피동의 의미론

Abstract

The paper has two parts : (i) the universal characterization of passive, ( ii ) the hisorical explanation of Korean passives. Typological studies have shown that the universal characterization of passive cannot be achieved in terms of formal properties like case markings, verb morphology, word order or the syntactic transformations. Thus it is contended that the univeral definition of passive should be based on the meaning, the meaning of passive that could consistently and uniquely delimit the set of passive expressions. The passive phenomenon is located on the sentential level: Thus, first the semantics of sentence. The sentences of human language are best analyzed as having multiple structures : (i) the structure of the state of affairs, ( ii ) the topic-comment structure and so on. The first structure could be descrived in terms of predicates and arguments or semantic roles, the latter in terms of the topic, i.e. what the talking is about, and the comment, i.e., the talk itself. The concept `subject` takes the most central position in the characterization of passive. And it is shown that the definition of subject in terms of Dixonian trichotomy S, A, P and Keenan`s Subject Properties List are not valid. I propose that the concept `subject` should be identified as the concept `topic`. I insist that the two concepts covey the same meaning, they have the same meaning in their terminological definition, i.e. the subject or the topic is what the sentence is about. The meaning of passive is now defined as the patient topic, or patient subject. Thus, a sentence is passive if the topic or the subject is the patient. It is maintained that the active and the passive are independent way of forming the topic-comment struture out of the state-of-affairs structure. The defintion is truly universal in that languages can vary in their manifestation of the passive meaning, but they have the same property when they have the passive meaning, but they have the same property when they have the passive meaning, i.e. the patient topic. On the basis of this definition it is shown that the so-called impersonal passives are not true passives and that the English sentence `John was given a book` is not the passive, either. The second part of the paper concerns about the origin of the Korean passives. The passive expression in the nominative-accusative type languages are not prototypical, when compared to the actives. Thus the passive expression, especially the passive veb morpholagy could have been developed historically from the other basic expression. It is established that the Indo-European passive verbs have their orgin in the middle forms of` the verb. In a similar fashion, Korean affixed passive verbs are shown to have been derived from causative verb forms that are found in the sentence having the causative and middle structures. The logic of the delvelopment is put forward and some evidences thereof are given. And it is shown that the analysis of Korean passives is best attained when we take the diachromic aspect of the passives into account. Some treatments of Korean passives are criticized and the historical illumination of the Korean passives has its validity.