<p>Integrating knowledge from different disciplines is considered key to realizing the potential of interdisciplinary collaboration and is central to many interdisciplinary research and education initiatives. However, how to stimulate knowledge integration remains elusive. As a middle ground between metatheories that provide all-encompassing conceptualizations of integration and accounts of specific interventions and methodologies to support integration, we set out to formulate heuristics that we find to shape a specific form of knowledge integration: conceptual integration. Using a single case study action research approach, we studied conceptual integration processes in a team of eight students from diverse Master’s programs. We found that they often made quick decisions about concepts, definitions, and relationships between concepts. However, there were several instances where such decisions later turned out to be false senses of shared understanding. We considered these instances to be indicative of a lack of “conceptual disambiguation”. Instead of first thoroughly making sense of the overlap and non-overlap between their conceptual understandings, they quickly agreed on terminology. So, only at a surface level, their work appeared to be integrated. We continued to explore how we could stimulate conceptual disambiguation and integration. We used Conversational Learning Theory and found that its five dialectics were helpful in identifying, understanding, and addressing the dynamics that hindered conceptual integration. Based on these findings, we argue that conceptual disambiguation should be granted greater attention in facilitating and training conceptual integration, and that conversational learning theory can be adopted as heuristic to recognize and address bottlenecks in conceptual integration processes.</p>