|
Corpus-based lexical bundle analysis of English podcasts by Thai creators |
|---|---|
| รหัสดีโอไอ | |
| Title | Corpus-based lexical bundle analysis of English podcasts by Thai creators |
| Creator | Lalita Saikaew |
| Contributor | Supong Tangkiengsirisin, Advisor |
| Publisher | Thammasat University |
| Publication Year | 2568 |
| Keyword | Lexical bundles, Corpus-based study, English podcasts |
| Abstract | This study examined the four-word lexical bundles in English podcasts made by Thai creators on YouTube, aimed at identifying the common lexical bundles in English podcasts and analyzing the structural and functional categories of lexical bundles. This study collected a 160,534-token corpus from 20 of the most popular English podcasts by Thai creators. The data were analyzed using AntConc 4.4 software, applying a minimum frequency of seven occurrences, and a minimum range of five texts, generating a total of 187 four-word lexical bundles. The structural and functional analysis followed Biber et al.’s (2004) taxonomy. The findings showed that “a lot of people” was the most frequent occurrence. Structural analysis resulted verb phrase fragments were the largest lexical bundle type (65.2%), while functional analysis found that stance bundles were the most dominant category (46.5%). These findings demonstrate that Thai-produced English podcasts represent an informal, interpersonal, and stance-oriented spoken style, used when speakers express attitudes, generalize experiences, and maintain conversational flow. This study offers pedagogical implications for integrating authentic podcast-based lexical bundles into instruction in the Thai context. |
