Critical realism and Comparative Sociology

With the rise of critical realist social science time has come to ask whether this approach can enhance comparative methodology in sociology? The paper contends that the answer contains two paradoxes. Although critical realist sociologists often argue in favour of comparative studies, they are yet to formulate a genuine comparative methodology (first paradox). They refer to the comparative tradition without scrutinising it from a critical realist perspective. When such an assessment is made of core comparative methodologies the conclusion is that post World War II comparative sociology has important methodological flaws being rooted in Mill’s inductive logic (1843) and empirical-analytical science philosophy (second paradox). A critical realist comparative methodology must reveal both substantial and contextual features of the phenomenon being researched, and even the most promising comparative methodologies do not integrate these two aspects in a single methodology. However, the conclusion is that a pertinent and critical realist comparative methodology may emerge from streams of contemporary comparative sociology if critical realists duly prioritise the issue of comparative methodology. 1 The paper is based on Wad 2000a, 2000b and 2000c. Introduction Critical realism presents a methodological perspective which takes a critical stance towards positivism and hermeneutics on the one hand and tries to integrate strong points from both theories of science on the other hand (Wad 2000a, 200b). Contemporary critical realism is formulated by Roy Bhaskar and his like-minded circle of British colleagues (Archer et. al. 1998). In the Nordic countries, critical realism has been promoted in Sweden as a new philosophical platform for sociology from the mid-1990s (Djurfeldt 1996, Danermark et.al. 1997). In Denmark, critical realism has been adhered to by several geographers and political scientists, but it has not taken the central position it deserves being a challenging philosophy of science (see thematic issue of Grus nr. 60, 2000, edited by Ougaard). The ‘newness’ of critical realism is, however, contended (Johansson & Lind 1999), and in this paper I want to discuss whether critical realism adds new insights to a more limited field, i.e. the comparative methodology of social science in general and comparative sociology in particular. Answering this core question we must address five interrelated questions: First, what is the aim and principles of critical realist social research? Second, does critical realist sociology give priority to comparative studies? Third, what constitutes a critical realist comparative methodology? Fourth, does the prevailing comparative methodology meet the requirements for critical realist research? Fifth, if this is not the case, can we construe such a methodology from the arsenal of existing comparative sociological methods, or do we need to invent from scratch? The first question will be answered in a very crude manner, assuming that the reader as a IACR conference participant is more or less familiar with the philosophy of critical realism. The section will mainly function to shortly clarify the author’s interpretation of this philosophy. The second question is addressed by discussing the status of empirical research in critical realist social science before we turn to the more specific issue of the position of comparative empirical studies within this school of science theory. If comparative studies are considered important then the next and third question is what kind of comparative methodology is relevant, appropriate and valid in a critical realist perspective? Is comparison in critical realist studies based on particular methodological principles or is it similar to comparative methodologies in other theories of science? Arguing that critical realism demands a specific methodology for comparative social science, the fourth question arises whether the mainstream tradition or counterpoints of comparative sociology comprises this type of comparative methodology. If this is not the case, and that is what the paper argues, then we face the fifth and final question, how can we develop such a methodology? Do existing strains of comparative methodological thinking and practice provide us with a useful platform or points of departure, or do we need to turn against the sociological legacy and invent a new comparative methodology? The paper is not about comparative sociology as such, but only about the methodology of that