paper, I will compare some terms related to Cyrtosperma taro and its cultivation customs in Yapese with those in Micronesian and non-Micronesian Austronesian languages to examine how much this claim can be supported or not supported with linguistic data.

 

 

 

On the development of the number systems of Oceanic pronouns

 

Ritsuko Kikusawa

Tokyo University of Foreign Studies

 

Proto-Oceanic pronoun system has been reconstructed as having a simple singular-plural contrast (Loss, Ross and Crowley 2002). A comparison of the reconstructed system for Proto-Extra Formosan and that for Proto-Polynesian supports this reconstruction. However, Oceanic languages today commonly exhibit multiple-number pronominal systems, where the dual (and other) number(s) are observed as well as singular and plural. A question arises here then as to how the Oceanic languages developed the multiple-number systems, and what the motivation was for these parallel changes to take place.

In this paper, pronominal forms in Oceanic languages are first examined to confirm that the development of the multiple-number system took place relatively recently, certainly after the split of the Oceanic languages. Then, possible motivations for these changes are discussed, paying particular attention to the existence of “first person dual inclusive” (or, “1+2 singular”) in the original system, which is considered to be one of the preconditions.

 

 

 

Being or not in Banoni

 

Piet Lincoln

Hawaii

 

In 2002, I reported on some interesting problems with marking of case roles in Banoni, particularly that with givng the recipient is coded on the verb like a transitive object.

1. Na ko ta mani-gho vai borogho
    I 1sg:real fut give-2sg obl pig
    I am giving you a pig.

Now, we will look at some other verbs for which the transitive object suffix is even less expected:

2. teese pode-gho?
    coconut born-2sg
    Do you have a coconut?

3. no tani-gho va-daame?
    you:sg be-2sg cause-good
    Are you well?

4. ghamam ghi-ghina-mam na dzoko
    1excl:pl REDUP-not-1excl:pl human:pl child
    We have no children.

Parallels are sought in nearby languages that have cognates of the Banoni podo: Mono-Alu poro 'born' and Roviana podo 'born.' Mono-Alu does have a similar possessive use of poro, more explanatory observations are still being sought.

 

Phonological and lexical convergence in central New Ireland, Papua New Guinea

 

Eva Lindström

Stockholm University

 

This paper re-examines the area of phonological convergence in north central New Ireland first noticed by Malcolm Ross (1988). This area centres on Kuot, the only non-Austronesian language in the province, and extends with varying force through two languages on either side of it. The distribution of shared phonological processes and phonotactic features is studied using additional data from field notes and published sources.

 

In trying to interpret the history of contact that has given rise to the convergence, reference is also made to lexical borrowings, and some preliminary conclusions are drawn.

Ross, Malcolm 1994. "Areal phonological features in north central New Ireland". In Dutton, Tom and Darrel T. Tryon (Eds) Language Contact and Change in the Austronesian World. Berlin/New York: Mouton de Gruyter

.

 

 Japanese loanwords in the Micronesian region

 

Daniel Long

Tokyo Metropolitan University

 

In this paper, I explore the state of Japanese loanwords used in several languages of the former Micronesian region colonies (including Chamorro, Carolinian, Ponapean, Kusaiean, Trukese, Palauan, Mokilese and Woleaian) .To understand the socio-historical context in which such words entered these languages, I briefly review linguistic aspects of the Japanese administration .In my analysis of loanwords, I first focus on entries in available dictionaries, but then analyze data from current fieldwork on Saipan into various sociolinguistic facets of Japanese loanwords there including: semantic fields covered by Japanese loanwords, differences in words which have entered Saipan Chamorro and Saipan Carolinian, (un)awareness of the Japanese origin of the loanwords, (in)comprehensibility of Japanese loanwords in communication between Guam and Saipan speakers, role that communication with Guam people plays in the realization of words’ Japanese origin, the use of Japanese-origin loanwords by Saipan speakers as an argot in the presence of Guamanians.

