Explanatory Fictions and Private Events
One of the concepts of behavior analysis that I consistently see students struggle with mastering is an explanatory fiction. On the surface, it seems like something that should be a no-brainer. But until you have a strong grasp of what behaviorism is about, it can be very confusing. It is like things being partially hidden behind a curtain of irrelevant details. (Yeah, I had to explain the associated picture.)
We just completed an assignment in the class I teach where the students were to identify whether a scenario was a testable hypothesis or explanatory fiction. I kept seeing the same mistakes across multiple sections of the class. And they generally related to a mis-identification of what behavior analysis covers.
Several of the scenarios used private events to describe the cause of the behavior. “Marie bites her nails when she is nervous” was one such scenario. This was identified as a testable hypothesis way more times that it should have been. Being nervous is a private event. It is a hypothetical construct that we assign to others based on our own personal learning history. We can’t know that Marie is nervous. All we can do is examine her behavior. And, how do you ethically test nervousness?
Another example was “Sandy gets cranky when she’s hungry.” This one posed a couple of problems. First, the behavior, “cranky” is neither objective nor measurable and could easily mean different things to different observers. That makes it almost impossible to test. More than that though, hungry is another private event. We can’t know that she is hungry. The best we can do is say she has not eaten for X number of hours and, based on our own experience having not eaten for that long, assume that she is feeling the same. So, explanatory fiction.
And that’s how I explained it to those who missed those things. Private events, while still a part of behavior analysis, can’t be part of an explanation of behavior since we want observable and measurable dimensions of behavior. I also suggest considering how you can test the hypothesis. Can you take each part of the scenario and devise a way to empirically test it using what you know about behavior?
I think that many people assume that everyone knows what nervousness or hunger is and could identify it consistently. But how do you identify nervousness? By the behavior you see the person exhibiting. Nail biting. Fidgeting. Those behaviors can also be associated with other states of being. Fidgeting could come from an uncomfortable seating position or itchy clothing. Neither of which reflect being nervous. So, the consistency is lacking in using nervous as a cause of an observable behavior. Same for nail biting. There could be an uncomfortable hangnail that needs attending.
Remember that private events are a part of behavior analysis, but we can only access them through the observable behavior of others. Using our own construct of the private event to attribute a cause to the behavior of others is not helpful and could actually lead us down the wrong path. Things are hidden from our view. A fascinating topic for another post.