| From https://www.zazzle.com/ |
This is the last 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. The first post can be found here, the second here. You may also download a complete PDF.
None of this is intended to deny the importance of structural insights offered by constructivist analysis. The argument, rather, is that “psychology provides the microfoundations for the motives behind normative behavior and identity change” examined by constructivist analysis (Shannon 2012, p. 14). Rhetorical coercion is an important mechanism, one I suspect underlay the inability of liberals and thoughtful moderates to articulate a resonant alternative to Bush’s “war on terror” narrative. But it is not the only mechanism of importance. As Kowert (2012) argues, norms are socially constructed, but they require that norm-holders both believe that something is right or wrong, and that they care about the outcome at stake. Understanding norms therefore requires understanding the “ideational triangle” of cognition, norms and social construction. I would be inclined to make it a “quadrangle” to include the pivotal influence of emotions.
Furthermore, for many of the issues in which discursive norm-production is important, there is yet another mechanism whose impact cannot be overlooked: the role of social networks. Neither constructivist discourse analysis nor individual or group psychology is very useful for explaining who becomes active in social movements and who does not. To explain who is likely to join a protest movement or a rebellion, we must look to social network theory as articulated most prominently by Tilly (2005, e.g.). The people who join social movements or rebellions are not consistently the people who feel most strongly about the issue at stake a priori; it is the people with the closest personal ties to those already involved. Explaining the rise of social movements, therefore, requires following Tilly’s insistence on looking to the “social appropriation” of existing institutions for social mobilization; to the brokers who create links between diverse social networks; and other similar mechanisms.
The result of taking seriously the importance of all of these different literatures would be a set of interlocking theories in which each piece of the puzzle fit into its neighboring pieces, each mutually supporting the other. Individual-level psychology would provide foundations for assumptions about human motivation and action tendencies—or, more precisely, which motivations are important when—but would then fade into the background as the focus of analysis shifted to social (including rhetorical, social psychological and sociological) mechanisms involved in political life. The role of the theorist is creatively to link the findings of these disparate fields into more or less coherent explanations of specific phenomena, instead of starting from ad-hoc assumptions considered risible in other disciplines. This approach is not too different from what constructivists typically do now, bracketing issues of agency to focus on discursive structure. I am only calling on constructivists to be more psychologically aware in the assumptions they make.
The centerpiece of this approach, then, is to identify when different modes of analysis are appropriate. For example, from a psychological perspective, the hypotheses of bureaucratic politics theory (“where you stand depends on where you sit”) is easily explicable in terms of the well-known mechanisms of socialization, commitment and role assignment. Bureaucrats advocate their organizations’ interests, that is, because it is their job to do so (role assignment), because once having done so, they feel committed to those values (commitment), and because they come over time to be socialized into those values by their senior colleagues.
Furthermore, attention to these psychological mechanisms helps to explain not only when bureaucratic effects are most important, but also when they are less so. For example, Rhodes’s (1994) finding of the insignificance of intra-service rivalry within the U.S. Navy (between airmen, submariners and surface warriors) should come as no surprise, because naval officers’ primary socialization (and training) is into the navy, not any particular “union” within it. Furthermore, Rhodes is analyzing the behavior of Chiefs of Naval Operations, whose role assignment is to advocate the interests of the entire Navy, not their “union”. On the other hand, these same mechanisms suggest that, especially before the 1986 reorganization, the Joint Chiefs of Staff should have fought like cats and dogs along bureaucratic lines, because the mechanisms of socialization, commitment and role assignment all pointed in that direction, yielding in turn cognitive biases and motivated biases all pushing the Chiefs each to defend his own service’s interests and values, as bureaucratic politics theory would suggest.
The same holds true of rational choice theory. While the assumption that people are rational utility-maximizers is almost always false and is often unproductive, there are circumstances in which social psychology would predict such behavior. Institutional behavior is again the clearest example. The mechanisms of socialization, commitment and role assignment yield the expectation not only that bureaucrats will pursue their institutional interests, but that lobbyists will lobby for their employers; and, indeed, individual self-interest is likely to point in the same direction. When individual self-interest does not align with institutional interest, rationalists rightly point out, greed or ambition may trump socialization, leading to shirking or corruption. Rationalists don’t talk about greed and ambition, but as long as those motives accord with assumptions of individual self-interest, the difference does not matter. Assuming bureaucrats will behave like bureaucrats is not psychologically dubious.