 

 

 

The Central/Southern “boundary problem” in Vanuatu subgrouping

 

John Lynch

University of the South Pacific

 

In coining the term “boundary problem”, Pawley (1999:128) referred to “the messy, overlapping groupings that stem from old dialect chains”. A number of such problems can be found within the Southern Oceanic subgroup (consisting of the languages of Vanuatu and New Caledonia), though only one will be discussed here. The languages of Efate share a number of innovations exclusively with the reasonably well-established Central Vanuatu subgroup, and especially with the languages of Epi immediately to the north. At the same time, however, they appear to share other innovations exclusively with Erromangan, which on other grounds clearly belongs to the quite well-established Southern Vanuatu subgroup. The paper will assess the value of different kinds of innovations in making subgrouping hypotheses, and will also partly foreshadow an overall subgrouping hypothesis for this whole subgroup.

Reference:

Pawley, Andrew. 1999. Chasing rainbows: Implications of the rapid dispersal of Austronesian languages for subgrouping and reconstruction. In Selected papers from the Eighth International Conference on Austronesian Linguistics, ed. by Elizabeth Zeitoun and Paul Jen-Kuei Li, 95-138. Taipei: Academica Sinica.

 

 

Another look at nuclear-layer serialization: Positional slots in Saliba complex verbs

 

Anna Margetts

Monash University

 

Saliba, an Oceanic language of Papua New Guinea, makes extensive use of nuclear-layer serialization where two or more verb stems are joined together, forming a single inflected verb. In these constructions prefixes attach to the first stem and suffixes to the last stem of the sequence. A same-subject constraint governs the sequence and stems also have to agree in transitivity status. As has been discussed in the literature for a number of other Oceanic languages (e.g. Early 1993, Sperlich 1993) the non-initial stems of the constructions can express a wide range of functions, e.g. the result of an activity (expressed by the initial stem), directionality, or adverbial and aspectual functions.

(1)

ye-sikwa-he-beku-ø

(2)

ye-kabi-dobi        

 

3sg-poke-caus-fall-3sg.o

 

3sg-touch-go.down            

 

‘he poked it to make it fall’

 

‘he reached down’

(3)

ye-he-kata-namwa-namwa-i-gai

(4)

se-paisowa-gehe

 

3sg-caus-learn-red-good-app-1ex.o

 

3pl-work-finished

 

‘she teaches us properly’

 

‘they finished working’

In this paper I present an analysis of such complex verbs where the different functions of non-initial stems are associated with separate positional slots in the constructions. Saliba complex verbs are mostly composed of two stems but sequences of three stems also occur. Based on this, four positional slots can be distinguished and each slot is associated with stems performing a certain type of function. There is a functional transition between the initial open-class stems with their lexical meaning to the increasingly grammaticalized, closed-class stems in the slots towards the right.

V1 (head) - V2 (result) - V3 (directionality) - V4 (adverbial/aspectual function)

Structural slots in Saliba nuclear-layer serialization

These positional slots are not identical with the slots in the surface structure of a serial verb but are abstracted from examples with three stems that show the ordering constraints of co-occurring stems. The analysis of complex verbs in terms of positional slots allows us to describe different types of serial constructions in a more systematic manner than approaches which exclusively refer to the surface structure of the construction.

 

 

The attrition of Japanese negation: the case of Palauan Japanese

 

Kazuko Matsumoto and David Britain

University of Tokyo and University of Essex

 

This is a progress report on research investigating Japanese dialect contact and subsequent language obsolescence in the former Japanese territory of Palau (1914-1945). Japanese occupation led both to a large Japanese immigrant population and considerable Japanese-Palauan bilingualism on the part of the native islanders. Subsequent American rule (1945-1994), however, halted the expansion of the Japanese speech community there.

We will examine some methodological and theoretical issues involved in language death studies; with what should the use of a dying language (Japanese in this case) be compared? - with formal standard or colloquial Japanese of the mainland or with the fluent spoken Japanese of older speakers in Palau? We address this by comparing rememberers and semi-speakers in Palau with fluent speakers in Palau and with Japanese speakers in Japan.

Our results highlight the necessity of comparing rememberers and semi-speakers with fluent speakers from the same community, and not with fluent speakers in a geographically distant community.

