Learning by teaching can be effective, yet many students fail to engage in knowledge building, in which they actively generate inferences and connect the material to their existing knowledge. Recent research suggests creating drawings while orally explaining to others fosters knowledge building and long-term learning; however, the mechanisms underlying this effect--generation, visualization, or both--remain unclear. In the present study, college students will learn about the human respiratory system and then explain what they learned on video to a fictitious peer. Following a 2 × 2 between-subjects design, students will either generate their own words or visuals on a whiteboard while explaining aloud, or they will view instructor-provided words or visuals while explaining aloud. We expect access to visuals (generated or provided) will foster knowledge building and learning outcomes better than access to words. We will also test competing hypotheses regarding potential unique benefits of explaining instructor-provided versus learner-generated visuals.