On the So-called Adjunct Predicates in Korean

Youngjun Jang
Harvard University
yjang@fas.harvard.edu


This paper examines the distribution of the two types of adverbs/adjuncts in Korean, namely, i-suffixed ones (e.g. ppali 'fast') and key-suffixed ones (e.g. ppalukey 'fast'). I argue that while ppali-type is a real adverb/adjunct, ppalukey-type is an infinitival predicate. Evidence for this claim will be drawn from small clauses, depictive predicates, resultative predicates, purpose predicates, and causative clauses in Korean. Both ppali and ppalukey have often been considered as adjunct predicates and each having PRO argument (Shi (1993)), as shown in (1).

(1a)

cha-ka [PRO ppali] talinta. car-Nom fast run-Dec

(1b)

cha-ka [PRO ppalukey] talinta. car-Nom fast run-Dec '(both) The car is running fast.'

However, there is evidence that suggests they do not belong to the same adverb/adjunct class. Given that adjuncts are optional, the fact that in some cases ppalukey-type is obligatory while ppali-type is optional suggests that the former must be a complement and not an adjunct. The complementary distribution of ppali-type and ppalukey-type supports our claim. For example, in small clauses (e.g. "*I consider [SC him honestly]" vs. I consider [SC him honest]), only key-type and not i-type is allowed. Conversely, in cases like (2) where mel 'far' itself cannot serve as a predicate of kicha 'train'only i-type and not key-type is allowed.

(2)

kicha-ka meli/*melkey salchiessta. tain-Nom far disappear-Pst-Dec 'The train disappeared far away.'

---

Last updated July 20, 1997 by rblight@mail.utexas.edu Return to main program