This is the first in a series of guest posts by Stuart J. Kaufman of the University of Delaware. Stuart advances a long-running dispute with PTJ about whether “what goes on inside people’s heads” is relevant to social constructionism. PTJ doesn’t think so; Stuart disagrees. After the final post, we will make the entire piece available as a PDF — consider it our first true “working paper” publication.
When scholars hold up physics as a model for how social science should operate, they tend to focus on the neat equations that physics is able to generate to explain physical phenomena. A more useful lesson to learn from physics, however, is how it deals with the multiplicity of sources of physical effects that are often described by different physical laws. Basically, it adds them up.
Consider the forces that keep a star stable. On the one hand, a star is powered by the process of nuclear fusion of lighter elements such as hydrogen into heavier elements, with their nuclei bound together by the strong force. This fusion releases massive amounts of energy, which generate electromagnetic forces that push the star’s matter outward, in some cases leading to massive stellar explosions (novae and supernovae). Most of the time, however, stars do not explode, as the gravitational forces that make nuclear fusion (and thus the heat) possible also hold the matter in the star together. Understanding how these processes balance out, however, requires understanding of at least three (really all four) of the basic forces in physics—the strong, weak, electromagnetic, and gravitational forces—and how they interact, even though they operate according to different physical principles.
Unfortunately, debates in political science and international relations tend to be less constructive. One side says, “the ontological nature of reality is the dominance of the electromagnetic force; ergo the sun’s heat will cause it to explode”. The other side asserts, “the ontological nature of reality is dominated by gravitational forces, so the sun must collapse.” Both sides agree that because of these competing ontologies, no synthesis is possible.
Real debates in political science and I.R. have the same “blind man and the elephant” quality. Because neorealism and most versions of constructivism are structural theories, their ontologies allegedly rule out consideration of domestic political or—horror of horrors—individual levels of analysis. Any departure from strict structuralism is scorned. For example, the standard constructivist theory of norm change, Krebs and Jackson (2007, p. 40) scold, “relies on incompatible microfoundations in stage one (instrumental adaptation) and stage two (internalization).” As one constructivist scholar flatly declared to me, “I reject methodological individualism”.
It seems obvious to me that this attitude is both profoundly illiberal and profoundly anti-intellectual. If liberalism requires not just toleration but respect for people who think differently, then such a priori rejectionism should be normatively out of bounds for any decent person, and doubly so for a self-respecting intellectual. More to the point, unless scholars are prepared to defend the position, “all individual psychology is bunk, and all psychologists who write about it are wrong,” then they have to concede the likelihood that psychologists do indeed know something about human behavior, and some of what they know is relevant to the study of political science and international relations. The same is true of sociologists, anthropologists, historians and even economists.
One of the things psychologists know, for example, is that—pace Krebs and Jackson—there is no incompatibility between the processes of instrumental adaptation and internalization. On the contrary, as has been widely shown and even explained by political scientists, just this process is psychologically common. People often react instrumentally to specific situations, and then infer their broader preferences from their own behavior—preferences they then internalize. Larson (1985) showed that just this process, as hypothesized by cognitive dissonance theory, was operating among members of the Truman Administration during the run-up to the Cold War: they first took a few hard-line decisions, then internalized the hard-line anticommunist norms. Confronted with the cognitive dissonance between their uncertain attitudes toward the Soviets and their tough decisions toward them, they resolved the dissonance by adopting the hard-line values implied by their own behavior.
At the root of the Krebs/Jackson mistake is the silly assumption that theoretical assumptions imply an exclusive ontology. They do not; they are metatheoretical judgments about which aspects of reality—which partial ontology—is most useful for answering a particular question. In some of my work (Kaufman et al. 2007) for example, I accept basically realist premises to explain the politics of the ancient Middle East, generally endorsing hegemonic stability realism over balance-of-power realism. In doing this, I do not accept the notion that the fundamental nature of international reality—the ontology of the international system—is fully described by realist thinkers like Robert Gilpin or Kenneth Waltz. Rather, I make the metatheoretical judgment that for this particular time, place, and issue (balance or hegemony?), these alternative realist theories identify most of the key variables for explaining the phenomena under consideration. My sympathy for liberal and constructivist insights comes in large part from my recognition that the contemporary world system is different from the hyperrealist ancient system precisely in that the role of international norms, international trade, international institutions, and internal liberal democracy is so much greater now than then.
These metatheoretical judgments are based, in turn, on the implicit or explicit conditioning hypotheses that come attached to all theories. If neoclassical realism truly explained the ontology—the essential nature—of social reality, then John Mearsheimer would take literally his own injunction to be “the biggest guy on the block” by spending much of his time in the gym bulking up. He would spend the rest of his time endeavoring to become the successor to Al Capone in Chicago on the theory that only armed force ensures physical security. The fact that he does neither shows that Mearsheimer, uber-realist though he is, recognizes a critical scope condition for realist theory—it applies to the anarchical international realm, not the hierarchical domestic one. In his writings, he makes the metatheoretical judgment that anarchy itself contradicts the scope conditions of liberalism; thus he does not assume that humans or human groups are ontologically incapable of cooperation. He merely asserts that stable cooperation requires an authority to enforce it. I think he is wrong about that, but it is not an exclusive ontological argument about the nature of reality.
The way forward in political science and I.R. theory, then, is to recognize that all social and behavioral sciences have some valid insights to offer; that all of those insights are limited by both the quality of the supporting evidence and the restrictions of the theories’ scope conditions; and that complex social realities can only be adequately understood if we can work out how to draw on multiple sources of insights to explain them.
The ontological fallacy of Krebs and Jackson, inter alia, is to assume that social theory must be based on a single, simple assumption about human nature: the human is a fearful creature (realism); the human is a cooperative creature (liberalism); the human is a social creature (constructivism); the human is a creature of habit (Bourdieu); the human is an emotional creature (motivational theory); the human is a rational utility-maximizer (rational choice theory); the human is a biased decision-maker (cognitive theory). It is obvious that all of these assertions are sometimes true of some people, and that all also offer good descriptions of the behaviors of certain groups at certain times, because all have some good social scientific evidence in their favor. It should therefore be plain that none alone is likely to offer a satisfactory basis for a theory of any complex behavior. Metatheoretical choice is not ontology. What we need is to define the scope conditions and implications of theories that harness each of those insights.
To be continued…