말뭉치를 이용한 신문 표제어 연구 (A Corpus-Based Study on Korean Newspaper Headlines)

서지정보

· 발행기관 : 한국언어학회
· 수록지 정보 : 언어 / 36권 / 4호 / 1075 ~ 1099페이지
· 저자명 : 조영순

초록

Cho, Youngsoon. 2011. A Corpus-Based Study on Korean Newspaper Headlines. Korean Journal of Linguistics, 36-4, 1075-1099. This study investigates headlines' linguistic characteristics based on corpora with reference to their parts of speech and final forms. Two corpora are exploited: One is built from some <news> files in the Sejong tagged corpus (2009) and the other is from Korean five major newspapers. Analyzing frequencies of parts of speech of headlines and those of corresponding newspaper articles based on the first corpus revealed that nominals and symbols (or marks) are remarkably frequent in headlines unlike articles. Some in-depth investigations which were carried out by the categories of content words, symbols, and function words demonstrate the followings: Compared to the news reporting, nominals are increased in headlines but adjectives and verbs are decreased with bigger degree of decreasing found in verbs than in adjectives; Symbols are widely exploited for purposes of clearing information, emphasizing expressions, and assisting meaning recovery; Function words, which can be easily deleted, are retained when carrying hardly recoverable meanings and when performing some structural or emotional functions. Analyzing frequencies of six kinds of final forms of headlines based on the second corpus showed nominals, sentence endings, and connective endings are preferred as headlines' final forms in that order. From different degrees of preference in different sections we infer that headlines in the <press reportage> section are of a dry and emotion suppressing style, but those in the <press editorial> section are softer and show feelings more readily, and <the others> are in between in the degree. (Chonnam National University)

영어초록

Cho, Youngsoon. 2011. A Corpus-Based Study on Korean Newspaper Headlines. Korean Journal of Linguistics, 36-4, 1075-1099. This study investigates headlines' linguistic characteristics based on corpora with reference to their parts of speech and final forms. Two corpora are exploited: One is built from some <news> files in the Sejong tagged corpus (2009) and the other is from Korean five major newspapers. Analyzing frequencies of parts of speech of headlines and those of corresponding newspaper articles based on the first corpus revealed that nominals and symbols (or marks) are remarkably frequent in headlines unlike articles. Some in-depth investigations which were carried out by the categories of content words, symbols, and function words demonstrate the followings: Compared to the news reporting, nominals are increased in headlines but adjectives and verbs are decreased with bigger degree of decreasing found in verbs than in adjectives; Symbols are widely exploited for purposes of clearing information, emphasizing expressions, and assisting meaning recovery; Function words, which can be easily deleted, are retained when carrying hardly recoverable meanings and when performing some structural or emotional functions. Analyzing frequencies of six kinds of final forms of headlines based on the second corpus showed nominals, sentence endings, and connective endings are preferred as headlines' final forms in that order. From different degrees of preference in different sections we infer that headlines in the <press reportage> section are of a dry and emotion suppressing style, but those in the <press editorial> section are softer and show feelings more readily, and <the others> are in between in the degree. (Chonnam National University)

참고자료

· 없음