The Cognitive Science of Moral Failure: Prejudice and Blink

Are you a racist?

Okay, that's probably too harsh a question. I'm just trying to get your attention. Let's ask something more subtle:

Are you prejudiced?

That is, do you have negative stereotypes about ethnic groups, gays, Muslims, genders or other groups?

Here's the deal. You can't really answer these questions.

The reason has to do with the cognitive systems we've been talking about in the last few posts. Prejudice and stereotypes are driven by System 1, the system that is fast, unconscious and automatic. This is also the system that is largely walled off from introspection. So when I ask you if you are prejudiced what you end up doing is searching System 2, the conscious repository of our values, morals and ideals. And when you consult that repository of course it tends to look like you are not prejudiced. That is, when you consult System 2 this is what you find:

I believe, given my values, that it is wrong to be prejudiced.

I have as a goal for myself not to be prejudiced.

I consciously try not to be prejudiced.

Note, however, none of this tells us if we actually are prejudiced. Values, goals and desires don't necessarily translate into action. How many of us value, desire and set goals for more healthy living? Think of those New Year resolutions. A moment of reflection reveals that not wanting to be prejudiced has little bearing upon actually being prejudiced.

The reason for this is simple. Introspection only penetrates System 2. But prejudice is produced by System 1. You're self-analyzing the wrong system.

But how could you ever analyze an unconscious system? You can't rely upon introspection. You need something that assesses our quick, automatic appraisals. The "blink" of Malcolm Gladwell's book. How do we assess blinks? Enter the Implicit Association Test (IAT).

The IAT was developed by Harvard psychologists to test for the strength of implicit (i.e., System 1) associations, good and bad, for different targeted stimuli. The first test assessed Good/Bad associations for race. Since then a whole host of tests are available for associations regarding obesity, old age, and religion. You can take one of these many tests by going to Project Implicit at the Harvard host site. Click on the link and take one of the tests.

If you take an IAT you see what it does. It assesses reaction time as you sort a target category (e.g., White faces vs. Black faces) along with an attribute (e.g., Good vs. Bad). What the test reveals in the racial version of the IAT is that, generally for white people, when "Good" attributes are paired with "White" faces our ability to classify is improved (assessed as mean reaction time over the repeated trials) relative to the trials when "Good" attributes are paired with "Black" faces. In other words, we find it easier to associate White/Good and Black/Bad relative to when we have White/Bad and Black/Good pairings. We implicitly associate whiteness with goodness and blackness with badness. You don't know you do this (again, introspection is no help) but the test reveals that, in fact, you do.  

(Incidentally, you probably noted that IAT works because of the Stroop Effect. That is, "good" when paired with white faces is easy while "good" paired with black faces creates the interference of the Stroop Effect.)

The IAT has important implications for spiritual and moral formation. Specifically, church people tend to lean too heavily upon introspection. Church people very often want to be good people which leads them to assume that they are good people. In short, too many Christians think they are good when, in fact, they are not. We call this hypocrisy.

And it's not a willful hypocrisy. Just a mindless one. A failure to understand that values, goals, and desires are not necessary and sufficient conditions for being a good person.  Once again, the fissure between the two cognitive systems has risen up and bitten us.