국어 문장 의미 연구 : 사건과 발화상황을 중심으로 = (The) study of korean sentence meaning

다국어 초록 (Multilingual Abstract)

This paper is that I studied meanings of Korean sentences. One way I study these is that I make sentence's meaning a unit and inspected closely it's semantic construction. Many existing studies have the viewpoint that sentence meaning was put together its parts, namely composition principle. Particularly, verb is the core of sentence meaning and sentence meaning is constituted by meaning structure of verb and combinations of parts. But, contrary to this opinion, I have the viewpoint that the sentence meaning is constituted by the interrelation of whole and parts, namely holism. So we must make clear what the sentence meaning is as a unit. Also, we must make clear what each constitutes of the sentence predicate.
Following event semantics (Davidson (1967), Parsons(1990) etc), I assumed that sentences represent events. But I don't treat that the sentence meaning is completely equivalent to event's existence. Because in relation to the utterance situation, an event is re-constituted. This is reflected by 'pososa(semantic particles)', connective endings etc. So, to inspect closely meanings of Korean sentences, an utterance situation must be hypothesized with an event. I treat meanings of sentences meaning as events on utterance situations. In other words, the event is figure and the utterance situation is ground.
With this viewpoint and methodology, I made clear following.
In chapter 2, I studied what the sentence meaning is, and to which the sentence meaning is paraphrased. As I claimed previously, sentences represent events that re-constituted in utterance situations. In this chapter, I verified the validity of this viewpoint.
In 2.1, I verified the validity of viewpoint that the sentence represent the event. At first, I discussed the nature of event's ontology. According to Kim(1967, 1976, 1993/1994), an event is the exemplification by 'a concrete object' of 'property' or relation at 'a time'. I basically followed Kim, but I treated that events in real world is not same to events in sentences. In sentences, each constitutes represent event's properties. Second, I argued that introducing the event argument can explain many phenomenons of Korean with existing evidences and new evidences. Third, event conception is useful of explanation of syntactic structure, too. In Travis(in prep.), within the context of making a clear link between aspect (outer and inner) and phrase structure, she proposed the extended projection of V, namely 'EP(event phrase)' and 'AspP(aspectual Phrase)'. EP is relative to outer boundary of an event, AspP is relative to the telicity of event. Forth, I look about sentences that represented event repetitions. I studied how sentence constitutes represented these. In my viewpoint, constitutes of the sentence have not proper, fixed meanings but unspecified meanings. So these meanings and functions were specified in relation to sentences as whole. Since constitutes that don't have the semantic function of repetitions can also represent event's repetitions through interactions.
In 2.2, I verified the validity of utterance situation's introduction. Previous endings and connective endings of Korean hold up a true mirror to the role of utterance situations in sentence meaning constructions. In particular, because the grammatical person is relative to the construction of the event and the utterance situation, we can know the relation of these through the grammatical person. In this study, I proved the necessity of introducing the utterance situation through semantic function of '-deo-'(previous ending) and '-kilrae, -nika, -keodeun'(connective endings). These represent event perception of the speaker.
In chapter 3, the meanings of predicates also were specified by whole sentence meaning, namely the event. In previous researches, the lexical aspect is an inherent property that the verb has. But if we take notice of status that 'yaksok-ul jikida[keep one's promise]' or 'yaksok-ul eogida[break one's promise]' exhibited, we can't speak that the verb or the verb phrase always has inherent properties. Also, we can't explain this status with aspectual shift, coercion or re-interpretation etc.
This study has called type like 'yaksok-ul jikida[keep one's promise]' event-dispositional predicate. It appears to report two actions or events in one predicates. The 'yaksok-ul jikida[keep one's promise]'-type reports activities that promise content represented as well as judgements that keep a promise. I rightly examines event-dispositional predicates which are not generally discussed in the aktionsart literature. It is shown that event-dispositional predicate constructions are restricted not by inherent aspectual property, but by dependent property. They are restricted by aspectual property that content of the promise represented. Therefore, event-dispositional predicate's constructions are aspectual object not that they themselves have aspectual property but that they inherit other VP's aspectual property. These phenomenon shows that the predicate's meanings also were specified by the interactions with meaning of whole sentence.
In Chapter 4, I inspected closely the semantic functions of the adverbials in sentences. In previous studies, adverbials play roles that modify VP. But I consisted that all constitutes of the sentence play roles that predicate the event. So the adverbials predicate an event that sentence represents. In chapter 4.1, I studied about semantic functions of 'adjective+ge' adverbials. Since 'adjective+ge' adverbials have various semantic functions at sentence meaning constructions, we must take notice of them. We have the view that sentence meanings were generated and interpreted through interaction between events and constitutes. So semantic functions of constitutes are determined by interactions of event. In accordance with this view, 'adjective+ge' adverbials have following semantic functions. First, since these adverbials support their predications, they have two functions according to whose theme arguments they have. Possible theme arguments are events and constitutes of the event. Second, in case 'adjective+ge' adverbials have theme argument as the event, they represented manners of the events. Third, in case 'adjective+ge' adverbials have theme argument as the constitute of event, they represented a state that have causal relation of the event. So only constitutes that have causal interpretations can be theme arguments of 'adjective+ge' adverbials. Fourth, according to precedent studies, 'adjective+ge' adverbials that have resultive interpretations can appear at sentences on condition of particular verbs and adjectives. But this is not fact. The causal relations between event that the sentence represented and state that the 'constitute+adjective' represented are the true constraint condition.
In 4.2 and 4.3, I inspected closely semantic roles of the temporary adverbials. The temporary adverbials also do not have semantic functions to themselves but be specified by interactions with the event or the utterance situation. For example, adverbials such as 'eoje'(yesterday), 'oneul'(today), 'naeil'(tomorrow) etc, were specified through relation to the event time and the utterance time.
In 4.4, I inspected closely semantic functions of the comparative adverbials. In Korean, the comparative adverbials were appeared by particles such as '-cheorum', '-mankeum' etc. These do not compare semantic properties of NP but compare the event that is constituted to the NP. These show that the sentence meaning is constituted by the interaction of the event and the constitutes of the sentence.
In chapter 5, I inspected closely the independent constitutes. Because these are alien from other constitutes, the status of the independent constitutes was the controversial issue. But at holistic viewpoint, these aren't alien from other constitutes. That is, as other constitutes, these describe what is event and utterance situation. First, the exclamations take all the event argument and the utterance situation argument and judge the value of the event in the utterance situation. Second, the vocatives take only the utterance situation argument and specify who is the hearer. Third, the connectives take only the event argument and specify the semantic relations of events.