  

A computationally assisted analysis of Tahitian oral tradition

 

David Meyer

University of  Edinburgh

 

There is a long history of manual linguistic analysis of poetry and oral tradition .A method is described here that applies computational linguistic algorithms to assist in the identification of poetic patterns of Tahitian oral tradition .The linguistic patterns are considered to be poetic when they exhibit repeated structure that is not found in similar analyses of Tahitian prose .The data comes from a corpus collected by John Orsmond in the early 19th century .The principal technique applied is a modification of Rens Bod’s Data-Oriented Parsing algorithm.

 

 

 

A quantitative study of animacy effects in Bislama

 

Miriam Meyerhoff

University of Edinburgh

 

It has been suggested that animacy constrains several variables in Bislama. These include: the alternation between zero and pronominal anaphors in subject and object positions (Crowley 1990:241, 326-7); the alternation between oli and i with 3p subjects (Crowley 2002:227). Previous quantitative studies of these variables provide mixed support for the animacy hypothesis. Crowley’s prediction holds for 1st and 2nd person subjects, but is less clear for 3rd person subjects (Meyerhoff 2000:219). Recent data analysing the distribution of null objects in a corpus of conversational Bislama finds an animacy constraint may hold in the sub-corpus of village speakers but not with urban speakers. This paper will summarise these findings and examine the frequency of oli/i with 3p subjects and identify where the variables show statistically most robust patterns. The limits and advantages of using even a small corpus of conversational Bislama will be discussed.

References:

Crowley, Terry. 1990. From Beach-la-Mar to Bislama: The emergence of a national language. Oxford: Oxford University Press.

—— 2002. Serial Verbs in Oceanic: A descriptive typology. Oxford: Oxford University Press.

Meyerhoff, Miriam. 2000. The emergence of creole subject-verb agreement and the licensing of null subjects. Language Variation and Change. 12. 203-230.

 

Reflexives and intensifiers in New Caledonian and Polynesian
languages

 

Claire Moyse

University of  New Caledonia

 

In New Caledonian and Polynesian languages, the semantic middle notion may be either unmarked or marked with the Proto Oceanic reciprocal prefix *paRi- in order to describe some grooming actions, collective events or natural reciprocal events. On the other hand, reflexive constructions - although they may present several marking patterns including this type of "reciprocal" derivation - usually only require the presence of an intensifier marking the coreferency between the subject and the pronominal object. I will present the different reflexive constructions in which intensifiers are attested and will try to draw a typological classification of these intensifiers.

 

 

Determination and quantification in Pileni

 

Åshild Næss

University of Oslo

 

 In Næss (2000), a grammar sketch of the Polynesian Outlier Pileni, I listed eight different forms under the heading of ‘determiners’. However, it is not entirely obvious that all these forms have the same formal status, and labelling them ‘determiners’, rather than, for instance,  ‘quantifiers’, in some cases seems to be a rather arbitrary decision. For example, some but not all of the ‘determiners’ may be combined, and the precise semantic parameters that govern their distribution are not at present clearly understood. In this paper I will reexamine the Pileni ‘determiners’, partly on the basis of new data, and make an attempt at a clearer characterisation of their function and distribution, as well as discussing the question of whether the forms should all be attributed to the same formal class.

Reference:

Næss, Åshild. 2000. Pileni. Languages of the World/Materials 325. Munich: Lincom Europa.

 

 

 

Proto Oceanic fish names

 

Meredith Osmond

Australian National University

 

Building on the work of Paul Geraghty and Robin Hooper, I have reconstructed around 180 fish names for Proto Oceanic. Other than providing the full list, my intention in this limited time is to discuss briefly the difficulties of working in this particular semantic field and to single out a handful of examples which pose interesting problems of semantic reconstruction.

