ESHS/HSS Annual Meeting

A Dispute on Diagrams: Epistemic Conflict on the Graphic Method in Dutch National Statistics, 1870–1890

• In Event: Data Visualization and the History of Statistics 1

English Abstract

When, around 1880, the Dutch government started to structurally publish graphics as part of their statistical reports, fierce criticism was voiced: “Each table is followed by a statistical graphic. We are generally not in favour of this last method of processing statistical data. Lines of different colours can never say anything that cannot be said simpler and clearer by efficiently coordinated numbers.” Other reviews followed similar lines. Proponents, on the other hand, would argue that graphics finally posed a solution to the “unusable” stream of statistical tables in official reports. Overall, statistical graphics remained contested objects in Dutch governmental statistics until well into the 1880s.

The unclear status of diagrams stemmed from the actors’ belief that statistical graphics not only visualised statistical facts, but also the scope, meaning, and interpretation of these facts; a duality that was both embraced and problematized. Proponents believed graphics added epistemic and rhetorical value that tables could not convey. But critics rejected graphics as imprecise and expensive.

In the standard historical narrative, the rise of government statistics in the nineteenth century has been explained as a consequence of political, bureaucratic, and administrative needs: quantification served to standardise knowledge, support governance, and claim legitimacy through objectivity. In this paper, I explore the dispute about Dutch governmental graphics to argue that this standard narrative on quantification does not hold for statistical graphics. Instead, I claim that graphics ultimately obtained their place in governmental reports because these graphics could appeal to the interpretation or meaning of statistical data. By tracing how the newly coined graphic method was debated, legitimised, and institutionalised, this paper highlights the situated nature of statistical visualisation and shows how ordinary, non-expert mathematical conventions evolved over time.

Author

• Marieke Gelderblom, Utrecht University