@article{oai:minpaku.repo.nii.ac.jp:00004215,
 author = {崎山, 理 and Sakiyama , Osamu},
 issue = {1},
 journal = {国立民族学博物館研究報告, Bulletin of the National Museum of Ethnology},
 month = {Aug},
 note = {In this paper I compare certain postpositional particles in Hiri Motu
and Japanese. Hiri Motu is one of the SOV type Austronesian
languages, known as Austronesian Type II, which is spoken around Port
Moresby, Papua New Guinea as the principal lingua franca. In particular
I compare the use of Hiri Motu topic maker be and actor marker
ese with Japanese postpositions wa (topic) and ga (nominative) . The
study is based on the language data from the Hiri Motu version of the
New Testament (the Gospel) , which has been made into the Key-Wordin-
Context (KWIC) at the National Museum of Ethnology. In this text,
be and ese are used up to 3,789 and 2,108 times, respectively.
Hiri Motu and Japanese apparently follow the same word order, i.e.
SOV, but as far as the syntactic device is concerned, the affix order in
Hiri Motu, i.e. sVo (s: subjective prefix, o: objective suffix) should be
recognized as a distinct type of Japanese word order, reflecting
anaphorically S and 0 within a sentence and forming a smallest sentence
nucleus grammatically.
As the result of an exhaustive survey, the present paper points out
that be in Hiri Motu is able to topicalize the whole elements including ese
phrase in a sentence, such as `taunina ese be ... (the body-actor ...)' (the
only example: Mat vi. 25) , `daika be ... (who ... ?)' (9 examples: Luk
ix. 46, etc.) , being different from Japanese in that the chaining ga + wa,
or interrogatives + wa is ungrammatical, by reason that the ga-marked
subject expresses non-topic in opposition to the wa-marked topic, and
interrogatives cannot be topicalized for their indefiniteness. Regarding
this phenomenon, the chaining `... ga + interrogatives' doesn't occur in
Japanese. This improbability is caused by the reason that the chaining
`... ga + interrogatives' presumes such an implicitly topicalized sentence
as ' (interrogative + wa) ... ga + interrogative', and that, after all, the
interrogative becomes topic.
What is interesting is that the combination `... ese + interrogative'
is unacceptable in Hiri Motu too, but this outward coincidence results
from the different reason that the ese-marked actor semantically forms
the exclusive relation with interrogatives in Hiri Motu, and it seems that
the new tendency is appearing, rather focussing the actor in itself
pragmatically, as seen in such a sentence `Tau ese daika is dogoatao?'
(Whom [daika] is the man [tau] holding [dogoatao]?) .},
 pages = {1--17},
 title = {ヒリモトゥ語の類型：辞順と後置詞 : KWIC資料に基づく通言語的研究},
 volume = {19},
 year = {1994},
 yomi = {サキヤマ , オサム}
}