 

 

 

Patterns of stability and change in Oceanic terms for plant and animal taxa

 

Andrew Pawley

Australian National University

 

Oceanic languages spoken by coastal communities on high islands in Melanesia and western Polynesia generally have between 1000 and 2000 named categories (taxa) of plants and animals .The largest domains are plants (usually between 500 and 1000 taxa), fish (200-500), and marine invertebrates (100-250), with smaller terminologies for land invertebrates (50-100), birds (30-150), reptiles, mammals and amphibians. Proto Oceanic was almost certainly spoken in the Bismarck Archipelago, a region relatively rich in land and marine life, and it is likely that number of taxa for each of these domains fell at the high end of the range, totalling around 2000 taxa in all. Some observations will be made on the structure of Oceanic taxonomies and nomenclatures, though we are handicapped in this by the paucity of systematic descriptions.

Reconstructions of the lexicon of Proto Oceanic have so far yielded barely 400 names for animals and plants (including about 150 fish, 44 marine invertebrates, 73 land invertebrates, 27 birds and over 100 plant names). This amounts to roughly 20 percent of the estimate of 2000 POc taxa .It appears that certain kinds of lexemes have been more stable than others. Explanations for patterns of stability and change will be sought in (a) the geographic distributions of particular species and genera, (b) inherent properties of the referents and their relation to humans (morphological and behavioural salience, usefulness, prominence in ritual and cosmology), and (c) the linguistic properties of lexemes, e.g. their internal structure and their rank in the taxonomy.

 

 

 

The various uses of ko in Māori: A unified analysis

 

Elizabeth Pearce

Victoria University of Wellington

 

Although there is general agreement that ko belongs to the class of prepositional particles in Māori (Biggs 1969, Bauer 1997, Harlow 2001), the singularity in the category assignment of ko appears not to match up with a multiplicity in its uses. In this paper I argue that, despite the seeming diversity of its uses, the common characteristic of ko is that it has a specificational function. This means that the apparent diversity is not  an effect of multiple functions or meanings associated with ko itself, but simply that there are a number of different syntactic  constructions which involve the specificational function.

I therefore argue that the role of ko is specificational when it occurs as a marker introducing constituents of all of the following types: (i) equational predicates, (ii)  topic constituents, (iii) focused constituents, (iv) individuated referents of a pronoun, and (v) naming phrases (Bauer 1997: 221). In developing this argument, I show first that ko is unlike other prepositions in that it fails to contribute a locational meaning (place or time) and/or it fails to mark an argument role in the sentence (of the type: agentive, benefactive, comitative, etc.). In this respect, ko is like a copular in its ‘bleached’ semantics.  The second part of the argument consists of showing and motivating the various syntactic configurations in which ko-constituents can be located.

I conclude that ko can be viewed as a prime exemplar of the particle class, the class of items which are ‘grammatical words’ (Biggs 1969: 51).

References

Bauer, Winifred. 1997. The Reed Reference Grammar of Māori (with William Parker, Te Kareongawai Evans and Te Aroha Noti Teepa). Auckland: Reed.

Biggs, Bruce. 1969. Let's Learn Maori. Revised 1973. Wellington: Reed.

Harlow, Ray. 2001. A Māori Reference Grammar. Auckland: Pearson Education.

 

 

 

Sulka of East New Britain, a mixture of Oceanic and Papuan traits

 

Ger Reesink

Max Planck Institute for Psycholinguistics, Nijmegen

 

Sulka is known in the literature as a Papuan isolate of New Britain. There are a number of typical Oceanic features in this ‘Papuan’ language: the marking of realis and irrealis in pronominal proclitics to the verb, presence of prenominal articles, attributive adjectives having a nominalized form and requiring their own article, some deictic elements clearly cognate with forms in Oceanic Tolai.

On the other hand, Sulka’s possessive constructions and its plural noun formation (many declensions) suggest an older Papuan strain. There seems to be evidence of an early contact with Oceanic languages of southern New Ireland, while a more recent layer of evidence (lexical items and social organization, as expressed in names of the clans belonging to the matrilineally defined moieties) points to contact with Sulka’s present neighbour, Oceanic Mengen.