Outside institutional bounds, in contrast, people usually do not act to maximize their material interests in politics because their judgment is at different times driven primarily by fear (explained terror management theory), group identity (explained by social identity theory), bias or prejudice (explained by cognitive bias or prejudice theory), motivated bias (explained by motivational theory), personal connections (social network theory), the desire for self-expression, or a host of other motives.
It seems plain, then, that interlocking theory provides the only way to move forward in international relations and political science, based on using the findings of allied disciplines. In international relations, the answer to the paradigm debate lies in determining under what conditions key actors behave like realists, or liberal institutionalists, or domestic politics liberals. What are the conditioning hypotheses for each theory? Again, a number of different factors play a role, each explaining a different piece of the puzzle. From this point of view, constructivism in international relations functions as a partial metatheory, pointing out that sometimes international actors behave like realists (“Hobbesean” systems), sometimes like international liberals (“Lockean” systems), and sometimes more like liberals in a domestic setting (Kantian systems). The trouble is that constructivist analysis is terribly thin in identifying when each sort of behavior should occur.
Again, social psychology provides a host of suggestions for how to sort these questions out. Realists note that for their theory, fear is the driving force, and indeed, terror management theory essentially explains why people behave like realists when they feel under threat. But when do they feel under threat? Personality has something to do with it, with trust playing a huge role: only relatively trusting people are inclined to behave the way liberal institutionalism would predict (Rathbun 2011). On the other hand, those who are less trusting tend to see the world as a competitive place—a syndrome identified as “social dominance orientation” (Sidanius)—and to respond aggressively to challenge. Ergo, the hypothesis: states led by people with social dominance orientation are likely to behave in realist fashion; those led by more trusting individuals are more likely to act as liberal institutionalists predict. Prejudice also has something to do with it: people are more likely to perceive threat when they hold negative stereotypes of the source of potential threat, and when they have negative emotional feelings about that outgroup. This approach also helps to explain why past behavior matters in some cases but not others: prejudiced leaders will tend to discount evidence of moderation on the part of the target of their prejudice (e.g., Cold War anti-Communists), and therefore act competitively.
To break down a more specific example, asking why the U.S. behaved like a neoconservative sort of realist toward Iraq in 2003 is actually asking a set of distinct questions, each of which has answers in a different area of theory. From an institutional perspective, there were at least three veto players regarding a war with Iraq: the President, his party, and Congress, with Congress’s position in turn partly dependent on public opinion. Therefore, to explain the war, we must explain the positions of all three veto players. First, George W. Bush decided he wanted to invade Iraq for a variety of reasons explained by personality theory (such as his ethnocentrism) and small-group dynamics (e.g., groupthink). Second, his party enthusiastically supported this course due to a combination of prejudice, institutional incentives in the party, and calculations of electoral advantage. The calculations of electoral advantage, in turn, depended largely on intuitive understanding of prejudice and terror management theory—how threat perceptions and anti-Saddam bias after 9/11 drove public opinion to the right on the issue of the war. Finally, as discussed earlier, a combination of constructivist and psychological factors explain why Democrats in Congress felt they had to go along with the Bush “War on Terror” narrative and vote in favor of the war—and why that course was popular with voters.
The final element of truly progressive theorizing, as suggested by these examples, is attention to the balance of counteracting forces. In the astrophysics of stellar stability, all of the interest is in the balance between the gravitational forces holding the star together and the countervailing forces pushing its mass outward. Similarly, almost any problem in contemporary international relations is likely to be driven by some factors emphasized by realism, some emphasized by liberal institutionalism, some by domestic politics liberalism, and some by constructivism. In the battle for public opinion over the Iraq war, for example, international institutional constraints—notably the position of the UN Security Council—were manifestly significant in constraining the march to war, yet were ultimately swamped by other factors pushing the other way. Meaningful theory means thinking about how to measure these counteracting effects, not simply assuming some of them away.
Parsimonious theories of politics are possible, of course, but not parsimonious theories that work. If we want to achieve anything like scientific progress, we need to put aside debates about which paradigm is best, and begin focusing on when each paradigm best applies, to what degree and in which circumstances.
What a great set of posts. Looking forward to PTJ’s response…
My own response coming soonish.
‘None of this is intended to deny
the importance of structural insights offered by constructivist
analysis. The argument, rather, is that “psychology provides the
microfoundations for the motives behind normative behavior and identity
change” examined by constructivist analysis.’
