Collaborative discourse in computer-supported collaborative learning (CSCL) environments is a multifaceted endeavor that encompasses more than just students talking to each other. Rather, it involves interactions aimed at fostering various knowledge practices, presenting challenges such as promoting student participation, engaging them in the continual evolution of their ideas towards higher-level knowledge practices, and effectively evaluating and capturing the complex dynamics of the discourse. My dissertation tackles these challenges through the design and implementation of a series of CSCL designs that leverage the affordances of digital technologies to enhance collaborative discourse in higher education. The three studies reported in this dissertation unfold over a multi-year research-practice partnership with higher education instructors. The first study focuses on the design and implementation of a scaffolding framework that integrates a web annotation tool named Hypothesis to enhance students’ social and cognitive engagement. The second paper expands this effort by introducing a knowledge synthesis intervention that facilitates the connection, analysis, and synthesis of student ideas and leverages these synthesis artifacts to promote sustained knowledge creation across various learning activities. The third study develops a comprehensive analytical framework that leverages natural language processing techniques and network analysis to capture the multifaceted nature of collaborative discourse, focusing on student engagement from their initial attention to the creation of knowledge artifacts. Together, these studies aim to foster productive interactions and sustained knowledge creation within and beyond the social annotation platform, providing actionable insights for the design and evaluation of collaborative learning environments.