Task engagement and comprehensibility in interaction

Abstract

This exploratory study examined the relationship between second language (L2) English speakers’ comprehensibility and their interactional behaviors as they engaged in a conversation with fellow L2 speakers. Thirty-six pairs of L2 English university students completed a 10-minute academic discussion task and subsequently rated each other’s comprehensibility. Transcripts of their conversation were coded for eight measures of task engagement, including cognitive/behavioral engagement (idea units, language-related episodes), social engagement (encouragement, responsiveness, task and time management, backchanneling, nodding), and emotional engagement (positive affect). Speakers who showed more encouragement and nodding were perceived as easier to understand, whereas those who produced more frequent language-focused episodes and demonstrated more responsiveness were rated as harder to understand. These findings provide initial evidence for an association between L2 speakers’ interactional behaviors and peer-ratings of comprehensibility, highlighting L2 comprehensibility as a multifaceted and interaction-driven construct.

Comprehensible to Whom? Examining Rater, Speaker, and Interlocutor Perspectives on Comprehensibility in an Interactive Context

“…The present analysis, which appears to be the first to examine self-rated comprehensibility in interaction, extends prior work (Nagle et al, 2022;Trofimovich et al, 2020) by showing that L2 speakers self-rate their comprehensibility similarly to how it is rated by their interaction partners. Regardless of the task or the timing of self and partner assessments, the two interlocutors' ratings were on average within 2-3 points of each other (on a 100-point scale), implying that the interactants had a fairly accurate view of how effortful it was for their interlocutors to understand their speech.…”

“…Nevertheless, drawing on theoretical perspectives that view interaction as socially and linguistically coordinated action (Garrod et al, 2018), where comprehensibility would involve a dynamic adaptation of the interlocutors to each other, we anticipated that self and partner assessments of comprehensibility might be closely aligned. Based on our previous analyses of this dataset, where partner ratings of comprehensibility tended to improve over time (Trofimovich et al, 2020), we also expected that the speakers' self-ratings would demonstrate a similar upward trend. In light of consistent gaps between L2 speakers' self-assessed comprehensibility and external listeners' assessments (Isbell & Lee, 2022), we also anticipated that the speakers' self-ratings would differ from rater assessments.…”

“…Comprehensibility has been extensively studied from the perspective of external listeners, who are typically either trained or naïve raters evaluating L2 speakers' recorded or live performances (Trofimovich et al, 2022). For external listeners, comprehensibility is primarily a speechcentered construct, such that ease of understanding is associated with many linguistic features (Isaacs & Trofimovich, 2012;, including lexis and grammar (e.g., appropriate and rich vocabulary, accurate and complex grammar) and pronunciation (e.g., accurate word stress).…”

“…The least explored perspective on comprehensibility is the one from a speaker's interlocutor (i.e., a person communicating with the speaker). In a recent study, Trofimovich et al (2020) engaged pairs of L2 English speakers in three tasks, asking them to evaluate each other's comprehensibility on seven occasions (2.5 minutes apart) during approximately 17 minutes of interaction. Results showed that comprehensibility ratings followed a U-shaped function, where the speakers rated their partners' comprehensibility as high during the first task, after which the ratings dropped during the second task (likely as a function of the task's difficulty) and then slowly increased during the final task (presumably in response to the speakers' increased familiarity with their partners).…”

“…Although comprehensibility ratings generally seem to reflect various linguistic properties of the speech being evaluated, to depend on the individual profile of the person providing the rating, and to show various degrees of change over time, assessments by external raters, speakers, and their interlocutors would likely yield different perspectives. For instance, during conversation, speakers may become more familiar with each other's speech patterns, so they will require less effort to understand each other as they continue speaking (Trofimovich et al, 2020). The social coordination that arises as speakers work together to accomplish a common task could manifest itself through enhanced collaboration and decreased anxiety, which may further bolster their mutual comprehensibility (Nagle et al, 2022).…”