What is rule-governed behavior?
Following the tradition of B. F. Skinner, the famous Harvard University psychologist who pioneered the study of operant conditioning, behavior analysts initially examined the behavior of rats and pigeons because nonhuman subjects could be studied under well-controlled conditions in the experimental laboratory. This made it possible for operant psychologists to discover a number of important behavioral principles and to demonstrate that much of the behavior of their experimental subjects was shaped and maintained by contingencies of reinforcement. “Contingencies” can be thought of as cause-effect relations between a context (in operant terms, a “discriminative stimulus”), an action (“response”), and the consequence (“reinforcement”) it produces. For example, if pressing a bar is followed by food only when a light is on and never when it is off, a rat’s behavior is gradually shaped by these contingencies until the rat presses the lever only when the light is on.
When operant researchers began to bring human subjects into the laboratory, however, the analysis went beyond behavior directly shaped by contingencies to include behavior under the control of instructions or rules. According to Skinner, a rule is a “contingency-specifying stimulus.” It functions as a discriminative stimulus (SD), but it differs from other SDs in that it is a description of a behavior-outcome relation. Other SDs are stimuli in the environment that acquire control over behavior only through specific training; rules, in contrast, have an immediate effect on behavior because they make use of an already existing language repertoire. For example, through a history of careful shaping, a seeing-eye dog can be trained to stop at red lights and cross the street only when the light is green. A verbal child, however, can be taught the same discrimination simply by being told, “Go when the light is green; don’t go when the light is red.”
Proverbs, maxims, advice, instructions, commands, and so forth all function as rules when they control behavior. In complete form, rules specify an antecedent condition, an action, and its consequences, and often take the form of if-then statements, as in “If you want to get to the other side safely, then cross the street only when the light is green.” Most rules, however, are only partial statements of contingencies, specifying exclusively the antecedent (such as a male figure or the word “men” on a door), the behavior (a sign reading do not enter), or the consequence (“Lose twenty-five pounds in one month!”), and it is left to the individual to fill in the blanks.
Despite an abundance of rules in the human environment, many people are not reliable rule-followers. Control by rules is often deficient because rules only determine the topography, or form, of behavior, but they do not impart the motivation to act. Stated differently, rules tell people what to do, but whether people actually do it depends on other circumstances.
For a rule to be followed, it must be part of an effective contingency: Either the outcome specified in the rule must function as a reinforcer, or the rule giver must be able to mediate aversive consequences (punishment) for noncompliance. Psychologists Steven C. Hayes and Robert D. Zettle have drawn an important distinction between contingency-shaped and rule-governed behavior. They assert that contingency-shaped behavior is controlled by one set of contingencies, usually consisting of a situation, an action, and a consequence (such as being offered a beer, drinking, and feeling relaxed).
In contrast, rule-governed behavior involves two sets of contingencies. One of them is the behavior-outcome relation specified in the rule itself (“If you want to avoid addiction, just say no”). The second involves social consequences for rule following, such as praise or criticism from significant others or social pressure to comply with peer norms. As the following examples will show, at times both sets of contingencies support rule following, but sometimes they compete with each other. In one example, a man is lost and his wife insists that he ask for directions. He is told to “turn left at the light and then follow the signs to the interstate.” The man is likely to follow these directions, because both sets of contingencies surrounding rule following are congruent: The natural consequences of finding the highway are indeed reinforcing to him, and the social consequences are reinforcing because following the directions will satisfy his wife and spare him criticism. In another example, however, a child is given a box of candy. Her mother says, “You may have only one piece of candy before dinner, or else you will spoil your appetite.” The contingency specified in the rule may be ineffective, because eating only one piece of candy if there is more may never have been reinforcing to the child. Hence, if the child obeys, it is not for the contingency specified in the rule but for the parental consequences that would result from noncompliance.
Behavior under the control of a description of contingencies does not involve a new process but is consistent with an operant framework postulating that the probability of behavior is controlled by its outcome. Rule governance results from an extensive history of reinforcement in which rule-following has directly led to contact with the contingencies specified in the rule, to social consequences associated with compliance and noncompliance, or both.