 

 

 

Linguists, Literacy, and the Law of Unintended Consequences

 

Kenneth L. Rehg

University of Hawai’i at Manoa

 

In 1970, the Pacific and Asian Linguistics Institute of the University of Hawai‘i launched a fourteen year effort designed to document and support the languages of Micronesia. The first goal of this undertaking was to prepare grammars and dictionaries of these languages, the second was to train Micronesian educators in the principles and practices of bilingual education, and the third was to develop vernacular materials for use in Micronesian schools. This paper assesses the consequences of those endeavors, both intended and unintended. In particular, it focuses upon the concept of ‘standard orthography’ and how that notion, in Micronesia and elsewhere, has commonly impeded the development of vernacular language literacy. More contentiously, it considers the possibility that the conventional goals of vernacular literacy programs might, in some circumstances, be counter-productive; that is, rather than enhancing linguistic vitality, they might, in fact, diminish it.

 

 

 

Proto Oceanic flora terms

 

Malcolm Ross

Australian National University

 

The present phase of of the Oceanic Lexicon Project concerns the reconstruction of Proto Oceanic terms for flora and fauna. In her paper for this conference Bethwyn Evans examines the organisation of Proto Oceanic plant terminology and higher-order flora taxa. In this paper I present reconstructions of terms for individual plants and lower-order taxa and discuss some of the difficulties I have encountered in making these reconstructions.

Galeya dictionary project

 

Lawrence Kenji Rutter

University of Hawai’i at Manoa

 

Galeya is a member of the Papuan Tip cluster spoken on Fergusson Island in the Milne Bay Province, Papua New Guinea. The language has never been documented. The presentation will include some background information on the language along with a brief description of the dictionary project.

Enseignement expérimental des langues et de la culture kanak à l'école primaire publique de la Nouvelle-Calédonie

 

SAM Léonard &VERNAUDON Jacques

Université de Nouvelle-Calédonie

 

Afin de mettre en œuvre l'accord de Nouméa qui dispose que "(les) langues kanak sont, avec le français, des langues d'enseignement et de culture en Nouvelle-Calédonie", et que "(leur) place dans l'enseignement et les médias doit être accrue et faire l'objet d'une réflexion approfondie", un enseignement expérimental des langues et de la culture kanak a été mise en place à partir de l'année 2002 à l'école primaire publique de la Nouvelle-Calédonie. En 2004 ce dispositif concerne 9 sites.
Notre communication s'attachera à rappeler les circonstances politiques de la misa en place de cette expérimentation (Accords de Nouméa) et les finalités qui lui sont assignées.
Nous présenterons également l'organisation du dispositif d'enseignement, de formation, d'évaluation, les programmes et les supports
.

Language and cultural preservation as a cultural goldmine:
an indigenous researcher's perspective on integrity and intimacy in the dialogue between language and culture

 

Ruth Saovana-Spriggs

Australian National University

 

This paper is in the spirit of sharing ideas and opinions with those in the Pacific Island countries who are passionate about keeping local languages and cultures well and alive in local communities, as I am.
When I consider the issue of integrity and intimacy in the dialogue between language and culture, I begin to wonder if there is indeed a need to re-consider three (at the moment) most important related areas with regards to language and cultural preservation(s).
The first issue deals with the criteria used in the classification of languages into two categories: languages under threat of dying out and those that are perceived as not immediately facing the threat of extinction. The second issue addresses western traditional means and strategies for preservation of local languages and cultures in contrast to oral traditions prevalent in Pacific Island countries. Thirdly, at this point in time in the history of a research tradition introduced in the Pacific region since Euro-american colonization, I feel the time has come to raise the level of interaction and participation between internal/local and external/international academic researchers with regards to studying and researching local languages and cultures. This would be instead of external academic researchers, as now, working exclusively with individuals in local communities who do not have the necessary scientific training but are experts of their specific cultures and languages.

Subject in Pukapuan: absolutive or agentive argument or no subject?

 

Mary Salisbury

Massey University

 

The category 'subject' in ergative Polynesian languages has long been debated. It was widely argued or assumed during the 1970s and 1980s that the ergative and not the absolutive argument is the subject of a transitive clause, although some linguists held the opposite view. More recently it has been argued that the notion of 'subject' has no usefulness or that the features used to diagnose the category subject are spread between the ergative and absolutive arguments, or even that ergative clauses are in fact intransitive and that the absolutive argument is the subject.

