Mapping the changing landscape of American experimental psychology through data-driven topic modeling.

Most empirical research has been conducted by countless psychologists whose names are now largely forgotten, and our work stands on the accumulation of their efforts. Quantitatively summarizing the collective practice of psychological research and tracing its changing orientations across periods offers a useful way to clarify what psychologists, as a professional community, were concerned with and how they pursued empirical inquiry. To this end, the present study comprehensively extracted article titles published from the end of nineteenth century to 1980 in two of the oldest and most influential American journals of experimental psychology, American Journal of Psychology and Journal of Experimental Psychology, and applied topic modeling using non-negative matrix factorization. This method decomposes high-dimensional vectors into multiple components, which are expected to capture the field’s canonical research themes. Even without information about authors or schools of thought, learning, perception, and cognition emerged automatically as distinct topics, and their relative prominence shifted across periods. At the same time, the trajectories of rise and decline differed substantially across topics. A more fine-grained analysis of topic strength further revealed an asymmetry: after the rise of cognitive science, learning-related research increasingly incorporated cognitive elements, whereas cognitively oriented topics gradually became less tied to learning. The historical shifts in psychological interests documented here are therefore better understood not as a sequence of milestone events driven by a few great figures, but as a collective record of the quiet yet steady efforts of many researchers who cumulatively shaped the field.