Published by EH.Net (October 2019)
Claire Lemercier and Claire Zalc, Quantitative Methods in the Humanities: An Introduction. Charlottesville: University of Virginia Press, 2019. viii + 177 pp. $19.50 (paperback), ISBN: 978-0-813-94269-8.
Reviewed for EH.Net by Pat Hudson, Department of History, Cardiff University.
This introduction to quantitative methods arose from an earlier French publication (2008) that largely confined itself to historical research and French data. Here the scope has been enlarged to serve a wider humanities readership with examples drawn from (predominantly) North American, as well as French publications. The authors are Parisian Research Professors, one based in the Center for the Sociology of Organizations, the other in the Institute for Early Modern and Modern History at the School of Advanced Studies in the Social Sciences. The pre-history of the volume and the research and teaching backgrounds of the authors is perhaps important in explaining both the strengths and weaknesses of the book.
The volume is based upon a belief in the potential and importance of carefully conceived quantitative work in the humanities across a range of sources, geographical contexts and research questions. The authors argue that practically all sources are potentially quantifiable. Their aim is to produce a survey of the range of techniques applicable to historical data together with a critique of those methods, bearing in mind their inbuilt assumptions and biases. This is laudable, as is the aim to highlight methods applicable across a wide range of humanities. There is also a much repeated concern to suppress “numerical fetishism” (“the belief that reality is best summed up by numbers, regardless of how those numbers were arrived at” (p. 4)). It is hard to argue with that.
The first chapter covers the history of quantification. It charts the rising tide of the 1970s, critiques and crises of the 1980s, the relative doldrums of the succeeding two decades and resurgence in the form of the digital revolution and its quantitative potential. Later chapters consider sources and sampling, the converting of untidy and complex source materials into more neatly categorized data in rows and columns, leaving those categories flexible and open to new questions as they arise (often from data inputting itself). The strength here is the familiarity of the authors with a wide array of relevant examples from published historical research, particularly on France and the United States. Chapters 4 and 5 cover correlation, the assessment of causality, regression and factor analysis followed by sections on network analysis, “trajectories,” visualizations and word counting. Thanks to a translation produced by Harvard academic and award-winning translator Arthur Goldhammer, technical terms, procedures and arguments have been processed into smooth and lively English prose. The volume is thus readable and accessible; a major achievement given that most chapters take us through a selected myriad of possible techniques, their pitfalls, potential and exemplars in whistle-stop fashion.
The authors state in their introduction that they “do not explore the mathematical principles underlying each method” (p. 4). They nevertheless necessarily trespass into doing so here and there in partial fashion because it is not easy to relay the advantages or the disadvantages of many techniques and their goodness of fit for interrogating certain sources and posing particular research questions without simultaneously addressing the nature and pitfalls of statistical theory underlying them (as with their treatment of random sampling and the Chi-Squared test). This sometimes makes for dense reading as in chapters 4 and 5. And in a volume aimed to attract the uninitiated it seems surprising that the opportunity to emphasize simple purposes of statistical manipulation (both pros and cons) in conveying information in summary form in graphs, charts and other diagrams, is underplayed. The first figure in the book appears on page 77 and they are infrequent thereafter even within the short chapter devoted to “visualizing history” where there is undue concentration upon a narrow range of network diagrams to the exclusion of others.
The main criticism that one might levy at the range of statistical techniques and areas of research discussed is that most concern static or cross sectional analysis. Time series take up just a few pages in a section where the focus is event history, within a chapter devoted largely to networks and research using collective biographies. From my own perspective steeped in economic and demographic history is seems surprising that diachronic research in the humanities and especially in mainstream history appears to have such a limited following compared with synchronic analysis, a bias that is reflected in this book. The final chapter on exploring texts is also disappointing. It is by far the shortest chapter at just twelve pages and although it covers an area entirely compatible with humanistic interpretation it is unduly limited in its appreciation of the potentialities of computerized quantitative work in this area in identifying authenticity and authorship for example and in examining the persuasive messages conveyed in vocabulary choices and repetitions: areas important in understanding the social, political and human impact of social media in the present day.
Overall, this is a useful book well worth recommending to students for certain purposes but readers may be disappointed that a full appreciation of the techniques discussed will depend upon learning them from other texts and that the texts recommended are written by economic historians relatively unconcerned with other humanities subjects. The authors also promise to highlight “nonstandard” approaches and to reveal opportunities for methodological innovation but these are hard to trace in the text. Finally, despite a plea at the start for quantification in East Asian and West African history there are few examples in the text from anything that might be described as world history or global history and no discussion of the pitfalls of comparing data across space as well as time. This is perhaps unsurprising given the sparse and complex nature of sources relating to non-western regions and the difficulties of integrating evidence generated by entirely different cultural and political traditions into the corpus of western science.
Pat Hudson is Emeritus Professor at Cardiff University, UK. She is joint author (with Mina Ishizu) of History by Numbers: An Introduction to Quantitative Approaches (London: Bloomsbury, 2017) and is writing a new history of the Welsh Woollen industry using qualitative and quantitative research methods applied across the spectrum from artefacts to economic analysis.
Copyright (c) 2019 by EH.Net. All rights reserved. This work may be copied for non-profit educational uses if proper credit is given to the author and the list. For other permission, please contact the EH.Net Administrator (email@example.com). Published by EH.Net (October 2019). All EH.Net reviews are archived at https://eh.net/BookReview.