Recent research has begun to explore the role of diagrams as cognitive tools. Here I develop new conceptual and methodological tools for exploring the sociality of cognition involving diagrams. First, I distinguish two varieties of group-dependent cognition. Second, extending Nersessian’s method of cognitive-historical analysis, I show how a suitably-informed “literature review” of diagrams published in scientific articles offers a window into the group-dependent cognition of scientists. I end by sketching future avenues of inquiry, and how this approach may inform science education.