Pukapukan clauses which are semantically transitive (containing an agent and a patient in a two argument clause) may exhibit casemarking ranging from 'ergative' to 'accusative' as well as a third option 'passive'. The features of subjecthood are investigated for each core argument of clauses in the different casemarking patterns. While the identification of subject is clear for intransitive clauses and transitive clauses of the 'accusative' pattern, it is concluded that the notion of 'subject' is not very helpful for either the 'ergative' or the 'passive' pattern since both arguments (the agent and the patient) exhibit subject-like properties. Tests for 'object-hood' are also largely inconclusive. Two conflicting views of the category 'subject' in ergative Polynesian languages are discussed with reference to the 'ergative' pattern in Pukapukan.

Pukapuka's place in Polynesia:
"East is east, west is west and never the twain shall meet?"

 

Mary Salisbury and Kevin Salisbury

Massey University

 

Pukapuka in the northern Cook Islands lies on the border of the western and eastern cultural areas of Polynesia. Pukapukan belongs to the Samoic-Outlier subgroup of Nuclear Polynesian according to the traditional subgrouping hypothesis of Pawley (1966), yet it also shares a number of features with East Polynesian languages (including 25% of the exclusively Eastern vocabulary listed in POLLEX). On the basis of shared sporadic changes, Marck (2000) proposed a new subgrouping of Ellicean languages that does not include Pukapuka but which does include its near neighbours Samoan and Tokelau. Marck's evidence is re-examined and Pukapuka's relationships within (and beyond) Polynesia are discussed in relation to vocabulary, grammar, cultural items, fauna and flora and with reference to oral traditions of migrations and voyaging.

  

  

Chronology of Rotuman consonant changes

 

Hans Schmidt

Afrika-Asien-Institut, Abteilung für Südseesprachen, Hamburg

 

Many more sound changes have occurred in Rotuman than in its rather conservative relatives, the Polynesian languages and Fijian. The most recent one was documented by Horatio Hale and some missionaries in the 19th century as a “change in progress”. It was one crucial piece of evidence that helped Biggs (1965) to identify loanwords from closely related languages (“indirectly inherited words”) within the Rotuman lexicon.

In this paper, I will try to group some consonant changes together and determine their sequence as well as compare the Rotuman development with general beliefs about the order of consonant changes.

 

 

 

An analysis of the function of te in Apma (Central Pentecost, Vanuatu)

 

Cindy Schneider

University of New England

 

This is a work-in-progress analysis of the te particle in Apma. Te exhibits a variety of functions, but it is difficult to come up with a single defining term that encompasses all of its uses. While it can frequently be glossed as a completive or emphatic marker, there are also numerous examples where a completive/emphatic gloss would be inappropriate – and alternative glosses reflect an assortment of disparate functions. It is possible that te’s function is tied to the TAM of the verb, or to the type of verb (for example, active versus stative). Furthermore, when an overt object is specified, te can appear either before or after the direct object. In this paper I will give examples of the contexts in which te is used in Apma. I will also look at the way similar-looking particles are treated in other Vanuatu languages.

 

 

 

Genres in Kilivila

 

Dr Gunter Senft

Max-Planck-Institute for Psycholinguistics, Nijmegen

 

It has always been problematic to classify text genres for non-Indo-European languages, especially if this classification is based on the tradition of, and the technical terms defined in, European philology and text or discourse analysis. After a brief discussion of the technical term ‘genre’ and the function the various ‘genres’ are claimed to fulfil, this paper presents a first emic typology of genres in Kilivila, the language of the Trobriand Islanders (Papua New Guinea) providing illustrative examples of various types of these genres. The typology is not only based on the Trobriand Islanders’ own metalinguistic terms for these genres, but also on the relationship these genres have with the ‘situational intentional varieties’ which are also distinguished - and labelled - by the native speakers of this language.

