Reward or Reinforce?

Are we concerned with "rewarding" or "reinforcing" behavior? Does it matter? Although perhaps not the most burning question in behavior analysis, it does bear on some important issues to those who use and study reinforcement…or is it reward?  

The Random House Dictionary of the English Language's definition of reward is “n. 1. Something given in return for or recompense for service, merit, hardship, etc. … -v.t. 3. To recompense or requite (a person or an animal) for service, merit, achievement, etc. …” Catania (1991, p. G32) defined reinforcement as “the response-produced presentation of positive reinforcers or termination of negative reinforcers (or the increase or maintenance of responding resulting from this operation)….”. He then noted that “[t]he operation is to reinforce responses, not organisms (organisms are sometimes said to be rewarded, but this term often implies other effects of stimuli than reinforcing effects).”

Staddon (1993, p. 442) observed that “we cannot be certain that all the phenomena that fit the conventional definition of reinforcement (response-rate increase caused by contingent reinforcer presentation) in fact belong together, because we still have no consensus on the underlying process (or processes). Hence to insist on the word reinforcement as opposed to reward or simply food delivery is to imply a greater precision than the state of knowledge warrants. …most [behavior analysts] are alert to the improper connotations of the vernacular word reward so that few will read into it anything more than the delivery of food to a hungry animal.” Practicing behavior analysts have come down on either side of the issue, with a few prominent ones practicing what Staddon preached.

Many have stressed the practical need to moderate, or even eliminate, technical behavioral language when communicating with non-behavior analysts (e.g., Bailey, 1991). On the one hand, it of course makes sense to use language that people without backgrounds in behavior analysis understand.  Would a theoretical physicist present her work to a lay audience by using differential equations? We all use the words we have to describe our world. We can’t do otherwise. But, we all learn. It's not an uncommon experience among behavior analysts to find parents and other clients adopting behavior-analytic descriptions as they learn both the behavior-analytic concepts and the accompanying language. With experience, “rewards” often become “reinforcers” for many partaking of behavior-analytic services. The use of behavioral language can reveal a growing understanding of the concepts thus described, but pushing it on clients never is helpful and may be detrimental.

The advantages offered by approximations to precise language among behavior analysts (an important qualification – see the previous paragraph) outweigh those of general-use terms: (1) Contrary to Staddon’s claim, even though behavior analysts may understand the difference between the two, I too often hear colleagues slip into describing the “feel good” dimension of a reinforcer, without mentioning its behavioral effect. As I have noted previously, this is conceptually muddled. (2) There indeed may be ambiguities in the concept of reinforcement (and other concepts, too), but if we wait until they are all "straightened out" to start using terms like  “reinforcement” or “reinforcer’ instead of “reward,” we will never get to any kind of a technical vocabulary.  (3) Didactically, using terms can teach people concepts.  This doesn’t mean that a term, or concept, shouldn’t or can’t be questioned, but terms give people a common framework from which to question both the observations leading to the concept and its label (e.g., the observation from the 1970s, still not explained to everyone’s satisfaction, that electric shocks can function as positive reinforcers for the very response that produce those shocks). (4) By limiting the terms in common use for describing behavioral phenomena, and defining those terms with as much precision as the data allow, we enhance communication among ourselves. Selecting or using terms idiosyncratically helps neither behavior analysts nor the people they work to help.

Common sense and conceptual clarity should guide the use of any term, with any audience.  When in doubt, define and explain. When you are reasonably sure of your audience's acumen in the science of behavior, I say go for "reinforce."  


Bailey, J. S. (1991). Marketing behavior analysis requires different talk. Journal of Applied Behavior Analysis, 24, 445-448.

Catania, A.C. (1991). Glossary. In Iversen, I. & Lattal, K. A. (Eds.), Research methods in the behavioral and neural sciences:  Experimental analysis of behavior. Part 2 (pp. G1-G44). Amsterdam:  Elsevier.

Staddon, J.E.R. (1993). The conventional wisdom of behavior analysis. Journal of the Experimental Analysis of Behavior, 60, 439-447. 

Posted by Andy Lattal, Ph.D.

Dr. Andy Lattal is the Centennial Professor of Psychology at West Virginia University (WVU). Lattal has authored over 150 research articles and chapters on conceptual, experimental, and applied topics in behavior analysis and edited seven books and journal special issues, including APA’s memorial tribute to B. F. Skinner.