Allen Neuringer: Behavioral Variation
Allen Neuringer’s Insightful Research on Behavioral Variation
The Aubrey Daniels Institute is pleased to recognize the outstanding contributions of Allen Neuringer, Professor Emeritus at Reed College in Portland, Oregon, to the advancement of behavior analysis. In the opinion of the Institute, his research on the problem of variation in behavior is among the most creative and insightful in our discipline.
Variation in behavior is a double-edged sword for behavior analysts. On the one hand, shaping new behavior depends on there being at least some variation in a child’s present behavior. If behavior is always exactly the same (which of course it never is) there is no basis for shaping a response that comes closer and closer to the target behavior than one that is further away from that target. On the other hand, too much or too persistent variability may make it difficult for a child to learn a new response. High baseline variability also makes it difficult to assess whether an intervention has made a difference.
Allen Neuringer’s research shows how variability can be controlled -- increased or decreased -- by reinforcers that are directly contingent upon the variability.
In an early experiment, he asked if he could teach himself to type random patterns of the digits 0-9, that is, to behave in a random manner. To the surprise of many psychologists who had claimed that random behavior was not possible, with the help of a computer that gave him feedback about whether his sequences were random or not (based on comparing them to a computer program that measures randomness), he indeed taught himself to approximate a random model, that is respond with maximum variation. Feedback was the key.
Many people have criticized the use of reinforcement because it leads to stereotyped behavior – the antithesis of variable behavior - and thereby is said to stifle creativity. Neuringer has conducted a long series of experiments with pigeons showing that if stereotyped behavior was reinforced, stereotyped behavior was what you get. But if you reinforce variation in responding (defined in the experiment by the way in which the pigeon distributed a sequence of 8 responses between two response keys (link to response key commentary), you get variable – “creative” if you will – behavior. His work, then, showed that rigidity or flexibility of behavior – stereotyped or creative behavior - was not an inevitable consequence of context or general conditions. Rather, stereotypy or creativity depends on what is required for reinforcement, on the contingencies. The take home message? If you want people to be creative, set up contingencies that require creative behavior for reinforcement. If you don’t want that, then set up an environment that allows the same old response to be reinforced time after time. His own creative research has too long been overlooked by the acrimonious critics who cite the deleterious effects of reinforcement on creativity.
What about the “bad” side of variability? When it interferes with determining whether our treatment has made a difference because there is too much noise in the baseline behavior level to see the effect of the treatment. Neuringer’s research implies that “bad” variation also is a function of circumstance, not an inherent part of the behavior itself. Something in the environment is causing the “nuisance” variability. Change the environment and the variation can be reduced to a level that allows assessment of an intervention’s effects.
Professor Neuringer’s systematic, wide-ranging, and creative research program has stimulated many people to think about behavioral variation in new ways, and to see the implications of his insights for theory and practice, as well as for that broader bedrock principle of every science, determinism. For all of these reasons, the Institute recognizes his insights and contributions as outstanding.