 

 

 

Topicalization in Palauan revisited: a text-based study

 

Shimoji Michinori

Tokyo University of Foreign Studies

 

This study will focus on topicalization in Palauan, which involves three sentence structures schematically shown below.
Structure 0: Pron. [ VP ] (NP0)
Structure 1: NP1 [ VP ]
Structure 2: NP2 [actr-VP ] (NP3).
Structure 0 is an unmarked structure from which Structures 1 and 2 are derived by topicalization. Structure 0 has the pronominal subject, which may be specified by a lexical NP (NP 0). In Structure 1, NP1 indicates the topic/ Actor. Structure 2 has the non-Actor topic (NP 2). The verb in Structure 2 obligatorily carries a pronominal prefix co-referential with the Actor argument, which may be specified by NP2 following the verb.
In previous studies (Josephs 1975; Lemaréchal 1991), no statements have been made regarding the presence or absence of restrictions on topicalization of non-Actor NPs. I will show that, in recorded texts, the frequency of topicalization differs depending on the semantic role of the NPs.

 

Grammaticalization and structural scope increase - Evidence from possessive-classifier-based benefactive marking in Oceanic languages

 

Dr Jae Jung Song

University of Otago

 

This paper discusses the development of possessive classifiers into benefactive markers in Oceanic languages. On the basis of Tabor and Traugott’s (1998) Diachronic String Comparison, this change episode will then be demonstrated to involve structural scope increase contrary to the widely held assumption that scope decrease is a manifestation of grammaticalization. The paper also identifies as an empirically testable hypothesis the strong connection between scope increase, and the formal identity between source items and their grammaticalized descendants.

 

 

Kuwae: the story in Namakir

 

Wolfgang B. Sperlich

UNESCO

 

Following a rendition of the story in Namakir by Chief Masoeripu, I will present the text and its analysis along the lines established by Ellen Facey in her Nguna Voices. After introducing the ethnographic context I will touch on dilemmas of transforming oral performances into text, and then I will focus on style and rhetoric and the grammatical devices used.

 

 

How long do we have? Dating the changes of Oceanic through the witness of archaeology

 

Matthew Spriggs

Australian National University

 

The linguistic sequence of changes that produced Proto-Oceanic and its sub-groups can be compared to aspects of the spread of the Lapita Culture, its ancestors and successor cultures in the Pacific as a datable archaeological sequence in the same region. This comparison is particularly easy for areas not settled pre-Lapita in Remote Oceania, but perhaps is also as persuasive for areas settled for much longer time spans in Near Oceania and Island Southeast Asia. The latest thinking on such dates will be aired and compared with the relatively dated sequence (or sequences) of linguistic change for the Oceanic languages.

 

 

 

Taivosa: A case of deliberate language shift

 

Apolonia Tamata

University of the South Pacific

 

One way to mourn the death of a close relative is to cease speaking one’s dialect and shift to speak the dialect spoken by the deceased relative. This is called biuvosa, or taivosa, in Nasarowaqa Fijian and its occurrence anywhere else is not known. This paper will provide some background on the custom of taivosa and a brief description of the dialects concerned to portray their relationship. The extent of the shift to the other dialect will also be explored as well as the reactions of those who have had to tolerate the change in the language.

 

 

 

The role of second language acquisition theory and practice in Pasifika language maintenance in New Zealand

 

Melenaite Taumoefolau

University of Auckland

 

This paper is about how second language acquisition theory and practice can inform Pasifika language maintenance by communities and by the school system in New Zealand. Studies show that Pasifika youth in New Zealand are shifting from Pasifika languages to English. I draw on research and literature on second language acquisition to explain why and how this shift is taking place, and to explain how second language acquisition theory can inform language maintenance efforts and make them truly successful. When groups of people shift from one language to a more dominant one, they are becoming engaged in language acquisition of the dominant language. This happens because the conditions for acquiring another language are satisfied, whether intentionally or not. For language maintenance to be successful, language acquisition of the dying language needs to take place.

Linguistic stratigraphy in the central Solomon Islands

 

Angela Terrill

Max-Planck-Institute for Psycholinguistics, Nijmegen

This paper examines the extent to which linguistic borrowing can be used to shed light on the existence and nature of early contact between speakers of Papuan and Oceanic languages. The question will be addressed by taking one Papuan language, Lavukaleve, spoken in the Russell Islands, central Solomon Islands, and examining lexical borrowings between it and nearby Oceanic languages, and also with reconstructed forms of POc.

