Subclassification of HI-Causatives in Korean

The purpose of this paper is to categorize Korean HI-causatives. Kim (1998) suggests a division into bi-clausal compositional causatives and mono-clausal lexical causatives. According to her suggestion, while compositional HI-causatives possess two verbs, lexical HI-causatives possess only one, since when the originally causative meaning is lost, the verb becomes one lexical item. However, some of the examples she provides cannot be accepted intuitively. An experiment was carried out in order to prove it. Each subject was given eight sentences, five with lexical and three with compositional causative verbs, and was asked to choose from among two pictures which they considered corresponding to the given sentence. Contrary to Kim’s (1998) suggestion, the result shows that Korean native speakers do not consider that HI-causatives are divided into two categories, compositional and lexical, but, instead, they are on a continuum from direct to indirect meaning semantically. According to this experiment, the continuum has mek-HI-ta ‘cause to eat’ at one end and pes-HI-ta ‘cause to take off’ at the far end. In addition, HI-causatives are context-sensitive, so that the context should be considered when native speakers use them.