<p>This investigation examined the effectiveness of different types of explicit classroom instruction on second language (L2) development of the Spanish discourse marker (DM) <em>pues</em>. While several studies have addressed the positive effect of explicit instruction on L2 DM development, the current investigation moves beyond the explicit-implicit method debate by examining the comparative effectiveness of different types of explicit instruction, specifically by comparing the effects of concept-based instruction (CBI), rule-based instruction (RBI), and a control group (CTRL). This investigation contributes to the field of instructed pragmatics by demonstrating how different types of explicit instruction can affect the robustness of pragmatics learning outcomes. Furthermore, with the inclusion of CBI, this investigation expands theoretical paradigms for L2 pragmatics teaching to include a less explored framework for instruction — sociocultural theory (Vygotsky, 1987).</p> <p><br></p> <p>Given the learnability problem posed by the low perceptual salience and extensive multifunctionality of DMs, along with their conceptual connections to important interactional practices and to a semantic core that guides their use in discourse, DMs and <em>pues</em> specifically were identified as a potentially ideal candidate for development through CBI. <em>Pues</em> is translated in English to ‘so’, ‘then’, ‘cos’, and ‘well’ and is a highly frequent feature of Spanish conversation (Domínguez García, 2016; Stenström, 2006a; Stenström, 2006b). DMs like <em>pues </em>contribute to speakers’ ability to communicate effectively and participate in social interaction (Crible & Pascual, 2020; Hayano, 2011; Hoshi, 2017; Thörle, 2016) and, thus, they are important linguistic features for L2 speakers.</p> <p><br></p> <p>Using a pre, post, and delayed post design, data were collected using an oral decision-making task and a dialogue reflection task. The analyses addressed whether CBI, RBI, and CTRL produce the same effect on L2 Spanish learners’ <em>pues</em> frequency of use, <em>pues</em> functional range, and use of <em>pues</em> in interaction. As a secondary objective, the analyses also considered whether CBI and RBI produce the same effect on L2 Spanish learners’ ability to transfer learning to unlearned DMs.</p> <p><br></p> <p>Quantitative analyses, which addressed the learners’ ability to use <em>pues</em> frequently and for a range of functions, indicated an advantage for the CBI group, with CBI learners using <em>pues</em> with greater frequency of use and functional range than RBI and CTRL learners. The qualitative analysis, which addressed the learners’ ability to use <em>pues</em> in interaction, also indicated an advantage for the CBI group, with CBI learners using <em>pues</em> to express a stance of givenness towards utterances as well as to manage turn-taking in interaction. CBI learners’ use of <em>pues</em> for accomplishing these two interactional practices indicated a deeper understanding of <em>pues</em> and how the DM can be used to accomplish social actions in interaction.</p> <p><br></p> <p>Overall, the findings provide support for the claim that CBI is superior to traditional explicit models of instruction (Erikson, Lanning, & French, 2017) and suggest that the learnability problem of DMs may be lessened by providing a conceptual structure that presents learners with a framework for organizing DM multifunctionality and that also highlights the interactional importance of DMs as tools that can be used to accomplish social actions. The findings are discussed considering the role of conceptual knowledge in L2 DM development and implications for pragmatics instruction.</p>