The results of this type of study provides information on the nature of cultural contact during the last 3500 or so years since speakers of Oceanic languages first arrived in Melanesia.

 

 

 

Pacific And Regional Archive for Digital Sources in Endangered Cultures

 

Nicholas Thieberger

University of Melbourne

 

PARADISEC (paradisec.org.au) is a distributed digital archive of material related to languages of the Pacific and Southeast Asia. We have been digitising fieldtapes that were located at the Coombs Building at ANU, and are now seeking any kind of relevant digital data for safe-keeping. In this talk I will outline the process of building this digital archive and the way in which it is already responding to the needs of both researchers and speakers of endangered languages.

 

 

 Developing a linked media corpus of South Efate

 Nicholas Thieberger

University of Melbourne

 

While it has been possible for some years now to play audio files on computers, there has been no simple method for working interactively with field recordings in order to provide citable references. Such references provide an authority for example sentences that we routinely use in our research as linguists. As part of my PhD I developed a methodology for linking text and audio which I will demonstrate in this paper.

 

 

 

Teanu and other languages of Vanikoro

 

Benjamin Tua

Institute of Solomon Islands Studies

 

On Vanikoro in the Solomon Islands, the Vano and Tanema languages are now extinct, and the Teanu language has only a small number of speakers left. In this paper I will describe the linguistic situation on Vanikoro today, and will also present some information about my on-going work on the Teanu language.

 

 

Structure, style, and content in dictionary entries for an Oceanic language

 

David S. Walsh

 

This paper arises from current work on a bilingual dictionary for an Oceanic language---Raga, North Pentecost, Vanuatu. Some differences between a basic entry structure model and an enhanced entry structure model which attempts to meet the needs of the Raga lexicographic situation are considered and speculatively evaluated. Some problem areas in entry style and content are also discussed.

 

 

 

What's in a name, an orthography, or a cluster bomb?

 

David S. Walsh

 

Three recent problematic published statements concerning an OC language, Raga, are considered. The first concerns an allegedly preferred language name; the second concerns an allegedly established orthography; and the third concerns the alleged non-presence of consonant clusters. Some associated issues are explored, and an attempt is made to set the record straight.

 

 

Aspects of noun classification in Savosavo

 

Claudia Wegener

Max Planck Institute for Psycholinguistics, Nijmegen

 

Savosavo is a Papuan language spoken in the Solomon Islands. Nouns in Savosavo belong to one of two gender classes, masculine or feminine. For nouns referring to humans the class membership is determined by sex; this is also the case for nouns referring to other animate beings if their sex is obvious in appearance or behaviour or important in the context. Nouns referring to inanimate objects are by default masculine. However, in some contexts a noun can be assigned temporarily to the feminine class, e.g. to emphasize the smallness of its referent. Thus when two objects are compared, the larger will be masculine as usual, but the smaller can be treated as feminine. In other examples it is not the actual size of the object, but politeness or aspects of the situation described that make a change in class assignment possible. This talk will explore the extent of this phenomenon.

 

 

 

The Hawaiian “ka/ke” article system: The historical development of a gender-like system

 

William H. Wilson

University of Hawai’i

 

Hawaiian is distinctive within Eastern Polynesian in dividing lexemes into classes based on their use of reflexes of the Proto Eastern Polynesian “definite” article *te .There is both phonological and morphological conditioning for the distribution of the two allomorphs ka and ke (and a few less common variants.)  A number of minimal pairs exist, e.g., ka pā ‘wall’ and ke pā ‘plate.’ While ka is clearly the innovative form, it is also the unmarked form in Hawaiian. The change from ke to ka in Hawaiian can be related to phonological rules of Hawaiian that resulted in surface strings of article plus base that could be interpreted as deriving from more than one underlying form. An underlying form with ka rather than ke was chosen and later surfaced in the language in deliberate speech. Morphological rules predicting the distribution of ka were later expanded in both their phonological and semantic conditioning.