Andrew Mendelsohn, Queen Mary University of London

Speaker: Andrew Mendelsohn, Queen Mary University of London

Title: Where did infectious disease modelling come from? Three early examples in context

Abstract: 

What can historians and modellers learn from each other? This seminar provides an opportunity by examining the three earliest known examples of modelling epidemics and situating them in historical context. 

Since the 1970s, modellers and mathematicians have reconstructed historic contributions (e.g., Fine on Ross and Brownlee, Dietz on En’ko, Barrow-Green on Hudson & Ross), or traced the evolution of key concepts (Heesterbeek on mass action and on R), or told a wider story beyond infectious diseases (Kucharski). 

Historians and sociologists have illuminated modelling through key episodes in the development of pedagogy (Engelmann on Frost), epidemiology (my own work on Hamer), and policy (Bickerstaff & Simmons on FMD, Mansnerus on UK public health). Given the recent expansion of modelling research and applications around emerging infections, the time is ripe for asking the elementary question of how and why modelling arose at all – and why this might matter. 

I address that question by identifying commonalities across disparate societies and settings in which epidemic diseases – measles, scarlet fever, influenza, malaria – were first modelled and simulated mathematically. The generative places were a boarding school in Russia, a metropolitan office in England, and army regiments in India, between 1880–1910. The three innovators – P.D. En’ko, William Hamer, and Ronald Ross – independently conceived different methods. Analysis of their equations, computations, tables, graphs, occupations, and worlds of work – from Russian community medicine to Atlantic Progressive government to sustainable colonialism in India – enables reconstructing the invention of a way of knowing and of guiding policy. 

The method I am trialling here is twofold: first, examine (early) models not primarily as applications of mathematical techniques, but as translations of unusual ‘real worlds’ that already structure, simplify, and idealise; second, investigate contexts not for their biographical, social, and institutional specificities, but to identify shared structural preconditions of modelling. I conclude by sketching some legacies these origins may have.