Clemans on Wilkinson, 'Deadly Quarrels: Lewis F. Richardson and the Statistical Study of War'
David Wilkinson. Deadly Quarrels: Lewis F. Richardson and the Statistical Study of War. Berkeley: University of California Press, 2018. 216 pp. $39.95 (paper), ISBN 978-0-520-30381-2.
Reviewed by Paul Clemans (Air University, Air Command and Staff College) Published on H-War (February, 2021) Commissioned by Margaret Sankey (Air University)
Printable Version: https://www.h-net.org/reviews/showpdf.php?id=55375
David Wilkinson’s Deadly Quarrels (1980) represents a full-length examination of an earlier work by Lewis Fry Richardson, titled Statistics of Deadly Quarrels (1960). Richardson had conducted and published an in-depth statistical study of factors that potentially contribute to the causations of war and the conditions of peace. Many subsequent scholars incorporated his factor analysis into their studies. However, Wilkinson believed the more creative parts of the work had been neglected, such as Richardson’s coding, counting, and distribution models. Wilkinson’s desire to see the political science field more fully embrace these aspects of the work provided the impetus behind his Deadly Quarrels. In short, Deadly Quarrels offers an assessment of what Richardson’s Statistics of Deadly Quarrels was intended to do, what it did, what others have done with it, and what is potentially left to be examined.
Deadly Quarrels affords a broad perspective of the significant scholarly works on the causations of war and conditions of peace up to 1980. The field was relatively young up to that point in time. In fact, Wilkinson credits Richardson and Quincy Wright, author of A Study of War (1965), as the two pioneers responsible for initiating studies into the causes of war a mere twenty years earlier. Wilkinson also incorporates at least a dozen significant scholars who leveraged Richardson’s data in their own studies up to that point, such as Frank Denton, William Horvath, and George Modelski, among others. Wilkinson generally provides a rich, even-handed treatment of these scholars’ alternative interpretations throughout the volume, usually by way of comparison. Although, he takes issue with several aspects of J. David Singer’s and Melvin Small’s works. Wilkinson registers disagreement specifically with their decision to omit civil wars and to not publish their coding and, more generally, their theoretical imagination. His critiques call into question the two scholars’ judgment calls regarding the data and, perhaps, their more conservative approach. It is worth noting that Singer and Small offered one of three war list data sets at the time, the other two coming from Richardson and Wright.
Wilkinson starts the volume with some welcome insights into Richardson’s background and motive for publishing his original study. Often a scholar’s previous experience and disposition are absent from the discussions, despite the many subjective choices made in studies of this breadth and depth. That is not the case here; Wilkinson includes a good brief description of Richardson’s religious background and strong statistical expertise in other fields, especially in meteorology. The description helps scholars to gauge whether Richardson’s choices are being driven by the quality and limitations of the data or personal persuasion. That determination may help scholars decide what aspects of the datasets they deem worthy of further examination or reinterpretation. For example, Wilkinson characterizes Richardson’s decision to not pursue religion factors further as a “conservative” decision (p. 89). Given that the data covers over three hundred wars, 780 belligerent pairs, and dozens of factors over a 130-year period, scholars may feel similarly about other factors based on the short biography in the beginning.
The chapters form an intuitive reevaluation of Richardson’s original study. The first chapter briefly, and compactly, covers Richardson’s purpose and assumptions. Wilkinson clearly expounds on Richardson’s terms and explains his methodology in the second chapter. Although analysts will appreciate the detail in this chapter, only a basic grasp is necessary to understand the concepts discussed in later chapters. The next eight chapters examine sets of factors with associated discussions of Richardson’s original analysis, Wilkinson’s assessments and interpretations, other scholars’ interpretations, and potential areas of future study. A discussion of Richardson’s strengths and weaknesses rounds out the chapters with an emphasis on his degree of success in achieving his Jominian aims. The conclusion contains little more than the chapters’ factor assessment summaries. The following nine appendices are half the size of the book and are chock-full of data and further analysis. Richardson’s full war list, the factor coding definitions, the raw war coding data, and the 780 belligerent pairing list may be found there. Discussions of the Wright and Singer and Small war lists may be found there as well.
At first, it appears Deadly Quarrels only evaluates factors with an admirable attention to an extended historiography, but Wilkinson does develop an argument based on Richardson’s factors. Halfway through the volume, the argument begins to congeal with the “participation of named nations” factor. Wilkinson reaches the finding that conditions of peace are most likely reached with “a realist/conservative peace strategy of pacifying, or to a utopian/radical strategy of liquidating, the great powers” (p. 70). From here, the work moves from examining the factors affecting the causes of war to seeking means to impose conditions of peace. The volume subsequently turns its focus toward schemes that might effectively achieve this pacification. Theories of prosperity, religion, language, modernization, and cultural similarities and their relationship to war are considered in light of these schemes. While not discounting any of the other factors, Wilkinson settles on prosperity and homogenization as the two significant factors leading to pacification and conditions of peace.
The logic Wilkinson uses to arrive at this conclusion and this strategy is satisfying with a few exceptions. Richardson assumed that given a large sample size he could overcome mistakes in factor categorization of any individual war. He did not consider that he may make the same categorization mistake across all wars, leading to systemic errors. Whether or not this occurred, and those errors exist, is left to the discretion of the reader. A second concern emerged from the occasional tendency to use numbers as opposed to percentages over time. For example, when considering the increasing or decreasing “warlikeness” of nations, Wilkinson compares five million deaths from 1820 to 1884 to forty-two million deaths from 1885 to 1949, that is, the deaths for two separate sixty-four-year periods (p. 26). However, the population was not the same and not constant for either of these periods. The use of percentages may be more appropriate here, but even percentages may be inadequate given the variations in population growth rates over such a duration. Finally, and curiously, Richardson and Wilkinson excluded the contribution of private manufacturers and arms traders from the list of potential economic causes. While reconsideration of these factors may not change the main finding or the subsequent factor analysis, they should have been given greater examination. That said, any such shortcomings would further the purpose of Wilkinson’s volume—to spur further research and assimilation in the field!
The value of Deadly Quarrels today lies in the logic leading to Wilkinson’s strategy for maintaining the conditions for peace. Wilkinson evaluated dozens of factors using quantitative data to identify the participation of the nations in war as the pivotal factor. The reasoning and choices he made to arrive at this determination is worth considering, even today. Also of value may be the consideration of his reasoning to discard other factors and alternative conclusions of other scholars. Wilkinson presented his work as an early effort to scientifically derive a promising direction for peace from qualitative data, but after forty years it remains valid. His ideas of pacification, prosperity, and homogenization can be seen and are valued in the world today.
Wilkinson wrote this volume for the political science community, specifically peace studies scholars. While the work itself has been overtaken by further developments in the field, the slim volume would be a good introduction into the subject’s multifaceted considerations. It should be noted that the volume altogether is over two hundred pages, but the appendices provide all of Richardson’s raw data beginning on page 122. The raw data offers a ready-made project for students to analyze for themselves. A wider audience would receive edification from the work as well with a basic familiarity of statistics. In either case, Deadly Quarrels most likely will be used as a survey from which to delve deeper into particular factors of study or relationships.
. The book under review is part of the University of California Press’s Voices Revived series, and is a re-issue of the 1980 work.
Citation: Paul Clemans. Review of Wilkinson, David, Deadly Quarrels: Lewis F. Richardson and the Statistical Study of War. H-War, H-Net Reviews. February, 2021. URL: https://www.h-net.org/reviews/showrev.php?id=55375This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.