목차 (Table of Contents)

• 제1장 서론 = 1
• 1.1 연구 목적 = 1
• 1.2 연구 대상 = 3
• 1.3 선행 연구 검토 = 4
• 1.3.1 문장 의미에 대한 주요 국내 논의 = 4
• 1.3.2 문장 의미에 대한 주요 관점 = 6
• 1.3.3 합성성 원리의 한계 = 8
• 1.4 연구 방법 = 12
• 제2장 사건과 발화상황의 문장 의미 = 15
• 2.1 문장 의미로서의 사건 설정 = 15
• 2.1.1 사건의 본질 = 15
• 2.1.2 사건 논항의 설정 = 20
• 2.1.3 사건의 구조적 실현 = 33
• 2.1.4 사건의 실제적 적용과 분석 = 55
• 2.2 문장 의미로서의 발화상황 설정 = 76
• 2.2.1 발화상황 논항의 설정 = 77
• 2.2.2 발화상황의 인칭제약 = 83
• 제3장 서술어구의 문장 의미 구성 = 92
• 3.1 동작류에 대한 기존 논의 분석 = 92
• 3.1.1 서술어구의 상적 분류 = 93
• 3.1.2 상적 추이(Aspect Shift), 혹은 강제(Coercion) = 95
• 3.1.3 의사-이중 술어(擬似-二重述語) = 97
• 3.2 사건-성향 서술어구의 문장 의미 구성 = 98
• 3.2.1 동작류로서의 특성 = 99
• 3.2.2 사건 내용 점검과 상적 특성 = 100
• 제4장 부사어구와 문장 의미 구성 = 108
• 4.1 부사어구에 대한 기존 논의 분석 = 108
• 4.1.1 ‘형용사+게’ 부사어구 = 108
• 4.1.2 시간 부사어구 = 116
• 4.1.3 기간-지속 부사어구 = 120
• 4.1.4 비교 부사어구 = 122
• 4.2 ‘형용사+게’ 부사어구의 문장 의미 구성 = 123
• 4.2.1 ‘형용사+게’, 사건의 방식과 결과 = 123
• 4.2.2 ‘형용사+게’와 서술어의 관계 = 126
• 4.2.3 ‘형용사+게’와 사건-성향, 발화상황의 관계 = 130
• 4.3 시간 부사어의 문장 의미 구성 = 132
• 4.3.1 사건, 발화상황에 따른 시간 부사어의 분류 = 133
• 4.3.2 동형이의(同形異意)의 시간 부사어와 중의성 문제 = 139
• 4.4 기간-지속 부사어구의 문장 의미 구성 = 143
• 4.4.1 ‘동안’ 부사어구의 의미 = 143
• 4.4.2 ‘만에’ 부사어구의 의미 = 147
• 4.4.3 기간-지속 부사어구와 사건 구성 요소의 관계 = 149
• 4.5 비교 부사어구의 문장 의미 구성 = 151
• 4.5.1 비교 부사어구의 성상 부사적 특성 = 151
• 4.5.2 비교 부사어구와 사건 구성 요소 간의 관계 = 154
• 4.5.3 비교 부사어구와 사건, 발화상황의 관계 = 157
• 제5장 독립어구의 문장 의미 구성 = 160
• 5.1 독립어에 대한 기존 논의 분석 = 161
• 5.1.1 독립어의 일치·호응 = 161
• 5.1.2 독립어의 위상, 하위분류 = 162
• 5.2 독립어의 의미 역할 = 164
• 5.2.1 느낌말, 사건과 발화상황의 관계 = 165
• 5.2.2 부름말, 발화상황과의 관계 = 166
• 5.2.3 이음말, 사건 간의 관계 = 168
• 제6장 결론 = 170
• 참고문헌 = 175
• <Abstract> = 194