This perspective rests upon a certain philosophy of the mind and a certain understanding of structure. Basically, it presumes that constructivism requires the assumption that a) structures are made out of individuals and b) individuals do things because they have motives, so understanding action requires knowledge of those motives. And it is true that much constructivist scholarship does proceed from these assumptions.
But it is possible to study society, and human behaviour, without those assumptions. It is possible to study situations without imagining that they are made up of individuals with discrete minds, each possessing a certain psychological state. It is possible to imagine that individuals do exist, and yet not need to have them in the model. It is possible to imagine that individuals are themselves entirely social constructions, no different in essence than what we might call ‘macrostructures’.
Perhaps you want to exclude work such as this from the label ‘constructivist’, but if you do, you end up with a large collection of ‘orphans’ that would otherwise self-identify as constructivist, and conflate the community of scholars interested in looking at how social reality is constructed with one particular methodology for doing so. This strikes me as unnecessarily restrictive.
On the other hand, if you want to make the point that a great deal of constructivist scholarship presumes and assumes certain psychological processes in order to render its theories coherent, I think you’ve done that quite well. But if you think that Krebs and Jackson fall within that category, then you haven’t read their article in the same way that I have, that’s for sure.
“It is possible to imagine that individuals are themselves entirely social constructions”Well, Simon, it’s possible to conceive of a lot of things, but most of those conceptions aren’t true. Maybe you consider yourself entirely a social construction, but that’s not what I understand people to be. I’m trying to be scientific here, and the science is pretty clear that people are NOT just social constructions, but the product of a complicated interplay of nature and nurture. People have personalities, ideologies, ideosyncracies, biases, and as a matter of fact, they respond differently to structural pressures. Some people read jidahist discourse and are repulsed, while others strap on an explosive vest and blow themselves up. Structure without agency–without people–is a meaningless abstraction, and not very productive as social science.
That said, I DON’T think structures are made out of individuals. Structures arise from the interaction of individuals, groups, institutions, etc. I’m a mainstream constructivist on that. Emergent properties and all that. My whole point is that you don’t have to choose between structure and agency, theorizing only about unit or system. Sure, at times you bracket one to focus on the other, but you need an overall conceptualization that embraces both and explains how they fit together. People have agency, and the autonomy to react differently to structure.
I also don’t insist on explaining people’s “true motives”; I don’t know those either. What is knowable is certain psychological tendencies–some universal, some specific to certain populations like right-=wing authoritarians. Also knowable is that people have enduring biases: think, for example, about how predictable most Supreme Court justices are. These are measurable through survey research, psychological experimentation, and yes, discourse analysis.
My argument, in sum, is that more anchored constructivism is in social psychological and neuroscientific knowledge, the more productive it is likely to be. More broadly, the more political science and I.R. scholars look to bridge disciplinary knowledge, the more solid their theorizing will be. That’s the point about interlocking theory. If some constructivists want to do it differently, they’re free to do so. I just think they won’t get as far; and t hose using purely interpretivist methodologies are leaving themselves wide open to personal bias
‘I’m trying to be scientific here, and the science is pretty clear that
people are NOT just social constructions, but the product of a
complicated interplay of nature and nurture.’
We may be talking past one-another here. When I say that it is possible to conceive of people as being social constructions, what I mean is that we can examine those personalities, ideologies, ideosyncracies, and biases as social things or properties. I am not suggesting that the physical organism homo sapiens doesn’t have traits that influence its behaviour, Rather, I’m suggesting that there is something essentially social, and entirely social, to how we observe, define, and analyse that behaviour. I think that this conception has a strong philosophical tradition behind it, and I think that Peter Winch’s Wittgensteinian manifesto ‘The Idea of a Social Science and its Relation to Philososphy’ provides a good foundation for a scientific ontology made of nothing but the social. That’s not discussing the post-structuralist traditions, of course.
‘That said, I DON’T think structures are made out of individuals.
Structures arise from the interaction of individuals, groups,
institutions, etc. I’m a mainstream constructivist on that. Emergent
properties and all that.’
You seem to be talking about a certain kind of emergence here, though: strong emergence. The idea that emergent entities and properties are causally irreducible to the micro-potentialities upon which they supervene. I know that this vision of emergence has some good philosophers amongst its proponents, but my understanding is that most of those proponents view consciousness as the only truly emergent entity. David Chalmers is one of these. One of the most preeminent philosophers on emergence and supervenience, Jaegwon Kim, has written in several places that while he views emergence as an important concept, it should not be understood as anything more than explanatory, rather than metaphysical. In short, that emergent entities are not natural kinds, and they are not irreducible parts of really real reality.
