Publications

Detailed Information

Embedded Topicalization in Korean Factive Complement Clauses: An Experimental Approach : 한국어 사실성 구문에서의 화제화에 대하여: 실험적 접근

Cited 0 time in Web of Science Cited 0 time in Scopus
Authors

주은선

Advisor
고희정
Major
인문대학 언어학과
Issue Date
2018-08
Publisher
서울대학교 대학원
Description
학위논문 (석사)-- 서울대학교 대학원 : 인문대학 언어학과, 2018. 8. 고희정.
Abstract
Complement clauses of factive predicates are considered to be structurally reduced compared to their non-factive counterparts. Semantically, they are said to lack assertion – their propositional content is not delivered from the viewpoint of the speaker but simply presupposed, or presented as is (Hooper and Thompson 1973). Syntactically, it is said to disallow root transformations inside of them. Root transformations are a group of syntactic movements including negative constituent preposing, left/right-dislocation and topicalization. They are only available in root clauses and root-like embedded environments (Emonds 1976, Hooper and Thompson 1973).

The root transformation that this study focuses on is topicalization. Topicalization is one of the root transformations that have been thoroughly debated since the original inception of Emonds (1976). When a phrase is topicalized, it moves to the clauseinitial position and receives the sentential topic interpretation. Factive complement clauses do not license this movement inside itself.

To explain why factive complement clauses do not allow topicalization, one must take into consideration the precise structure of factive clauses as well as the nature of topicalization. This study examines previous theories on this issue and recognizes two different approaches that account for lack of topicalization inside factive complement clauses: the operator approach and the clause-size approach.

The operator approach is more semantically motivated. It is inspired by the intuition of Melvold (1991) that a factive complement clause is a definite description of an event. Watanabe (1993) and Haegeman and Urögdi (2010) advocate this approach and aim to explain lack of topicalization as a result of competition for the same landing site or of an featural intervention effect.

The clause-size approach appeals to structural differences between factive and non-factive complement clauses. Mostly laid out in the early works of Haegeman (2004, 2006), it views factive complement clauses as structurally smaller than nonfactive complement clauses. Her view connects to the classic arguments of Hooper and Thompson (1973) that factive clauses lack assertion. What is exclusively projected in non-factive complement clauses encode information such as viewpoint and assertion (Rizzi 1997, Haegeman 2004, De Cuba 2007). Under this view, topicalization is unavailable in factive complement clauses because it requires one of the functional projections that are present in non-factive complement clauses but missing in the factive ones.

However, the situation is more complicated in the case of Korean. In Korean, factive predicates can select two different forms of complement clauses. The large form projects a series of functional structures including the sentence final particle that appears in non-factive complement clauses and matrix clauses. The small form only projects a part of the functional structures projected in its larger counterpart. According to the logic of the operator approach, topicalization would be unavailable in both cases since the factive operator at the clause edge would hinder topic licensing no matter the size of the clause. On the other hand, the clause-size approach expects topicalization to be available in the larger clauses since it projects the necessary structure to license topic.

To the best of my knowledge, these two approaches were never applied to analyze Korean data in previous studies
therefore I judged it would be beneficial to attempt analyzing Korean data with these frameworks and see whether their predictions actually bear out. An acceptability judgment experiment was carried out in order to empirically verify the predictions of each approach. The results demonstrate that both the operator approach and the clause-size approach make correct predictions for factive complement clauses in Korean. A follow-up experiment was designed to verify whether the clause-size approach can be extended to explain the lack of topicalization inside islands. The results provide partial support for the clause-size approach.

This study analyzes a new set of data with existing theories, delineates the predictions of each theory, and verifies the prediction by means of a formal experiment. Based on the results, it calls for a reconciliation of the two approaches in explaining factivity in Korean. Emphasizing the explanatory adequacy of the clause-size approach in factive complement clauses as well as islands, it also advocates a splitCP framework approach to various phenomena at the syntax-pragmatics interface of Korean.
Language
English
URI
https://hdl.handle.net/10371/143812
Files in This Item:
Appears in Collections:

Altmetrics

Item View & Download Count

  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Share