Teaching people to follow rules is important for a number of reasons, which B. F. Skinner outlined in his book About Behaviorism (1974). Most important, many behaviors can be acquired much more quickly through rules than through shaping by the contingencies described in the rules. For example, it is easier to teach a boy the basics of a card game by explaining the rules to him than by playing with him until he gradually (if at all) figures out the rules for himself. Furthermore, there are cases when the contingencies are so complex or vague that most people would never understand them without the help of rules. Learning to type with ten fingers illustrates such a case. Without appropriate instruction, the immediate success accruing from a hunt-and-peck method will reinforce typing with two fingers, and the person will never learn to use ten fingers, even though in the long run this would have been much more efficient.
According to Roger L. Poppen in a 1989 essay, initially people learn rules from a multitude of external sources such as parents, peers, teachers, television, and books, and eventually they learn to extract rules from interacting with and observing environmental contingencies. Parents encourage the rehearsal and internalization of rules so that these self-instructions then help children guide their own behavior in similar circumstances.
The effects of rules on behavior have been extensively studied within a behavior-analytic methodology. A summary of this research can be found in a chapter by Margaret Vaughan in the book Rule-Governed Behavior: Cognitions, Contingencies, and Instructional Control, edited by Steven C. Hayes (1989). Most of these human operant studies use a method in which subjects press a button that, according to some schedule of reinforcement (an arrangement that specifies which responses within an operant class will be reinforced), occasionally produces points exchangeable for money. Depending on the preparation, button pressing may be controlled by the contingency between pressing and point delivery; in this case, the behavior would be contingency-shaped. Button pressing may also be controlled by experimenter instructions, in which case the behavior would be rule-governed.
A number of studies showed that experimenter-provided instructions quickly bring the behavior under stimulus control (behavior occasioned by a stimulus because the stimulus signals some consequence of responding) but also create insensitivity to the scheduled contingencies. For example, telling subjects that “the best way to earn points is to press the button fast” (a fixed-ratio contingency) immediately allows them to respond correctly and earn points. When the contingencies are then surreptitiously changed, however, subjects continue to follow the instructions for long periods of time although they have become obsolete and no longer produce rewards. In contrast, when subjects receive no instructions and their responses are shaped, sensitivity to changing contingencies develops; that is, when the schedule of reinforcement changes, subjects adjust their behavior to the new schedule and continue to earn points. This observation has led operant researchers to conclude that insensitivity to contingencies may be an inherent property of instructional control.
The insensitivity effect of rules has intriguing implications: Instructing people how to solve problems is immediately effective, but it may be counterproductive in the long run, because individuals may come to act in accordance with outdated rules. Their behavior may come to be guided by what cognitive psychologists call irrational beliefs or unrealistic expectations, which from an operant perspective would be considered inaccurate statements about contingencies resulting from broad overgeneralizations of old rules. The following example illustrates how an irrational belief may come to control behavior. A mother might tell her child, “Stop making noise! I don’t love you when you are bad.” This rule may quiet the child immediately, and because it is effective, the parent may use it in other situations. Over the course of her development, the child learns many instances of what her parent considers “bad” (perhaps disobeying instructions, perhaps asserting herself, showing anger, and so on). Gradually she internalizes a generalized rule, “I am only lovable when others approve of me,” and evolves into an adult who tries to please everybody and feels unworthy at any sign of disapproval, however ineffective this behavior may be.
Humans live in a world in which rules abound, in the form of instructions, advice, warnings, manuals, cookbooks, self-help books, laws, and social norms. They are intended to provide guidelines for effective behavior. Even when no external rules are available, most people can formulate their own plans of action. The greatest advantage of rules is that they can be extremely helpful and can establish effective behavior quickly. Their greatest disadvantages are that rules do not produce behavior unless other contingencies support rule following and (as Skinner has pointed out) that they may be troublesome rather than helpful when the contingencies change but the rules do not.
Originally, behavioral researchers attempted to replicate findings from experimental work with rats and pigeons to demonstrate the generality of the principles of behavior discovered in the animal laboratory. It soon became apparent that people often showed response patterns not comparable to those of animal subjects on the same schedules of reinforcement. For example, a cumulative record of responding on a fixed-interval schedule for animals typically shows “scallops” (a pause after reinforcement, followed by a gradually accelerating response rate until delivery of the next reinforcer). In contrast, human subjects typically time the interval by counting; toward the end, they respond as few times as necessary to obtain the reinforcer.
