한국어 종결어미 ‘-다’의 자향성 분석

정해권 ( Jeong¸ Hae-gwon ),정경아 ( Jeong¸ Kyung-a ) 국어학회 2021 국어학 Vol.- No.99

한국어 종결어미 ‘-다’는 대표적인 평서형 종결어미로 여겨지지만, 3인칭의 인칭 일치, 접속법 어미, 청자의 신정보, 즉각 증거성, 의외성 등의 특별한 의미를 나타낸다고 보는 입장도 존재한다. 본고에서는 ‘-다’가 기본적으로 1인칭 평서문과 2인칭 의문문에 분포하며 근원 인식자를 나타내는 자향성 표지라고 제안한다. ‘-다’는 평서문에서 1인칭 화자의 내적 상태, 의지, 경험, 주변적 참여를 나타내고 여기에서 약속 등의 의미로 확장된다. 이런 화자의 우월적 정보 지위를 바탕으로 ‘-다’는 2인칭 주어에 대해 예언, 경고, 명령 등을 나타낼 수 있으며, 3인칭에 대해 발견, 감탄 등을 나타낸다. 또한 의문문에서 청자의 내적 상태나 의지, 화자가 우위에 있는 정보를 묻기도 한다. The sentence-terminal suffix -ta has been treated as an unmarked declarative form in Korean. However, there were other opinions that this suffix could represent various meanings and was regarded as a grammatical marker such as third person agreement, subjunctive mood, new information to the addressee, immediate evidential, or mirativity. In this paper, it is proposed that -ta is an egophoric marker that represents a primary knower who has epistemic authority with the speaker of a statement and the addressee of a question. It basically represents the speaker’s internal state, volition, private experience, peripheral involvement, or extended meanings such as promise in a declarative sentence. It also represents prediction, warning, or command for a second-person subject, and discovery or admiration for a third-person subject based on the speaker's privileged access. And it represents the internal state or volition of the addressee or the information in which the speaker has epistemic authority in the interrogative sentence.

한국어 주격중출에 대한 구문문법적 접근

정해권 ( Jeong Hae-gwon ) 국어학회 2016 국어학 Vol.78 No.-

언어유형론에서 논의된 타동성과 관계화를 참고하면 한국어 주격중출 구문은 서술절 구성이 아니라 주어-보어 구성을 나타낸다. 이 구문의 두 명사구는 의미적 관련성을 가지므로 낮은 타동성을 나타내는 구문이다. 가족닮음의 원리에 따라 긍정의 계사 구문도 주격중출 구문의 한 유형으로 볼 수 있으며, 다의적 용법은 계사 구문을 기준으로 변성 구문, 대상 반복 구문, 대상 부분 구문, 대상 소재 구문, 심리형용사 구문으로 확장된다. 다의적 주격중출 구문은 방사 범주의 의미망을 구성하는데, 다의성 연결, 은유적 확장 연결, 하위부분 연결, 예시 연결로 나타나는 상속 관계를 통해 동기화된다. Korean language has a multiple nominative construction (MNC) such as [NP1-i/ka NP2-i/ka V]. This paper aims to reanalyse the MNC by the perspective of Construction Grammar. According to NP Accessibility Hierarchy, the subject can make a relative clause by the primary strategy. However the NP2 of MNC cannot be relativized. So it isn't the double subject and doesn't contain a Predicate Clause. Two arguments of MNC, NP1 and NP2, have closely related meaning, and the polysemy of MNC represents low-transitivity events. This characteristic of MNC makes a difference from the Transitive construction. Based on the Principled Polysemy approach, the prototypical MNC is the Copula construction and MNC is extended to several peripheral constructions, which are the Transform, the Theme Duplication, the Theme Part, the Theme Location and the Psychological Adjective constructions. They are connected in a semantic network and motivated by inheritance links.

