The appropriate methods for studying international relations have long been contested and controversial. Although, having said that, in the mid 1960s when I did an undergraduate IR degree in Britain and then a masters degree in IR in the United States I have no memory of the issue of methods ever being raised or discussed. More alarmingly, when I returned to Britain in 1968 to start my PhD I was left entirely to my own devices when it came to questions about how to carry out research in this field. By then, however, I had gleaned that there was an ongoing debate in the IR literature between behavioralists, who argued that it was necessary to adopt a methodology that was consistent with the methodology employed in the natural sciences and traditionalists, who argued that the analysis of International Relations was not susceptible to the methodology employed by natural scientists.
This debate was epitomized in what became a very widely cited article by Hedley Bull in World Politics in 1966 where he defends what he called the classical approach to International Relations against what he viewed as a putative scientific approach. He insisted that he was talking about a scientific approach rather than a scientistic approach, a term of opprobrium, so as not to prejudge the issue. However, now, over 50 years after the article was first published, what is most curious is how Bull argues that while scientists have helped, although only marginally, to advance our understanding of International Relations, they were far from operating on a united front. He notes not only that scientific studies in International Relations have varied enormously in terms of the methods that they employ but also that the advocates of the different methods all too often regarded each other with implacable hostility.
From Bull’s perspective, moreover, behavioralists adopting a scientific methodology only provided answers to questions that are relatively trivial for our understanding of international relations. When they endeavored to answer the more important questions that are central to the field he asserted that they inadvertently reverted to the traditional methods employed in the classical approach and, all too often, used such methods very badly. At the end of the article, however, Bull qualifies these bold and somewhat acid assertions by acknowledging that there were some behavioralists who had conducted research that has made a substantial contribution to the field but when they have done so, they have employed methods that can easily be accommodated within the classical approach.
Unsurprisingly, Bull’s article precipitated a number of dyspeptic responses from well-known behavioralists of the time. But what is most surprising about the article is that while Bull readily gives examples of behavioral methods, from game theory to content analysis, he is remarkably coy about identifying the nature of the methods most closely associated with a classical approach. Nevertheless, the idea that there are two competing methodological approaches to the study of International Relations and that there are different and distinct methods associated with these divergent methodologies had a significant impact on thinking within the discipline at that time.
Although I read round the literature on this emerging debate, when starting my PhD, I have to say that I did not find it very helpful when it came to thinking about how to do my research. I was interested in the norm of non-intervention and, in particular, the circumstances under which decision makers decided to violate the norm. I decided to focus on situations where the violations were unequivocal and concluded that sending troops into an on-going civil war provided the best example of such a violation. I decided to focus on Britain because of its central importance in the consolidation of the modern international society. I started by identifying every instance where Britain had sent troops into a civil war during the course of the nineteenth and twentieth centuries. This tactic, however, generated too many examples given the approach I had decided to adopt. I wanted to develop a model to account for norm violation and this required me to look in great detail at how decisions were taken and so, as a practical necessity, I needed to restrict my focus to only a very limited number of cases. I agonized for some time about how to select these cases but eventually focused on four cases drawn from across the nineteenth and early twentieth centuries. I chose a civil war in Portugal in the 1820s where Britain sent troops to support the existing regime, the American Civil War in the 1860s where the British decided against intervention, the Russian Civil War at the end of the First World War where Britain sent in troops alongside a number of allies, and the Spanish Civil War in the 1930s where Britain decided against intervention despite the interventions by other European states. These cases were selected because I thought that they looked interesting and encompassed the period I was focused on. Nevertheless, I worried that I could not give a more robust and logical reason for the choice of these particular cases and was relieved that the issue was not raised in my viva.
Despite finding the distinction drawn between the scientific and classical approaches unhelpful, it is worth noting that I then assumed nevertheless that I was operating on the basis of a scientific methodology. The assumption was justified to the extent that when the manuscript was refereed for publication in the United States, Stanley Hoffmann said that I was obviously a young man “intoxicated by the social sciences”. This was not intended as a compliment but the comment did reflect how deep-seated the bifurcation of approaches was at that time. Yet, retrospectively, the bifurcation looks increasingly unhelpful. My empirical data was based entirely on very extensive archival research and I mined the archives for data that could illuminate the processes that promoted or at least permitted the violation of a central norm that helped to maintain the structure of the international society. Half a century after starting this research it is no longer clear to me that it is meaningful or helpful to insist that the research was operating on the basis of a scientific rather than a classical methodology – or indeed vice versa. It seems like a false dichotomy.
In the intervening period since I originally worked on intervention and non-intervention there have been some very significant developments in the discipline. One that I find particularly instructive is the emergence of Constructivism, an approach that focuses on the socially constructed nature of the world we inhabit. The approach resonates with the research that I carried out all those years ago and undoubtedly my project would have been helped enormously if constructivist ideas had been available. On the other hand, I do not think that I would have adopted different research methods. I would still have focused on a very small number of case studies and I would have unquestionably relied on archival research. From my perspective, once I had decided on the basic orientation of the research, it essentially determined the kind of research methods I had to adopt.
This conclusion chimes very closely with most of the chapters in this very intriguing book. What the editors have done is to ask the contributors to reflect on the process whereby they arrived at the methods they used in a major piece of constructivist research. To some extent, I have also followed their injunction although in the context of an era before the emergence of Constructivism. While the book is not a primer or manual, the reader will undoubtedly learn a lot about constructivist methods along the way. A method should not be viewed as a hammer to be picked up at the start of a research project. With a hammer in hand, there is a danger that the researcher will seek out a nail to hit. What this book reveals is that appropriate methods emerge in the process of conducting research.
Richard Little
Emeritus Professor
University of Bristol