This study examines the use of discourse markers okay and yeah by Korean learners of English as a second language (L2) in conversational interaction within the framework of Conversation Analysis. Among several discourse markers, yeah and okay became t...
This study examines the use of discourse markers okay and yeah by Korean learners of English as a second language (L2) in conversational interaction within the framework of Conversation Analysis. Among several discourse markers, yeah and okay became the main subjects of the research because the two markers are acquired relatively early and occur with appropriate frequency, neither overused nor rare, but sufficiently frequent to allow systematic analysis. Analyses reveal that there are seven functions of the discourse marker okay in L2 conversations and six functions of yeah, which exhibit remarkable commonalities and differences in terms of functions. These uses were also analyzed in terms of distributions. Analyses of functions and distributions indicate that the two markers are actively employed to aid communication in L2 conversations.
Approximately four hours of English conversation data were collected from undergraduate and graduate students at Seoul National University with varying proficiency levels, categorized as low, intermediate, upper-intermediate, and advanced. Proficiency categorization was done through participants’ New TEPS scores, which correspond with criteria for English course assignment. From two to four L2 learners constituted one group, and nine groups in total had conversations with given topics.
The analysis identifies seven functional uses of okay. The discourse marker okay can function as a receipt token in second and third turns, marking the recipient’s response to the current speaker’s turn; as a continuer; as a device to signal one’s listening as a recipient; with dual characteristics that orient simultaneously to the prior turn and the forthcoming action, thereby leading the conversation both backward and forward at the same time; as a turn-initial prefacing device; as an interactional device that indicates one’s turn-starting; as a repair marker to correct a trouble source in repair practices; as a turn-exit marker, which indicates one’s turn ending; and as a device for granting turn-bidding, an interactional action that yields one’s role as the current speaker.
Four uses of okay, as a receipt token, with dual characteristics, for prefacing, and as a turn-exit device, align with those documented in native speaker interaction and previous studies. By contrast, other uses of okay, including its deployment as a continuer, as a repair marker, and for granting turn-bidding, are relatively rare in native-speaker data.
The analysis identifies six functional uses of yeah. The discourse marker yeah functions as a continuer, a turn-exit device, a boundary marker (to mark the boundary between distinct interactional actions by either temporarily suspending the prior action or aborting the prior action and launching a completely new one), a repair marker, a stance marker that exhibits one’s epistemic uncertainty, and a device for signaling proceeding difficulty, which is mostly accompanied by non-lexical perturbations. In the case of yeah, most of these functions are hardly analyzed in native-speaker conversations, with the exception of its use as a continuer.
The distribution of the functions of two markers was also analyzed. First of all, the distribution indicates that yeah is more frequent than okay. It also indicates that there are differences in reliance on certain functions of both markers depending on proficiency levels, suggesting that proficiency levels shape the frequency of the two markers. In the case of okay, three uses (with dual characteristics, as a repair marker, and for granting turn-bidding) exhibit limited frequency at lower proficiency levels. Interestingly, the use of okay as a turn-exit device is absent only at the upper-intermediate proficiency level. Similarly, in the case of yeah, its use as a repair marker shows limited frequency at lower proficiency levels, while three uses (as a turn-exit device, as a stance marker, and for signaling proceeding difficulty) exhibit limited frequency only at the upper-intermediate proficiency level. However, such distributions of some functions are too limited to certain individuals to generalize any pattern.
The current study leaves several implications. The two discourse markers yeah and okay, which are typically used as response tokens, exhibit highly diverse functions in authentic L2 conversations, and their distributions vary according to proficiency levels. In the case of okay, some uses are documented phenomena in native-speaker conversations, while others are not. In the case of yeah, all phenomena except continuer uses are L2-specific phenomena. Importantly, the uses of these two discourse markers are particularly related to discourse management. Despite the lack of generalizability, the distributions indicate that proficiency levels do play a role in whether certain functions occur.