한국어 피동문 처리에서의 통사 구조와 빈도의 영향 -그림-문장 검증 과제를 통한 행동 반응 연구-

정해권 ( Hae Gwon Jeong ),이선영 ( Sun Young Lee ) 한국외국어대학교 언어연구소 2013 언어와 언어학 Vol.0 No.58

The difficulty of passives is often explained in terms of their structural complexity (e.g., Double Dependency Hypothesis) or form-meaning mapping problem (e.g., Isomorphic Mapping Hypothesis). This study examined these hypotheses in Korean using a picture-sentence verification task measuring its accuracy and response time. The results from 40 native speakers of Korean showed (i) that passive sentences were more difficult to process than corresponding active sentences, (ii) that passive sentences with dative case (오빠에게 잡혀요) were more difficult than those with subject case (오빠가 잡혀요) and (iii) that no difference was found between matched and unmatched conditions within the same type of passives. The findings were discussed regarding structural account and relative frequency of different types of passives.

한국어 문법 현상과 역전태- 단체명사 주어, 주체 존대, 능격 동사를 중심으로 -

정해권 ( Hae-gwon Jeong ) 한말연구학회 2016 한말연구 Vol.- No.40

In the linguistic typological perspective, Korean grammatical phenomena may have relations to linguistic universalities. This study aims reanalyse the group noun subjects, subject honorific expressions, and so-called ergative verbs constructions as voice phenomena parallel with inverse voice. According to the universal nominal hierarchy, the first person pronoun (speaker) is the most topical and frequently coded as a subject. However inverse voice shows that the patient as subject is more topical than the agent, but the agent retains considerable topicality in the nominal hierarchy. Furthermore, Inverse voice represents the default cultural values of a community which are coded in a different linguistic form and are reversed for specific communicative purposes (Maldonado 2007). Similarly, a group noun subject and a honorific subject cannot be a first person. Subjects of so-called ergative verbs in Middle Korean have supernatural powers and are socially remote roles such as kings or generals. So group noun subjects with case marker eyse, subject honorific markers kkeyse and -si-, and so-called ergative verbs represent the way the norm has been reversed.

한국어 술어 부류와 구문 습득

정해권(Hae Gwon Jeong),이영진(Young Jin Lee) 언어과학회 2013 언어과학연구 Vol.0 No.65

Adjective(stative verb), intransitive verb, and transitive verb in Korean are used as a predicate in a sentence, and each type of predicate class requires a specific case marker within its construction. However, the mapping of semantic role of each argument and case marker is not consistent with each other. In cognitive and typological perspectives, the different case alignments from Korean such in ergative languages are case-marked based on another construal of events. Therefore, the learners of Korean as a second language could make some errors with the case markers. To find the acquisition of each construction using Korean predicate classes, we conducted a selective blank task with 157 items for fifty learners. The results showed that 1) in adjective constructions, learners acquired the animate subject arguments easier than the inanimate, 2) in intransitive constructions, learners poorly learned both animate and inanimate subjects, 3) in transitive constructions, learners understood more accurately inanimate objects with accusative markers than animate subjects with nominative markers. For more effective teaching of Korean, it is suggested to make distinctions between the predicate classes and the animate/inanimate arguments in each construction in order to make the functions of each case marker more explicitly noticeable for the learners as utilized in the input enhancement methods.

한국어 조사 ‘에게’의 인지의미론적 접근

정해권(Jeong, Hae-gwon) 담화·인지언어학회 2012 담화와 인지 Vol.19 No.2

It is argued that grammatical categories such as particles or prepositions have more abstract meanings than ordinary lexical categories. However each particle also has an image-schema that is based on embodied experiences, and forms a semantic network around its primary sense (Tyler & Evans 2003). This article aims to identify different senses of Korean particle eygey, that is usually known as a dative particle and also used in various constructions. The image-schema of eygey includes an animate landmark and a nearby trajactor. Its primary sense is a place of animate one and without directional meanings. This sense is consistent with the grammaticalization or historical development of this particle. The directional meanings of giver or receiver are secondarily extended senses in the semantic network of eygey. And the frames related to the particle’s five situations are an animate place, a comparer, a goer, a receiver, and a giver. These frames are associated with eygey constructions.

