Oceanic Linguistics, vol. 36, no. 2 (December 1997)

Contents

In Memoriam: Samuel Hoyt Elbert 1907-1997, pp. 199-204
Emily Hawkins

He Mele Aloha no ka Sweet Man (A Song of Affection for Sweet Man), p. 205
Ka‘upena Wong (translation by Puakea Nogelmeier)

Samuel H. Elbert: An Update to the 1976 Bibliography, p. 206-207
Jack H. Ward

ARTICLES

Semantic Differentiation Between Three Tokelauan Complementizers, pp. 208-226
Robin Hooper

Tokelauan has a number of subject-complement-taking verbs that express ideas of modality, aspectuality, and evaluation. In contrast with the situation in closely related languages, three different complementizers are available to introduce subject-complement clauses. Many predicates allow a choice between two, or in at least one case, all three of the complementizers. It will be demonstrated that the selection of complementizer is influenced by several semantic factors: the type of modality expressed, the agentivity of the complement clause, and whether the matrix clause is negative or affirmative. The semantic specialization that these morphemes have developed exhibits the characteristic of persistence commonly found in grammaticalization processes.

On the Origins of the Possessive in Central Pacific Languages, pp. 227-246
John Lynch

This paper's orginal impetus was to attempt to find the morphosyntactic origins of the well-known Polynesian o and a possessive-markers, and the origins of the phonologically and grammatically similar Rotuman markers ?o and ?e. In exploring these origins, some contribution is made to the development of the Oceanic possessive system in Proto Central Pacific as a whole, and especially within the Fiji dialect chain. In particular, evidence is adduced to suggest that the Polynesian and Rotuman systems resemble those of Kadavu and Gone Dau more than those of Lau, and that this suggests that Geraghty's of Lau, and that this suggests that Geraghty's (1983) Tokalau-Polynesian hypothesis may require some reassessment.

Proto-Kimbe: A New Analysis, pp. 247-311
Ward H. Goodenough

Using data gathered by Chowning, Valentine, and Goodenough in 1954, a reconstruction is undertaken of Proto-Kimbe, an Austronesian group of languages in North Central New Britain. It significantly revises the reconstruction by Johnston (1982). It confirms that Bali and Vitu, included by Johnston, must be excluded from the Kimbe group, though remaining closely related to it. This analysis shows that both the Kimbe languages and Bali and Vitu have had complicated histories, involving extensive borrowing from another Austronesian language by Bali and the merging of two dialect communities in the formation of what became Proto-Kimbe. This analysis also provides evidence that ProtoMalayo-Polynesian *p and *h had not merged in Proto-Oceanic as *p nor had *d and *D merged as Proto-Oceanic *r, as has been hitherto assumed, although such merger is widely attested in many Oceanic languages.

The Pronominal System of Mantauran (Rukai), pp. 312-346
Elizabeth Zeitoun

Rukai includes six dialects, Tanan, Budai, Labuan, Maga, Tona, and Mantauran, of which more is known about some than others. The position of Mantauran in the Rukai network remains controversial because its structure has been obscured by drastic phonological and syntactic changes. Phonologically, its voiced stops have been spirantized. Syntactically, it has undergone a process of grammaticalization whereby bound pronouns have coalesced with the verb stem, and third singular/plural oblique pronouns have been reanalyzed as "nonagent" agreement markers. Li (1977, 1996) has proposed two different versions of the Mantauran pronominal system in an attempt to shed some light on the position of this dialect among Rukai. Li 1996 maintains that there is sufficient evidence to support the claim that Mantauran is the first offshoot of Rukai. The present paper attempts to show that this conclusion may have been premature. Li's accounts of the pronominal system are reassessed and an alternative analysis is presented that permits greater crosslinguistic generalization among the Formosan languages. Comparative data and analyses from other Rukai dialects and other Formosan languages are presented.

On the Proto-Austronesian "Diphthong", pp. 347-361
Adrian Clynes

Arguments are presented against the reconstruction of diphthongs in the phoneme inventory of Proto-Austronesian. The synchronic and diachronic evidence indicates that the elements previously characterized in this way are better analyzed as sequences of two independently reconstructed phonemes, a vowel followed by a consonant. With the elimination of this phonological class, a simpler PAN phoneme inventory than that previously posited is obtained.

Grammar Notes on Siraya, an Extinct Formosan Language, pp. 362-397
K. Alexander Adelaar

This is an attempt to unravel the grammar of a gospel text in Siraya, an extinct West Formosan language. It includes a discussion of the historical setting, the spelling, phonological features, function words, relation markers, morphosyntax of the verb, and the use of deictic verbs as prepositions. Regarding verbal morphosyntax, special attention is given to case marking suffixes, verbal classifiers, compound verbs, and anticipating sequences. Verbal classifiers are lexical elements prefixed to a root (a verb, adverb, or noun) with which they constitute the overall meaning of the resulting verb. Compound verb constructions are constructions in which auxiliaries assume the functions of adverbs in English. These auxiliaries form an open class and are in fact the head of the verb phrase, as they carry most of the marking. An anticipating sequence is an clement of a verb that is prefixed to the preceding auxiliary.

SQUIB

Rukai Stress Revisited, pp. 398-403
Robert Blust

For over sixty years, Austronesianists have known that the reconstructed sound system of Proto-Austronesian cannot explain the contrastive stress of many Philippine languages. In recent years, efforts have been made to support the reconstruction of PAN stress through appeal to non-Philippine evidence. The Budai dialect of Rukai (south-central Taiwan) has contrastive stress, and Malcolm Ross has argued that it correlates significantly with Philippine evidence, thus providing the basis for a reconstruction of PAN stress contrasts. Closer attention to the Rukai evidence shows that this argument is unsupported, as the exceptions are nearly as numerous as the agreements, forcing us to conclude that the history of stress in Budai Rukai is unconnected with that of Philippine languages.

REVIEW ARTICLE

Darrell T. Tryon, ed., 1995, Comparative Austronesian dictionary: An introduction to Austronesian studies, 4 parts, pp. 404-419
Reviewed by Robert Blust

BOOK REVIEWS

Heinrich Zahn, 1996, Mission and music: Jabêm traditional music and the development of Lutheran humnody, trans. by Philip W. Holzknecht, ed. by Don Niles, pp. 420-424
Reviewed by Joel Bradshaw

Shigeru Tsuchida, Yukihiro Yamada, and Tsunekazu Moriguchi, 1987, Lists of selected words of Batanic languages, pp. 424-429
Reviewed by Paul Li

Arthur Holmer, 1996, A parametric grammar of Seediq, pp. 429-436
Reviewed by Elizabeth Zeitoun

Osada Toshiki, 1995, Mundajin no nokobunka to shokuji bunka: Minzoku gengogaku kosatsu [Farming culture and food culture of Munda: An ethnolinguistic study], pp. 436-438
Reviewed by Alexander Vovin

From the Editor, p. 439

Index to Volume 36 (1997), pp. 440-448