Behavior analysts suspected that the differences between human and animal responding mainly stemmed from people’s prior conditioning history and from instructions, both experimenter-provided and self-generated, with which they approached the experimental tasks. These assumptions began to focus the attention of operant researchers on the role of instructions. By the mid-1970s, instruction following became synonymous with rule-governed behavior and began to evolve into a field of study in its own right.
One importance of rule governance lies in the possibility of a rapprochement between behaviorist and cognitivist positions. Behaviorists have often been accused of disregarding or failing to acknowledge the importance of higher mental processes. Although such accusations are polemic and extremely misleading, it is true that it was not until the mid-1970s that operant psychologists began a systematic empirical analysis of cognitive-verbal processes. The study of rule-governed behavior marked the beginning of the experimental analysis of phenomena that until then pertained to the domain of cognitive psychology.
The analysis of rule-governed behavior is important for another reason. It provides some insights into causal mechanisms that may underlie current cognitive therapies. For example, in 1987 Poppen presented an excellent theoretical analysis of a self-efficacy approach and of rational emotive therapy, while in 1982, Zettle and Hayes presented a similar analysis of cognitive restructuring and cognitive therapy for depression. The common denominator of these diverse cognitive approaches is their assertion that people’s reactions to their environment are mediated by covert verbal statements, which, when dysfunctional, are given labels such as irrational beliefs, low self-efficacy, and negative expectancies. From an operant perspective, such formal categorizations are considered not very useful because formally distinct verbal statements may all have the same function, while statements identical in form may have different functions (one person might say “I can’t do it; I’m too dumb” to avoid an unpleasant task, while another person may say the same thing to request assistance).
Within a framework of rule-governed behavior, all these dysfunctional cognitions are considered partial statements of contingencies, and the behavior they produce is rule-governed. Hence, findings from basic experimental research on rule-governed behavior could conceivably be brought to bear on clinical phenomena, which eventually might lead to a better understanding of psychological dysfunctions and to the development of more effective therapies.
Catania, A. Charles. Learning. 4th ed. Cornwall-on-Hudson: Sloan, 2006. Print.
Freeman, Arthur, ed. Encyclopedia of Cognitive Behavior Therapy. New York: Springer, 2005. Print.
O'Hora, Denis, Dermot Barnes-Holmes, and Ian Stewart. "Antecedent and Consequential Control of Derived Instruction-Following." Journal of the Experimental Analysis of Behavior 102.1 (2014): 66–85. Print.
Pierce, W. David, and Carl D. Cheney. Behavior Analysis and Learning. 5th ed. New York: Psychology, 2013. Print.
Poppen, Roger L. “Some Clinical Implications of Rule-Governed Behavior.” Rule-Governed Behavior: Cognitions, Contingencies, and Instructional Control. Ed. Steven C. Hayes. Reno: Context, 2004. Print.
Skinner, B. F. About Behaviorism. 1974. Reprint. London: Penguin, 1993. Print.
Skinner, B. F. “An Operant Analysis of Problem Solving.” Problem-Solving: Research, Method, and Theory. Ed. Benjamin Kleinmuntz. New York: Wiley, 1966. Print.
Vaughan, Margaret. “Rule-Governed Behavior in Behavior Analysis: A Theoretical and Experimental History.” Rule-Governed Behavior: Cognitions, Contingencies, and Instructional Control. Ed. Steven C. Hayes. New York: Plenum, 1989. Print.
Watts, Amanda C., et al. "The Effect of Rules on Differential Reinforcement of Other Behavior." Journal of Applied Behavior Analysis 46.3 (2013): 680–84. Print.
Zettle, Robert D., and Steven C. Hayes. “Rule-Governed Behavior: A Potential Theoretical Framework for Cognitive-Behavioral Therapy.” Advances in Cognitive-Behavioral Research and Therapy. Ed. Philip C. Kendall. Vol. 1. New York: Academic, 1982. Print.