Some Remarks on Reconstructing Earlier Korean

Abstract

Prior to the fifteenth century, Korean had fewer pitch distinctions than it did in the Middle Korean period. Internal evidence suggests that many, if not most, of these distinctions arose from an earlier, predictable pitch shape. The principal hypotheses summarized in this paper are as follows: The accent classes and word tones of the modern Korean dialects can largely be derived from the Middle Korean pitch system. However, this same dialect evidence, together with philological evidence adduced by Kim Wancin and Ceng Yenchan, also suggests that pitch variation after the first high pitch in a Middle Korean phrase was not distinctive. Many of the tonal distinctions that did exist in Middle Korean were not original. There was once a typical pitch shape for native Korean morphemes in which the last syllable was automatically made prominent. For two-syllable nouns, for example, this canonical shape was Low-High. Morphemes with different pitch patterns were aberrancies. Some were loans. Other came to have a different pitch pattern because of a distinctive segmental shape. Still others may have been etymological compounds that had been reanalyzed. A significant cause of some exceptional pitch patterns was vowel syncope. Among other things, it is suggested that the monosyllabic, highpitched shapes of verb stems such as khú- 'be big' and thó- 'ride' were created through the loss of a first-syllable vowel. Recent philological evidence uncovered by Lee Ki-Moon shows what the earlier segmental shapes of these verb stems were.