Searching for Behavior in Behavior Science

Has the science of behavior forgotten behavior?

Posted May 17, 2008

A recent conference highlighted the dwindling presence of behavior in psychology, the so-called science of behavior. The theme of last week’s conference, held at Purdue University, was behavior. I thought this was too broad a theme for a small conference, though in fact the conference worked quite well. I had thought, isn’t behavior what most of psychology studies, in principle at least? How could two days of talks begin to cover the full topic of behavior?

Certainly nobody thought the topic was exhausted by the 16 talks.

Yet the conference was useful. The goal was not so much to provide exhaustive coverage of the topic of behavior, but rather to rekindle interest in something psychology has occasionally emphasized but increasingly has seemed to abandon.

Experimental psychology started in the 1800s. Wilhelm Wundt’s laboratory is generally credited with being a pioneer. It focused on introspection. Thus, analyzing conscious experience and reporting on inner states was at the center of psychology’s first efforts to become scientific. Direct observation of behavior was not seen as crucial.

That changed radically across the first half of the twentieth century. To be sure, Freudian approaches, with their corresponding emphasis on introspection (treated with considerable skepticism by the psychoanalysts who assumed that unconscious processes, immune to direct introspection, were powerful) remained strong. But the scientists in psychology, especially in American universities, rallied round the behaviorist flag. At its most strident, behaviorism dismissed all talk of mental states as unscientific and possibly irrelevant “black box” phenomena. The important thing was to observe behavior directly — in fact, scientists should observe and count the behavior of other people (and/or white rats!).

Behaviorism began to wear thin in the 1960s and 1970s. Theory could not advance without acknowledging that what people thought and felt made a difference. My own field of social psychology helped bring about the crisis, because adult human behavior depended so obviously on thoughts and feelings that denying them destroyed any chance of making sense of behavior. There were last-ditch efforts to salvage behaviorism by claiming that thoughts and feelings were behavior too. (Yes, we did feel more scientific when we said “emotion behavior” than if we had merely said “emotion,” though it is embarrassing to recall this!) Psychologists asked people to report on their inner states and insisted, usually with a straight face, that marking a scale to report on one’s inner states was behavior too.

Eventually the nonsense was recognized as such and just collapsed. The cognitive revolution of the late 1970s was the death knell of behaviorism in psychology.

Unfortunately, some rather precious babies seem to have gone out with the used and useless bathwater. In a recent survey conducted by myself, Kathleen Vohs, and David Funder, we found that many areas of psychology seem now to have abandoned behavior almost entirely. In social psychology laboratories today, the typical procedure is to seat participants in front of a computer, have them read things and report on their thoughts and feelings. The defensive joke of the 1970s — that making checkmarks on paper to report your inner states is a form of behavior — has become the dominant mode of doing psychology.

If psychology is the science of behavior, then most behavior is performed in a seated position and consists of giving ratings.

The reliance on these ratings is especially ironic given that skepticism of introspection has gained scientific punch. A classic paper in the 1970s by Richard Nisbett and Timothy D. Wilson showed that when people furnish introspective reports on inner processes, they are often demonstrably wrong (and don’t realize they are wrong). They say what seems to make sense and what they think they are supposed to say, but they do not, and often cannot, give a correct answer about what has happened inside their minds.

Many psychologists think that Nisbett and Wilson permanently discredited introspection. Yet many of them and their colleagues now use introspection as their primary, often even their only, research method.

Don’t get me wrong. I think there is much to be learned from asking people to report on their inner states, especially if we can maintain the skepticism based on the work of Nisbett and Wilson and their successors.

But I also think it is a tragedy for psychology. The behaviorists may have wildly overstated their case, but they had important points. Direct, objective observation of behavior should have a valued place in psychology. It makes contributions to advance the field that no other method can make. And the criticisms of introspection remain valid. We can try to correct for them, but there is no substitute for venturing — occasionally, at least — into direct observation of behavior.

We can ask romantic couples how they feel about each other and how they foresee the future over and over, but these data need to be supplemented, now and then, by measuring whether they actually break up or get married.  

We can ask people how they feel about various political issues and candidates. But these are not a fully satisfactory substitute for measuring whether and how they actually vote.

We can ask participants to consider several possible products and rate them on various dimensions, maybe even to rate which one of several they consider the best. But occasionally we must see which one the buy, or whether they buy one at all.

The strongest movements in psychology right now are all against behavior. Cognitive psychology has been dominant and influential since the 1970s, but it studies thought processes and rarely deigns to observe anything resembling get-up-out-of-your-chair behavior. Neuroscience and brain imaging have captured the imagination of many researchers (and most of the grant funding), but the favored research method requires participants to lie motionless in a scanner. Clinical psychology remains large and powerful, partly because of society’s need for help, but the center of its world is the therapy session in which therapist and client mainly talk about what the client is thinking, feeling, and possibly doing elsewhere, so direct involvement in actual behavior is almost impossible. Stereotyping and prejudice dominate social psychology labs, but the dominant methods have come to treat them in terms of slight differences in reaction times, as opposed to burning crosses and refusing jobs or housing.

Again, I do not wish to complain about what psychology is doing, and indeed much is being learned from these methods and approaches. But we need behavior too.

The sad fact is that direct observation of behavior, and indeed direct intervention in actual behavior, is relatively inconvenient. In the highly competitive world of psychology careers, people do what can yield the best results fastest. Asking people to report on imaginary responses and inner states is easier than observing actual behavior. These competitive pressures are crowding behavior out. Each individual decision is understandable, but the result could severely endanger and trivialize the future of our field. We need to find ways, incentives, opportunities, to promote a bit more behavior alongside all these other trends.