Statistics & Computing in British Agriculture & Biology

A review of “Making Sense of Figures:” Statistics, Computing and Information Technologies in Agriculture and Biology in Britain, 1920s-1960s, by Giuditta Parolini.

“During the last war our administrators learned, though perhaps with some reluctance, that men trained in research were essential for the success of the national effort. The remaining nucleus of my department, if I may speak in its praise, constitutes a unit for heavy mathematical computations as efficient, both in machines and men, as the country can command” (p. 162). So wrote R.A. Fisher (1890-1962) — the acclaimed statistician and geneticist — at the outset of the Second World War. It is but one of the important and intriguing new insights into the life, work and significance of Fisher that could be found in Giuditta Parolini’s groundbreaking thesis. It is a quotation that transplants Fisher from the context of evolutionary biology, population genetics and modern synthesis with which most of us are familiar, and sets him to work in the world of men and machines. This is the great strength of Parolini’s approach, choosing as she does to follow her actors into the specifics of their working environments, whether taking tea in the gardens of an agricultural experiment station, computing blood group surveys in a university laboratory, or negotiating the acquisition of cutting-edge equipment in Whitehall. In doing so Parolini establishes an exciting new perspective on the history of statistics and information management, explaining how they came to matter for the life sciences and how the life sciences have mattered for them.

In 1919 R.A. Fisher became an employee of Rothamsted Experimental Station, building the statistical department there while advocating for a wider and deeper involvement of statistical theory within its research programs. At Rothamsted the new statistics was not in and of itself sufficient to transform experimental practice, Fisher having to negotiate its introduction with colleagues, particularly through the formal mechanism of the “Field Plots Committee.” Moreover, Rothamsted was not just a fortunate place of patronage for Fisher, but the specifically agricultural work conducted there was actually formative of his statistical understanding (Chapter 1). Essential in his eventual success were his publications, including Statistical Methods for Research Workers (1925) and the often marginalized Statistical Tables for Biological, Agricultural and Medical Research (1938). The way in which such texts were prepared for publication, and the way in which this intellectual property was managed, point to their significance as discipline building tools for Fisher and his collaborators (Chapter 2). Fisher was also highly opportunistic, turning the records of the Emergency Transfusion Services during the Second World War into a valuable resource for geneticists and statisticians, advancing his mathematical and human inheritance interests — closely linked to his eugenicism — at one and the same time (Chapter 3). However, perhaps Fisher’s most important legacy for the history of statistics, particularly subsequent to his death, was his marriage of statistical research with the exploitation and exploration of new calculating machines. These two elements are found working hand-in-hand throughout the twentieth century, and only increased in significance following the arrival of the digital computer (Chapter 4).

Parolini’s research has been exceptionally extensive, drawing upon numerous archives from around the world. The vast majority of her primary source material cannot be found in existing publications, making this a valuable and useful new addition to the history of twentieth century British science. Parolini has also been generous with her appendices, of which there are no less than 11, ranging from detailed lists of the employees and researchers that visited Rothamsted, to significant statistical formulae, and transcripts of the interviews she conducted with retired station staff. Particularly worthy of mention is her use of the archives of the publishers Oliver & Boyd, housed in the National Library of Scotland. The correspondence between Fisher and his publishers is exceptionally revealing and has thus far been overlooked by his biographers. However, it would be misleading to label Parolini’s dissertation merely a biographical work. The figures of Fisher and his successor at Rothamsted, Frank Yates (1902-1994), certainly provide the backbone to the narrative, but the terrain explored throughout each chapter is rich and diverse.

Historians of statistics and historians of agriculture are clearly the primary target audience for the thesis as a whole, while the growing number of historians of agricultural science will simply not be able to ignore what is an all too rare, new addition to this field. However, outside of these broad categories, historians with less obviously direct interests will also find much valuable material. Historians of the book for instance, will benefit in particular from Chapter 2, dedicated to unearthing the preparation of Fisher’s publications, the significance of their status as “tools,” and how Fisher’s management of his intellectual property through negotiations with his publishers all illuminate the tactics that lay behind these manuscripts. Historians of medicine will be attracted in particular to Chapter 3, focusing as it does on the establishment of the Galton Laboratory Serum Unit under Fisher at University College London. This work on blood groups came to be interrupted by the Second World War, a context that ensures the interest of historians of British science, especially when one considers the way in which different governmental departments and scientific programs interacted with one another during the War, providing new opportunities for statisticians and advocates of Operations Research. Examples of the latter, which have become definitive for the historiography of this period, are replete throughout Chapters 3 and 4. Meanwhile historians of technology and computing will benefit from her sophisticated analysis of the role of “computing aids” in statistics, found in Chapter 4, which details the ways in which the world did and did not change following the arrival of digital computers.

Giuditta Parolini’s “Making Sense of Figures”: Statistics, Computing and Information Technologies in Agriculture and Biology in Britain, 1920s-1960s is thorough, persuasive, and full of surprises. It is of enormous value to anyone charting the history of science in twentieth century Britain and cannot fail to be influential within the same.

Dominic Berry
PhD Student
University of Leeds
ph09djb@leeds.ac.uk

Primary Sources

Rothamsted Research Library and Archive, Harpenden, United Kingdom
John Innes Centre Archives, Norwich, United Kingdom
National Library of Scotland, Edinburgh, United Kingdom
University College Special Collections, London, United Kingdom
Royal Statistical Society Archive, London, United Kingdom
Bodleian Libraries, Oxford, United Kingdom
Royal Society Centre for the History of Science, London, United Kingdom

Dissertation Information

University of Bologna. 2013. 287 pp. Primary Advisors: Giuliano Pancaldi and Staffan Müller-Wille.

Image: Ronald Fisher (on the right) during tea time at Rothamsted Experimental Station (1920s). Copyright Rothamsted Research Ltd. Used with permission.

Leave a Reply