A Corpus-based Study of the Global Issues in TED Talks

• Main Authors: Lee, Tieh-Shang, 李鐵生
• Other Authors: Huang, Da-Fu
• Format: Others
• Language: en_US
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/4qvmzh

碩士 === 南臺科技大學 === 應用英語系 === 106 === TED Talks cover a wide variety of inspiring topics and issues across different fields of knowledge. TED talks have also become a popular channel for English learners across the globe due to the convenient access to English speeches provided by the TED platform. This corpus-based research aims to analyze the target corpus, TED Talks on global issues, to understand the frequency distribution and lexical coverage of the vocabulary in the corpus. The main purposes of the study are to determine: (1) the vocabulary size on the BNC/COCA Word List needed to achieve the 95% and 98% lexical coverage of the target corpus,; (2) the lexical coverage of the target corpus by the Intermediate-level word list of the General English Proficiency Test(GEPT) to understand whether familiarizing with the corpus content will be beneficial to GEPT test takers; (3) the lexical coverage of the target corpus by the New General Service List (NGSL) alone and that by the NGSL combined with the New Academic Word List (NAWL) to compare respectively to the lexical coverage by the General Service List (GSL) alone and that on the GSL combined with the Academic Service List (AWL) in order to discover which possesses more lexical coverage. The target corpus consists of 506 global issue speeches (ca.1,139,093 tokens) retrieved from the TED Talk website; AntConc and AntWordProfiler corpus analysis tools are employed to analyze the data. The results of the study are summarized as follows: (1) Global issues TED Talks corpus has 69% of functional words and 31% of substantive words on the top 100 frequently appearing words. On the top 200 frequently appearing vocabulary, the proportion of functional words gradually decreases, while the proportion of content words gradually increases; (2) ca. 8000 words are needed to achieve the 98% lexical coverage of the target corpus on the BNC /COCA word list for adequate reading comprehension, while ca. 3000 words are needed to achieve the 95% lexical coverage for listening comprehension; (3) The lexical coverage of the target corpus by the intermediate GEPT word list reaches 95.25 %, and among the covered vocabulary, 85.21% appears over 12 times in the target corpus; (4) The lexical coverage of the target corpus by the NGSL reaches 92.47%, higher than that of 89.37% by the GSL, while the lexical coverage by the combined NGSL/NAWL list reaches 93.75%, only slightly higher than that of 93.56% by the combined GSL/AWL list. This study sheds light on the learning and teaching of English vocabulary, particularly the vocabulary size needed for English learners in general and for those intending to take the GEPT. The findings of the study also provide significant implications for designing English teaching materials and pedagogies across all educational levels in Taiwan.