The seduction of temporally proximate causes

We grow up seeing causation in the world around us in its simplest terms – the stack of blocks fell over because our little brother hit them, the pencil rolled off the desk because the cat pushed it, the car door closed because we pulled it. We don’t have to understand the formalities of physics to conclude that events can apparently be explained in terms of other immediately preceding – temporally proximate – events. This billiard ball model of causation is easy to absorb because we’re surrounded by countless examples every day.

It’s understandable that we extend this causal frame of reference to behavior. Behavior is just another event in our environment, after all, and things that happen immediately preceding our actions offer seeming obvious explanations of much of what we do. A little boy playing with a toy truck gets upset because another child took the truck away from him. A student responds to a professor’s question. Approaching an intersection, we start braking because the light turned red. Although behavior analysts know there’s a lot more going on in these examples than what happened before the behavior, the causal role of immediately prior events usually seems clear to everyone else.

When in doubt

If a behavioral circumstance seems to lack worthy prior events, or if there is any doubt about their causal contribution, our verbal community explicitly encourages turning this causal model inwards, where temporally proximate prior events are conveniently and endlessly available for causal assignments. All that is required is invention of a mental domain and willingness to attribute a causal role to its contents. The operating assumption is that mental events are always going on and therefore precede and accompany everything we do. Our vernacular dialect confirms the belief that mental processes are at the root of human behavior.

Thanks to this pervasive mental arena of potential behavioral causes, we’re never short of possible explanations of behavior. Many are so well established in common parlance most people don’t even recognize the causal assumptions. “I decided” to put up the holiday decorations today is seen as sufficient to explain this activity. He went to the mailbox because “he assumed” that the mail had been delivered. “I was worried” about it freezing tonight, so I took the plants inside. Our conviction that behavior is the result of temporally proximate prior events is preserved by reference to mental events apparently preceding our actions that are assigned the needed causal role. It’s that easy.

Why a behavior occurs when it does

This implicit reasoning is especially likely to show up when trying to explain why a particular behavior occurs when it does. When someone does something others don’t expect, it’s understandable that they look for an explanation of why the behavior occurred at that moment. The lack of an obvious justification may even encourage them to say the behavior was spontaneous, which implies that it might not have even had a cause. Of course, this assumption violates the deterministic posture so dear to behavior analysis and other physical sciences.

Failure to anticipate a behavior may also lead people to say that it occurred suddenly or without warning. This characterization pretty much admits that we don’t know much about that behavior or, perhaps, behavior in general. After all, the occurrence of a particular behavior is not necessarily forecast in any obvious way by events that preceded it. If our roommate “suddenly” gets up from studying and gets a beer from the refrigerator, we may not have anticipated the trip to the kitchen, but that should not suggest there is anything special about the nature of the behavior or its causes. Any particular behavior is not occurring before it then occurs, and there is no reason there should be noticeable environmental events or precursor behaviors alerting us to its impending appearance, even if we were paying attention.

In wondering why a behavior occurs when it does, admitting that we’re not aware of notable physical events that preceded the behavior or that we don’t find them up to the task of explaining why it occurred at that moment encourages people to turn to mental alternatives. This fictional domain provides plenty of prior, temporally proximate causal options, albeit without the luxury of falsifiability. Mentalism instead offers convenience and comfortable familiarity, both hard to resist given the reassurance of our verbal community for this kind of rationale.

How long has it been?

Another reason for turning to mental causes is that people accept causal attributions to prior environmental events only if the gap between an event and the following behavior is relatively brief. As Box 3.2 suggests, when we retrieve our food from the microwave shortly after the bell rings, it seems obvious that the bell caused our action. However, people tend to be comfortable with this conclusion only when the gap between the bell and the behavior is brief. If it’s only a few seconds, there’s no problem, but when the bell rang a few minutes ago, the tendency to insert a mental process into the causal chain grows. And if the gap extends to an hour or more, the role of the bell is usually superseded by reference to mental processes such as remembering. When gaps between prior environmental events and the behavior to be explained are measured in days, weeks, months, or even years, most people are unlikely to give much of a causal role to physical events, looking instead for alternatives in a mental domain that seem more immediate and credible.

Behavior analysts, however, know that there is very little about behavior that does not rely on historical influences, which may be recent or much further back in time. In general, the contribution of behavioral histories is cumulative, though recent events tend to play a more significant causal role than earlier ones. Nevertheless, our understanding about the nature of historical causation does not suggest that there is any point in the temporal gap between a behavior and possibly controlling prior events at which the nature of causal explanations needs to change. The mechanisms by which prior environmental events may influence behavior do not depend on the size of the time units involved.

In sum

Our experience with the physics of daily life does not provide a good guide to understanding the causes of behavior. The science of behavior analysis has revealed that the causes of behavior are multifaceted and complex, even in what might seem simple examples. Although physical events occurring immediately prior to a behavior participate in a causal analysis, they should not usually be assigned a lead role. The contrary conviction that behavioral causes must be found in temporally proximate prior events encourages a search for prospects in a fictitious mental domain, guided by colloquial dialect. The resulting references to mental processes may seem satisfying for everyday purposes, but they interfere with understanding the nature of operant behavior and reaping the benefits of the science of behavior analysis.