Ancient Koguryo
o
o
o˙˙˙˙
, Old Koguryo
o
o
o˙˙˙˙
, and the Relationship of Japanese to Korean
Christopher I. Beckwith
Indiana University
Linguists working on the historical relationship between Japanese and Korean have mostly ignored the
medieval or ‘Old’ Koguryo˙ (OKog) onomastic material, some arguing that it does not, in fact, represent
the language of the Koguryo˙ people in the medieval Koguryo˙ kingdom, but a language of Paekche (Toh
1987, 1989, 1994) or an ancient substratum language of the Korean Peninsula (Kim 1981, 1985), and the
reflexes of this lexical material in Japanese are loanwords introduced by immigrants during the Kofun
period (Unger 2001)
. Consider the examples in (1).
(1) a. OKog mir (SS) ‘three’ : OJpn mi ‘id.’
b. OKog ütsi (KS) ‘five’ : OJpn itu ‘id.’
c. OKog ku (SS) ‘child’ : OJpn
kú (Martin 1987: 452: *kwo) > NJ ko ‘id.’
d. OKog kìr ~ key ‘tree, wood’ : OJpn kì
~ ki ‘id.’
Recent arguments in favor of a genetic ‘Japanese-Koguryoic family’ (Beckwith 1999), however persua-
sive, may be criticized because they do not deal with the often remarked problem that no connection has
been shown between the language of the ancient Koguryo˙ kingdom, or ‘Archaic’ Koguryo˙ (AKog), and
the language of the later OKog toponyms from north of the Yalu on the one hand, and the OKog
toponyms from the central Korean Peninsula on the other. However, if these objections are answered the
argument that Koguryo˙ is related genetically to Japanese (as thought by most scholars who have worked
on the primary data, regardless of their ideas about further relationships) would be strengthened. In view
of the recent disproof of the most important putative Koguryo˙-Korean etymologies (Beckwith 2002), the
close relationship between the Japanese and Koguryo˙ languages could potentially disprove the Japanese-
Korean genetic theory, or at least force a major reformulation.
This paper examines all known linguistic data on AKog from ancient Chinese literary accounts (primarily
the San kuo chih), as well as the King Kwanggaet’o memorial inscription of 414. Since determining
linguistically whether AKog and OKog are related is heavily dependent on the reconstruction of the
phonology of the Chinese syllables used to transcribe these languages, characteristics of the archaic
Northeastern Chinese in use in the Korean area when the Koguryo˙ names were recorded are taken into
consideration in this paper. Analysis of both sets of Koguryo˙ material reveals a clear, regular relationship
between AKog and the OKog language attested half a millenium later, as in the examples in (2).
(2) a. AKog gapma (SKC) ‘great mountain’ ~ OKog §ap ‘high mountain’ : OJpn yama ‘mountain’
b. AKog kuru (SKC) ‘walled city, fort’ ~ OKog ku
ër ‘id.’ : OJpn kì ‘id.’
c. AKog kweyru (HHS) ‘yellow’ ~ OKog ku
ër ‘id.’ : OJpn ki, ku- ~ kú- ‘id.‘
Next, the problem of the OKog dialects, the existence of which was first suggested by Lee (1964, 1983),
and their relationship with AKog, is discussed. Some of the differences evident in the OKog material—
such as the examples in (3), which show that the dialect in question and Old Japanese are closer to each
other in some respects than to the dominant OKog dialect—suggest that one dialect (perhaps introduced
into Korea in Antiquity during one of the migrations of peoples related to the Koguryo˙, including the Ye
or Ye-Maek) may be even more closely related to Japanese.
(3) a. OKog kìr ‘tree, wood’ ~ OKog (dial.) key ‘id.’ (SS) : OJpn kì ~ ki ‘id.’
b. OKog kuërtsi ‘mouth’ ~ OKog (dial.) kútsi (SS) < *kutui ‘id.’ : OJpn kuti < *kutui ‘id.’
