Computing Without Computers?

This article explores Jacques Bertin's Graphique , an analog and visually driven method for processing data in the human and social sciences (HSS), developed at the École Pratique des Hautes Études in postwar Paris, France. Conceived as a semiological alternative to mathematical formalization, Graphique enabled researchers to analyze large datasets through perception and interaction rather than calculation. By reconstructing its rise and decline between 1947 and 1995, the article shows how Graphique embodied a human-centered vision of rationality that competed with computer-assisted multivariate statistical procedures, supported by institutional and national science policies. This case study reflects a broader shift in the history of HSS toward reproducibility, delegation, and shared infrastructures—values that marginalized Bertin's interactive and individual approach. Beyond recounting this technical competition, the article argues that the fate of Graphique reveals how choices about computing infrastructures redefined what counted as rigor and trust in the human and social sciences, while foreshadowing later developments in data visualization and human–computer interaction.