한국어 태 범주의 연속체

정해권(Hae Gwon Jeong) 언어과학회 2013 언어과학연구 Vol.0 No.67

Traditionally, voice categorization is understood as having its basis on an active-passive contrast (Shibatani 2004). However, in cognitive and functional-typological perspectives, voice is defined as a system of correspondences between event types and syntactic structures (Shibatani 2006, Kemmer 1993), and various voice constructions, including active and passive, fall within this system. Some voice constructions are associated with morphological markers, which can be polysemous. Functions of these polysemous markers can be identified using different sentence constructions (Goldberg 1995). Semantic maps represent functional domains of grammatical markers on the conceptual space and semantic boundaries between functions. This paper proposes several Korean voice categories based on semantic boundaries, sentence constructions, and restrictions in voice alternation. These voice categories ?active, passive, causative, potential, spontaneous, body action middle, emotional middle, reflexive, reciprocal, and etc.? reflect the differences in image-schemas of various event types. Consequently, voice continuum in Korean is proposed.

사용전 검사제도 및 전기안전의 중요성

정해권,Jeong, Hae-Gwon 한국전기산업진흥회 2000 電氣産業 Vol.11 No.6

영어권 교포 학습자의 한국어 격조사 처리에 관한 ERP 연구

이선영 ( Sun Young Lee ),정해권 ( Hae Gwon Jeong ) 이중언어학회 2014 이중언어학 Vol.57 No.-

This study investigates the processing of case markers in Korean using Event Related Potentials (ERPs) with a picture-sentence verification task. An experiment was conducted with 13 heritage speakers of Korean from English-speaking countries and 18 native speakers of Korean. There were three conditions: an agent semantic incongruence condition, a patient semantic incongruence condition and a syntactic violation condition. The results revealed (i) that different ERP components were elicited between heritage speakers and native speakers, (ii) that N400s were elicited for semantic incongruence conditions for heritage speakers (at the onset of verbs for agent incongruence and at the onset of case markers for patient incongruence) whereas P600s were elicited for native speakers (at the onset of case markers for agent incongruence and at the onset of voice marker positions for patient incongruence), and (iii) that P600s were elicited for syntactic violation condition for both groups. The findings were interpreted in such way that heritage speakers` processing of case markers is different from native speakers` probably due to their incomplete acquisition of case markers and related syntactic features such as voice morphology.

그림 - 문장 검증 과제에 나타난 한국어 격조사의 의미·통사적 처리의 차이

이선영(Lee, Sun-Young),정해권(Jeong, Hae-gwon) 담화·인지언어학회 2012 담화와 인지 Vol.19 No.3

This study investigates the semantic and syntactic processing of case markers in Korean sentence processing, reexamining two leading theories: Modular Theory (Fodor 1983) and Interactive Theory (Bates & MacWhinney 1987). The former proposes that semantic processing is independent from syntactic processing, whereas they are interacting each other simultaneously in the latter. The accuracy and response time of the judgment by native speakers of Korean were measured in a picture-sentence verification task where the participants judged whether a set of pictures and the aurally presented corresponding sentences matched each other. Control sentences were compared with corresponding sentences in a semantically unmatched condition and a syntactically ill-formed (thereby unmatched) condition. Also, in each condition, subject dropped sentences were compared with object dropped sentences. The results from 40 participants revealed (i) that syntactically ill-formed sentences were processed more accurately and rapidly than semantically unmatched sentences as well as control sentences, (ii) that semantically unmatched sentences were processed more slowly than semantically matched sentences in object dropped sentences, but no such differences were found in subject dropped sentences, and (iii) that subject dropped sentences were processed as rapidly as object dropped sentences in matched conditions, but object dropped sentences were processed more rapidly than subject dropped sentences in unmatched conditions. The findings were interpreted to indicate independent processing levels for semantic and syntactic processing, which conforms to the predictions of Modular Theory, not to the Interactive Theory.