@article{oai:minpaku.repo.nii.ac.jp:00003873,
 author = {崎山, 理 and Sakiyama , Osamu},
 issue = {3},
 journal = {国立民族学博物館研究報告, Bulletin of the National Museum of Ethnology},
 month = {Feb},
 note = {日本語は，北方のツングース諸語および南方のオーストロネシア語族の両文
法要素を継承する混合言語である。日本語の系統もこの視点から見直そうとす
る動きがすでに始まっている。これまでに発表したいくつかの拙稿では日本語
におけるオーストロネシア系語源の結論部分だけを述べたものが多かったが，
本稿では音法則を中心とした記述に重点を置き，意味変化についても民俗知識
に基づいた説明を行った。引用語例は筆者がすでに述べたものも含まれるが，
多くは本稿で初めて提示するものである。他者によって言及されている語源説
についてはその妥当性を検討した。また，論述の過程において，日本語音韻史
でこれまで隔靴掻痒の感が否めなかったハ行・ワ行歴史的仮名遣いの表記上の
問題点を摘出した。, That Japanese was formed through a mixture of Tungusic and Austronesian
has gradually become clear. While it is said to be difficult to locate Japanese
in a particular language family, this may be primarily because of the
great length of time over which present-day Japanese was formed. Progress
has been hampered by insistence on the “purity” of the language. Modern
Japanese was in fact formed through a mixing of languages.
After Austronesian first moved from Taiwan into the northern Philippines
around B.C. 2500, including the Malayo-Polynesian subfamily, a branch of
this group left for the Ryukyu-Japanese Archipelago and first settled in West
Japan in the late Jomon period. As people spread northeastward, rice might
have carried with them, as evidenced by the early presence of rice cultivation
in the southern Taiwan and the northern Philippines around B.C. 2000.
Several Austronesian rice-related terms are retained in Japanese while involving
semantic changes; for example, Proto Malayo-Polynesian (PMP) *bǝRas
‘husked rice’ became Ancient Japanese (AJ) *fiya-i (Old Japanese (OJ) hiyë
‘Japanese millet’), PMP *pajay ‘rice plant’ to AJ *fasai (OJ hasë/wasë ‘earlyripening
rice plant’), and so on.
In this paper I focused my concern to show about a hundred cognate
words based on regular sound correspondences between Proto Austronesian
(PAN) or PMP and AJ, earlier Japanese before the Nara period (A.D. 710-
784), and also to explain semantic changes brought about in AJ from the
viewpoint of ethnology. In the case of sound changes from PAN to AJ, the
most remarkable is the confusion of phonetically similar phonemes : the distinction
between *f /ɸ (the present ha column in kana) and *w (the present
wa column), inherited from PAN *p/*b and *w respectively, is suspected of
having become disordered mainly in the intervocalic position of AJ and that
this confusion carried over into the OJ Man’yo-gana writing system (Japanese
syllabaries originated from Chinese characters) used since much earlier in the
Nara period. Some words written with the ha column were rewritten with the
wa colomn, and vice versa sporadically, to which I pointed out the historically
logical etyma tracing back to Austronesian. Another case is the merger of etymologically
different phonemes; PAN *R and Proto Tungusic *d seem to have
chanced to become one phoneme y in AJ. As characteristics of sound change
PAN word-final consonants tend to be dropped in AJ, allowing stems to end
with an open syllable. This trend, i.e. apocope, appears as a common phenomenon
in Malagasy and Oceanic languages (particularly Polynesian languages),
which are found on the geographical rim of the Austronesian distribution.
And as a noticeable sound change from PAN the distinction between voiced
and voiceless consonants is being lost, and in most languages of Oceania as
well as in AJ, this tendency is argued to exist on reliable linguistic evidence.
It would not be surprising even if a similar drift has emerged in Austronesian,
which extended to Japan’s northernmost tip. The reason why the phonological
voiced and voiceless distinction revived at the latest before the Nara period is
inexplicable without taking into consideration on language contact and bilingual
mixtures brought about in the process of forming the Japanese language
after the late Jomon period. I wonder how conservative thought attached to
language monogenesis interprets this extraordinary innovation.
The pitch accent of Japanese is an important element when considering
the origins of the Japanese language. In Austronesian languages roots are
basically multisyllabic. I divide AJ vocabularies etymologically originating in
PAN/PMP into three groups : group 1 appears to inherit the penultimate syllable
of PAN, group 2 the final, and group 3 reflects the whole form. It is interesting
that Poi Tsaan (Huihui-hua), a Malayo-Polynesian Chamic language
of Hainan Island of China, has phonetically changed to resemble a tonal language
retaining in principle the final syllable of PAN (or Proto Cham). This
fact gives a crucial insight into the occurrence of an accent after the loss of
PAN/PMP syllables in AJ.
In a comparative list I classify Austronesian-derived Japanese vocabularies
into several categories such as rice-related terms, wind and direction
names, and basic color terms.},
 pages = {353--393},
 title = {日本語の混合的特徴 : オーストロネシア祖語から古代日本語へ音法則と意味変化},
 volume = {36},
 year = {2012},
 yomi = {サキヤマ , オサム}
}