And of course, the question remains, what are groups and institutions – those things which in addition to individuals comprise structures, according to your claim – except types of social structures?
‘People have agency, and the autonomy to react differently to structure.’
I don’t think this needs to be part of a social scientific methodology, these assumptions. I mean, they are part of many excellent methodologies, but I have read stuff that I’d call ‘good social science’ which doesn’t assume them.
‘My argument, in sum, is that more anchored constructivism is in social
psychological and neuroscientific knowledge, the more productive it is
likely to be…and t hose using purely interpretivist methodologies are leaving themselves wide open to personal bias.’
I am a great fan of social psychology and neuroscience. I think we should all study more of them. What I find troubling is the suggestion that they are necessary components in a productive scientific methodology, and that any research programme/tradition lacking them will not be productive or useful.
And I don’t think that personal bias is a bad thing, so long as we’re open about it. In fact, I’m not sure facts exist independent of such biases, or are somehow improved or that ‘more true’ statements come out of minimising such biases, though I can certainly see great value to methodologies that operate under the assumption that they can.
Simon, we’re definitely talking past each other. I’m thinking of the conception of structure articulated by Alex Wendt, not Wittgenstein (whom I haven’t read). As for bias, the key problem is that interpretivists tend to be insensitive to the fact that most people don’t read tests the way they do, so their interpretations don’t shed much light on what normal people think.
Am working on a reply now; hope to have it posted before the 4th. Kudos to Stuart for setting the table with a lot to chew on.
Thanks for an interesting set of posts!
“People have personalities, ideologies, ideosyncracies, biases, and as a
matter of fact, they respond differently to structural pressures.”
How deep do we have to go to be able to explain things “scientifically” enough? How do we decide which elements of personality, biases, traumas, sexual perversions etc. are relevant enough to be taken into account when analysing construction, identities, decisions, structures…? Should psychology take structures into account as well, if we start from an assumption that individuals are “the product of a complicated interplay of nature and nurture” or does this requirement of “all levels” hold only for political science?
Are we able to trust on psychology and similar fields? In Norway, the world’s best mind experts can’t even decide if Breivik is ‘compos mentis’ or not. By the way, he says that what he did was a political act – an answer to the structure.
Undoubtedly, we cannot explain everything if we don’t take everything into account. For me, this seems as a kind of lethargic and
indifferent attitude towards the field’s (and science’s) own limits. Aren’t those
theories the best ones which identify most logically and most coherently most of
the key variables for explaining the phenomena under consideration? Otherwise will happen what happened in the USSR in the 1930s, when Stalin wanted to have a comprehensive map of the motherland… in scale 1:1…
“How do we decide which elements of personality, biases, traumas, sexual perversions etc. are relevant enough …? Should psychology take structures into account as well…?”
Excellent set of questions, Riika. The answer to the first question–what do we need to take into account?–is to find out through study. Here’s an example. There’s an interesting literature showing that personality factors like “right-wing authoritarianism” (RWA) and “social dominance orientation” (SDO) are strong determinants of generalized prejudice (Altemeyer 1998). On the other hand, a well-regarded Dutch study (Hagedoorn and Verkuyten 1998) found that SOCIAL effects determined when personality mattered and when it didn’t. Basically, if people were primed to think about their ingroup stereotype (we Dutch are tolerant), that effect swamped the effects of personality in determining prejudice.
So, to answer your first question, the whole RWA and SDO thing sounds plausible to me, but since there is solid doubt about whether the effects matter in a social context, I leave those factors out of my model of ethnic politics.
To answer your second question, psychology should and does take structures into account. In this study, the experimenter played the role of the structure by priming people to think about different contexts. In other studies, structural factors such as media effects are studied directly. And again, the finding of this study is that the stereotype–aka discursive formulation–trumped the personality factor in influencing whether people expressed prejudiced attitudes.
Psychology doesn’t replace constructivism, it helps explain how and when it works.
I ran out of time, so I won’t be able to finish and post my reply to Stuart until I get back from my daughter’s national dance competition late next week. In the meantime, though, I’ve set up an essay of my own (on “relativism”) to auto-post in three parts while I’m gone…