Chair: Joan R. Piggott, Cornell University

Archaeology and anthropology tell us that modern Japanese are not the direct descendants of the Jōmon people who populated Japan until the introduction of wet rice agriculture under 3,000 years ago, nor are they related to the Ainu. Where did they come from? Since no one has yet been able to find the ancient ‘homeland,’ some deny that Japanese is related to any other language.

Joan Piggott, a historian of Japan, will chair this panel, which provides new evidence for the major proposals that have been made to find the proto-historical Japanese homeland elsewhere in Asia. Unger shows that semantically broad common Japanese words are affiliated with Kokwulye (=Koguryō), Puye (=Puyō) and Paykcey (=Paekche), unlike the vocabulary related to Korean, which is semantically more restricted. Kiyose shows that the third century Wa people already spoke identifiable Japanese, and argues that they had migrated earlier from coastal China. Sakiyama argues that Japanese is a mixed language with Austronesian grammatical features more deeply embedded than ‘Altaic’ elements. Beckwith shows that the Old Koguryō language is (as previously suspected) closely related to Japanese, though not to other languages in Northeast Asia, while loanwords indicate the speakers of Proto-Japanese-Koguryōic lived in China in Late Old Chinese times.

The consensus among these views is that Japanese is related to another Asian language, but that later loanwords have obscured the original relationship. Richard Pearson, an anthropologist specializing in Japan, will be discussant.

J. Marshall Unger, Ohio State University

Archaeologists like Gina Barnes, Walter Edwards, and Mark Hudson have criticized Gari Ledyard’s revised version of the "horserider theory," stressing that the larger size of Japanese tumuli preceded the appearance of horse trappings in them, and that both came decades after Ledyard’s date for a Puye incursion (369). But Ledyard’s account of Puye involvement in the formation of the Latter Paykcey kingdom is persuasive, and if Japanese and the Silla ancestor of Korean separated during the Yayoi period, non-Silla peninsular words in Japanese must have been borrowed later. In fact, Japanese morphemes cognate with Korean that are semantically specialized or restricted in distribution often have synonyms of broad meaning or wide use for which Korean cognates have not been found. Some of these synonyms nevertheless match words we know of the Paykcey and Kokwulye languages, a situation reminiscent of post-Norman English, in which Anglo-Saxon words that survive displacement by French synonyms are confined to semantic niches; e.g., English deer: German Tier ‘animal’ (preserving the original sense), displaced by beast and later animal. Is this not like Middle Korean kwol ‘valley’: OJ kura ‘saddle, seat’ = OJ tani ‘valley’? (Cf. tan id., preserved in Kokwulye toponyms.) The archaeologists may be loath to compare Emperor Ōjin with William the Conqueror, but the infiltration of French words into English had more to do with prestige than power, and Japanese could have been infiltrated by words of Puye, Paykcey, and Kokwulye origin in the same way.

Gisaburo N. Kiyose, Himeji Dokkyō University

The oldest material on pre-Nara Japanese is in the Wa section of the IIIc. San-kuo chih. 55 words, mostly names, are recorded. There are no initial voiced stops or affricates; all syllables are open, as in old Japanese. The Yamatö queen’s name Pimiko is analyzable as pi ‘sun’ plus mi-ko ‘honorable child.’ The word for ‘frontier guard’, pina-möri, shows syntax as in Japanese: pina ‘frontier’ followed by mör-i, the verbal stem mör- ‘defend’ with the deverbal nominal suffix -i. Thus, the language of the IIIc. Wa was a direct ancestor of old Japanese. The Wa states, mostly unified by Yamatö (=Yamatai), used this Archaic Japanese. The wave theory, applied to the distribution of dialect words meaning ‘snail,’ shows that Yamatö was centered in the Kinai, presently Nara Prefecture.

When the Yayoi period Wa brought bronzeware, ironware, and technical skill in wet rice agriculture from coastal China, somewhere between Shantung and the Yangtze, in the IIIc. b.c., they were already speaking Japanese, as their autonym Wa was the first person pronoun, as in Old Japanese. VIIIc. b.c. Chinese sources mention Wa dwelling on the southern coast; later, the northern coast. There must have been several Altaic-type languages in ancient China, but all including the Wa were Sinicized, so today no language genetically related to Japanese can be found anywhere on the Asian continent.

Osamu Sakiyama, National Museum of Ethology

The right column indicates Proto-Central-Eastern Malayo-Polynesian forms reconstructed by A. Capell (1969, 1978) and revised by myself.

Kena-no waku-go i pue-puki-noburo. (Nihon-Shoki, vol. 17)

"A young prince of Kena, HE (= i, III sg.) is going up (a river) blowing a flute."

Tösi-no-pa ni, ayu si pasira-ba. (Man’yo-Shu, vol. 19)

I cannot agree with the view that *-i can be explained through ‘Altaic’ more successfully. Old Japanese as well as Austronesian i have the locative function, which Ataic never does.

Christopher I. Beckwith, Indiana University

This paper is based on a new, thorough, philological and linguistic study of the Old Koguryō-Chinese bilingual glosses in the Samguk Sagi. It presents a systematic reconstruction of Old Koguryō phonology and adds new identifications of Japanese cognates, while rejecting or modifying many earlier identifications. It shows how Old Koguryō is related closely to Japanese by divergence (‘genetically’) from a common ancestor, Proto-Japanese-Koguryōic (PJK), and how several of the Korean-related words (thought to indicate a divergent relationship also exists between Old Koguryō and Korean, and thus Japanese) are apparently loanwords into Old Koguryō from Korean, or vice-versa.