The Ideological Symmetry of Motivated Reasoning

On the heels of the John Bullock article & his amplification of it below,  the ideological neutrality of motivated reasoning came up again in an informative exchange with Howie Lavine during my recent presentation at the University of Minnesota. So I’ve found myself continuing to ponder the matter.

In our work, we test the hypothesis that cultural cognition — a species of motivated reasoning that reflects the impact of group values on perceptions of fact — is responsible for conflicts over scientific evidence on issues like climate change, the HPV vaccine, & gun control (and for conflicts over non-scientific evidence on many legal issues, too). The hypotheses assume that those on both sides of such debates are being affected by cultural cognition, and our data seem to reflect that.

But at least some social scientists have been advancing the claim that motivated reasoning in politics is more characteristic of (or maybe even unique to) conservative ideology. Essentially, these researchers are reviving the “authoritarian personality” position associated with Adorno. The most prominent of these neo-Adorno-ists is John Jost (see here, here & here, e.g.).

I tend to doubt that motivated reasoning is ideologically lopsided. What’s more, I tend to believe that even if the effects are not perfectly uniform across the ideological continuum (or cultural continua; we use two dimensions of value in our work as opposed to the single “liberal-conservative” one that Jost and others use), the impact of motivated reasoning is more than large enough at both ends to be a concern for all.

But I acknowledge the issue of “motivated reasoning asymmetry” is an open one, and agree it is worth investigating.

Obviously, the investigation should consist in empirical testing. But there must also be attention to theory, which is necessary to tell us what we sort of evidence is relevant, and hence how tests should be constructed and interpreted.

To that end, I offer some thoughts on a couple of the theories that might result in contrary predictions on the asymmetry thesis & what they suggest about empirical testing of that claim.

As I read Jost and others, the asymmetry position grounds motivated reasoning in a general propensity (a personality trait, essentially) toward dogmatism that tends toward a conservative (or “authoritarian”) political orientation. On this account, we shouldn’t expect to see motivated reasoning among liberals, whose ideology is itself a reflection of their propensity toward open-mindedness.

In contrast, the symmetry position (as reflected in cultural cognition and related theories) sees ideologically motivated reasoning as simply one species of identity-protective cognition. As developed by Sherman & Cohen, identity-protective cognition refers to the dismissive reaction that individuals form toward information that threatens the status of (or their connection to) a group that is important to their identity.  “Democrat” and “Republican” (along with hierarchy and egalitarianism, communitarianism and individualism, in cultural cognition) are both group affinities of that sort, and so both create vulnerability to motivated cognition.

Simple correlations of the extent of motivated reasoning with partisan identity or ideology (or cultural worldviews) furnish the most obvious way to test the asymmetry thesis but are unlikely to be conclusive because of their modest magnitudes and their variability across studies (such asymmetries in lab studies will also raise toughter-than-usual external validity questions). One nice thing about specifying the  theories in this way, we can expand the search for evidence that gives us more or less reason to accept or reject the asymmetry thesis.

E.g., if personal self-affirmation works to reduce resistance to ideologically noncongruent information among both liberals & conservatives, Republicans & Democrats–that, in my mind, counts as reason to be skeptical of asymmetry. The effect of self-affirmation is evidence that the source of the motivated reasoning at work is identity-protective cognition; there’s no reason to expect self-affirmation to have any effect in mitigating motivated reasoning that arises from a generalized disposition toward dogmatism.  And, btw, we already know self-affirmation reduces the resistance of liberal Democrats as well as conservative Republicans to ideologically noncongruent information. See here & here, for example.

Also: If we see ideologically motivated reasoning operating through sensory perception, that’s a reason to be skeptical of asymmetry too. The neo-Adorno-ist dogmatic personality theory addresses responses to arguments and evidence that bears argumentatively on political positions; it is about closed-mindedness not sensory blindness. Identity-protective cognition doesn’t make any claim that self-defensiveness will be limited only to assessments of arguments, and so can fit motivated reasoning effects in sight & other senses.  Research using cultural cognition has shown that motivated reasoning can generate polarization of individuals of all values when they observe video of politically charged events  (e.g., abortion-clinic vs. miltitary-recruitment center protests or high-speed police car chases).

Lastly, if we can parsimoniously assimilate motivated reasoning in politics to a larger theory of motivated reasoning, then we should prefer that account to one that posits a patchwork of local motivated reasoning dynamics of which ideologically motivated reasoning is one. Identity-protective cognition offers us that sort of parsimony: individuals are known to react defensively against information that challenges diverse group identities — like being the fan of a particular sports team or a student of a particular university — and not only against information that challenges partisan or ideological identities.  The neo-Ardon-ist dogmatic personality theory doesn’t explain that (although it does seem to me that Yankees fans are very closed minded & authoritarian).  Thus, more evidence, I think, for the symmetry position.

More but not conclusive evidence. For me, the question is, as I said, very much an open one.  Also, I don’t mean to say that identity-protective cognition & the dogmatic-personality theories are the only ones to consider here.

The only point I am trying to make is that we are likely to get further in answering the question if we think about it in conjunction with theories of motivated cognition that offer competing predictions about symmetry and other things than if we just gather up studies & ponder correlations.

Or to put it more concisely, and on the basis of a (profound) truism from the philosophy of science: No theory, no meaningful observations.

Leave a Comment