The sharp contrast between the shared Japanese-Koguryo˙ material and Early Middle Korean shown in the
examples in (4), though superficially facile, when buttressed with further examples and analysis of shared
phonological innovations supports arguments against the Japanese-Korean genetic theory.
(4) a. OKog puk ‘deep’(SS) : OJpn puka- ‘id.’ : EMK kiph
u˙n (Sasse 1976:138)
b. OKog piar ‘level, flat’ (SS) : OJpn pira- ‘id.’ : EMK ‘y
o˙tho˙/nyo˙tho˙/’yathu˙n (Sasse 1976:138)
c. OKog tan ‘valley’ (SS) : OJpn tani ‘id.’ : EMK kolk
o˙i (Sasse 1976:105)
Since new publications in linguistics (Beckwith 2002) and anthropology-archaeology (Nakahashi 1998,
Hudson 1999) give powerful arguments against the Japanese-Korean theory, it is suggested that despite
important recent contributions to comparative Japanese-Korean historical linguistics (Whitman 1990,
Vovin 1995, Unger 2001b) the current divergence or genetic theory must be either rejected and replaced
with a convergence theory, or comprehensively revised to explain the Koguryo˙ data in a principled way.
Note on Transcription
All unstarred forms cited are reconstructions based on Chinese character transcriptions.
Source Abbreviations
HHS: Hou Han shu;
KS: Kory
o˙ Sa; SKC: San kuo chi; SS: Samguk Sagi.
References
Beckwith, Christopher I. 2000. Toward Common Japanese-Koguryoic: A Reexamination of Old Koguryo
Onomastic Materials. In J.J. Nakayama and Charles Quinn, Jr., ed., Japanese/Korean Linguistics,
Vol. 9. Stanford: CSLI, 3-16.
—2002. On Korean and Tungusic Elements in the Koguryo Language. Transactions of the International
Conference of Eastern Studies, XLVII, 2002, pp. 82-98.
Hudson, Mark J. 1999. Ruins of identity: Ethnogenesis in the Japanese Islands. Honolulu: University of
Hawaii Press.
Kim, Bang-han 1981. The relationship between the Korean and Japanese languages. Hangeul 173-4:657-
667.
—1985. Hankokugo no keitô. Tokyo: San’ichi shobô.
Lee, Ki-moon 1964. Materials of the Koguryo language.
Bulletin of the Korean Research Center: Journal
of Social Sciences and Humanities 20:11-20.
—1983. Kankokugo no keisei.
Tokyo: Narikai shobô.
Martin, Samuel E. 1987. The Japanese language through time. New Haven: Yale University Press.
Nakahashi, Takahiro 1998. Hokubu Kyûshû no Jômon — Yayoi ikôki ni kansuru jinruigakuteki kôsatsu.
Jinruigaku zasshi (= Anthropological Science) 106.1: 31-53.
Sasse, Werner. 1976. Das Glossar Kogury
o˙-pango˙n im Kyerim-yusa: Studien zur Entschlüsselung eines
chinesisches Glossars mittelkoreanischer Wörter. Wiesbaden: Otto Harrassowitz.
Toh, Soo-hee 1987, 1989, 1994. Paekche
o˙ yo˙n’gu, I. Seoul: Paekche Munhwa Kaebal Yo˙n’guwo˙n.
Unger, J. Marshall. 2001a. Layers of words and volcanic ash in Japan and Korea. Journal of Japanese
Studies 27.1: 81-111.
—2001b. Reconciling comparative and internal reconstruction: The case of Old Japanese /ti, ri, ni/.
Language 76.3: 655-681.
Vovin, Alexander. 1995. Once again on the accusative marker in Old Korean. Diachronica XII.2: 223-
236.
Whitman, John B. 1990. A rule for medial *-r- loss in pre-Old Japanese. Linguistic change and recon-
struction methodology, ed. by Philip Baldi, 511-545. Berlin: Mouton de Gruyter.