Multivocality, literally ‘many voices’ was developed over 5 years as an interdisciplinary method in the Learning Sciences community, within Computer Supported Collaborative Learning (CSCL). The initiative culminated in the Springer book Productive Multivocality in the Analysis of Group Interactions (2013). Our goal was to analyze shared data on collaborative learning from different perspectives in order to build a broader framework for understanding group interactions. In our context, the ‘many voices’ and perspectives refer to five researcher groups’ differing epistemological, theoretical, and analytical assumptions. Each group analyzed a different shared corpus, varying across domain of learning, population, institutional setting, and interactional media. Researchers defined a set of ‘pivotal moments’, a boundary object that allowed them to compare shared moments that had been located for different reasons and to explore the extent to which these reasons were compatible. Our collaboration concluded with reflections on when and how productive multivocality could be achieved.