Lexical Bundles in a Saudi General-Audience Podcast in English: A Corpus Analysis

Research finds that lexical bundles are vital to fluent language processing as they reduce cognitive load when memorized as chunked sequences of language, especially for second-language (L2) learners. Although lexical bundles in academic and political discourse have been studied, their use in podcasts is a less-researched domain in corpus linguistics even though podcasts are an increasingly popular medium that offers authentic public discourse for diverse audiences, including L2 learners and instructors. To address this gap, this study investigates the most frequently occurring and widely dispersed lexical bundles in an English-language, general-audience Saudi podcast. A specific corpus consisting of 10 podcast episodes (almost one hour each) was submitted to AntConc for the identification of lexical bundles based on three predefined parameters- length, frequency, and distribution. A lexical bundle was extracted if it consisted of a four-word sequence that occurred at least five times in at least five texts. From a podcast corpus of 111,174 words, 56 four-word lexical bundles were identified and ranked according to frequency, and their grammatical structures were analyzed. Results show that lexical bundles such as thank you so much, a lot of people, be honest with you, and and I was like are high on the list. Structurally, they consist of nominal, verb-based, and prepositional phrases among others, although verb-based bundles are the most common. The study concludes by providing a list of lexical bundles that may be used for podcast-based learning practice in a L2 listening class or for independent learning.