@article{oai:minpaku.repo.nii.ac.jp:00004376,
 author = {崎山, 理 and Sakiyama , Osamu},
 issue = {2},
 journal = {国立民族学博物館研究報告, Bulletin of the National Museum of Ethnology},
 month = {Dec},
 note = {There are two groups of Austronesian languages in New
Guinea: AN1 with S-V-O and AN2 with S-O-Va,c cording to
A. Cape11, who attributes the latter to Papuan (or Non-Austronesian)
influence. But in each verb phrase in itself different
structures are observed in Austronesian and Papuan languages,
basically the former with sVo and the latter with types other
than sVo.
I classified several types in Papuan languages such as 1. Vs,
2-1. Vso, 2-2. Vsb, 3. Vos, 4. Vbs, 5-1. oVs, 5-2. bVs, 6-1. sV, 6-2.
soV, 6-3. sVb. Most belong to 5-1 and 5-2. The verb phrase
composed of such affixes as s, o and b, which mean subjective,
objective and benefactive, respectively, functions as a rigid
unity or morphological kernel (E. Sapir). In other words, both
S and O as a topic have free positions in a sentence. In languages
of Southeastern part of the Hesperonesian family, in particular,
a word order like O-sVo-S occurs, as in Cebuano in the
I Southern Philippines, Palauan in Micronesia, for example.
The structure of this "crossover" type was apt to change
into a S-O-sVo, as seen in AN2. This was most likely influenced
by a Papuan substratum, without causing any confusion of
meaning.
The origin of the postposition also, which occurs with a
S-O-V type, can be sought in an anaphoric use of the preposition
in a sentence-final, such as C-S-V-prep.C (C : complement),
as seen in languages of Western Micronesia (e.g., Yapese,
Ulithian etc.).
The occurrence of the AN2 is explicable from external
evidence found in neighboring languages near New Guinea,
not only from internal reasons, for instance, innovation of
causative constructions, as J. Bradshaw thought.},
 pages = {355--382},
 title = {オーストロネシア語族とパプア諸語の言語接触 : とくに語順変化について},
 volume = {11},
 year = {1986},
 yomi = {サキヤマ , オサム}
}