A Herman Ottó Múzeum Évkönyve 32. Kunt Ernő emlékére. (1994)

TANULMÁNYOK - KI–MOON LEE: A koreai és az altáji nyelvek összehasonlító tanulmányozása (magyar és angol nyelven)

this comparison was made greater as research in the history of the Korean language showed that -r and -n were in fact verbal noun endings in Middle and Old Korean. At this stage what is needed is deeper research into the respective morphologies of the Altaic languages and Korean. I believe that this is a necessary precondition for further comparative research. 4. Ramstedt took up the topic of Korean-Altaic etymology in the last work of his life (1949). After retiring in 1943, he devoted all his energy to the compilation of this book (Aalto 1975), but as early as 1939 he made the following remark: "Ich selbst habe mit der Zeit ungefähr 5,000 Karten für eine Koreanische Etymologie zusammengestellt. Wie skeptisch man auch sei, so dürfte doch die Hälfte davon jeder Kritik standhalten." (Ramstedt 1939/1951) Of the 1,904 lexical items included in the work, only 1,599 have the forms from Altaic languages shown

9 This is surely the result of careful selection. Here it is necessary to point out two things. (1) 196 of the items were recognized by Ramstedt himself as borrowings. (2) 350 Sino-Korean words (marked by "sk.") are included. There are in fact more Sino-Korean words in addition to these. He had a rather unusual theory regarding Sino-Korean words, but it is difficult to justify. In spite of its many flaws, Ramstedt (1949) provided a good foundation for the lexical comparison of the Korean and Altaic languages. So much so, in fact, that it appears a kind of inertia has taken root among scholars that this work is "the final word" on the subject. Ramstedt himself revised his opinions several times in the course of his many years of research, and there has been considerable effort on the part of other scholars since then. Before new research in this area is undertaken, it is necessary to compile and organize all the research that has been done to date. I would like to briefly discuss two etymologies from Ramstedt (1924). These are the first examplws comparing Korean lexical items with Altaic and Japanese data that he gave after starting to learn the Korean language. The first example is a comparison of Turkic tas 'stone' with Korean tol 'stone' and Japanese to 'whetstone'. Mongolian cilajun is not mentioned, but is hinted at the statement that tas goes back to tilagun < tal 'a-gun. This comparison can also be seen in Polivanov (1927). Chuvash cul and Orochon |olo are also mentioned. As evidence that the liquid in South Korean tol and North Korean tor is originally an 1, Ramstedt (1928) cites Goldi |ollo, |olo, Proto-Turkic *tal' and Chuvash cul. Here he encounters the problem of word-initial *t and *t\ Ramstedt (1932) confronts this problem head on. He interprets Korean tol, Turkic tas *tal', Chuvash cul < *t'al, Mongolian cilajun < *t'ala-gun as resulting from the presence or lack of palatalization. In Ramstedt (1949), however, the only listing under the item tol is Tungus |olo. It includes the annotation "The identity is questionable because of t = | (?)." This can be interpreted as indicating that in his later years Ramstedt came to harbor serious doubts about his earlier comparison of tol with Turkic tas, Mongolian őilajun and Tungus £olo. On the other hand Poppe (1960) accepts the comparison of O-ld Turkic tas, Chuvash cul, Mongolian cilamn and Korean tol as valid.

10 He neglects to explain, however, the difference between word-initial *t and *t' appearing in the reconstructed forms ta *tas < *täs < *täl

2 < *täl

2 a, cul < *t' al

2

, őila^un < *tila-gun < *t'ala-bun. The explanation for this was provided by Poppe (1975). Here he chan9 There is a certain degree of inaccuracy in these figures. 10 It is noteworthy that Poppe (1960) does not include the Tungus words in this comparison. 100