REINFORCEMENT

The process of strengthening or supporting behavior is reinforcement, a term that is embedded in learning THEORY. As such, the use of the concept implicitly harbors assumptions about the mechanisms involved in learning and ultimately the nature of the human animal. Its application in sport and exercise is diffuse. Most coaches use rewards to reinforce SKILL ACQUISITION and punishment to deter improper behavior; whether they know it or not, they are using reinforcement. EXERCISE BEHAVIOR is often reinforced by the reactions of others to changes in appearance; this has been called social reinforcement, and it is one of several types of reinforcement, the most basic types being positive and negative. Positive reinforcement describes stimuli that, when presented fol-

lowing a response, increase the probability that the same response will be repeated; the rewards accruing to successful athletics performances are examples. Related to this is negative reinforcement, including painful stimuli, like electric shocks or deafening noise, which, when withdrawn emit the desired response. Negative reinforcement is not the same as punishment. The former involves an active deterrent, such as removing the physical distress resulting from a shock, while the latter entails denial of, or withdrawal from access to rewards and decreases the probability of a RESPONSE. For example, giving an animal an electric shock every time it presses a lever will weaken not reinforce the response-this punishes pressing the lever. Conditioning 1: Classical. The term ‘‘reinforcement’’ has origins in

Ivan Pavlov’s early twentieth-century experiments with dogs that were inadvertently conditioned to salivate in response to a signal as much as they did to the presence or smell of food. At first, the dogs salivated only when their food arrived-what Pavlov (1849-1936) called an unconditioned response (UR) to an unconditioned STIMULUS (US)—but the continual accompaniment of food with the signal forged an association, so that, even in the absence of food, the dogs salivated when the signal sounded. This was a conditioned response: the food (US) functioned as a reinforcer of the response to the conditioned stimulus (CS). Take the food away, and, eventually, the UR becomes extinct. The food satisfies a need or a desire. In this sense, the reinforcer is a REWARD. Later experiments

demonstrated that similar simple rewards could be used to induce rats to push levers and pigeons to perform often elaborate behavior, like ‘‘playing’’ table tennis. The conditioning process described is instrumental, and it works on the basic idea that the reinforcer is contiguous

to the action. In other words, the reinforcement must take place quickly after the appropriate response. When every response is reinforced in sequence, there is a schedule of reinforcement. This is a little too straightforward, however. Some human behavior does appear to operate according to the

same principle. Fan adulation, cash, and sincere congratulations from a coach reinforce an athlete’s successful performance, as do compliments about body shape. These may seem like functional equivalents of food for rats: the rats continue to press the right bar, while the athlete or exerciser attempts to reproduce the same behavior that brought him or her rewards-though, of course, they might not. Building cash incentives into contracts, raising prize monies, or awarding medals and extravagant gifts to those who reach goals work as reinforcers. Humiliation, exclusion from honors, and the EMOTION associated with failure may negatively reinforce, or punish, the behavior that brought defeat. Obviously, reinforcement does not work in this regular and predictable fashion. Conditioning 2: Operant. Athletes do not win every COMPETITION;

exercisers do not always hit their targets and draw applause; gamblers do not win every bet. Yet, athletes continue to play, exercisers continue to work out, and gamblers continue to gamble, at least until the losses are so consistent or punishing that they can no longer be sustained. While schedules of reinforcement are effective in developing, modifying, and maintaining many types of behavior, complex human behavior is more difficult. B. F. Skinner (1904-90) experimented with variable interval, compound, and other sorts of variations in schedules of reinforcement. His operant conditioning is one of the most comprehensive explanations available. An operant is any instance of behavior that effects a change in the environment, whether it is a rat pushing a lever or a football player scoring; when the operant is brought under the control of a specific stimulus or combination of stimuli, operant conditioning has happened. This is different from simple respondent behavior, which is a direct response to a stimulus, such as Pavlov’s dogs’ salivation. Operant behavior is controlled by its own consequences (the word operant is Latin for ‘‘at work’’). For instance, left alone, a baby may kick out or cry spontaneously in the absence of stimuli. Once the behavior occurs, the likelihood that it will be repeated depends on the consequences. So, a tennis player who inadvertently changes grip and immediately hits a series of winners is likely to use the same grip again. One of the premises of this APPROACH is that the human being is

plastic, capable of being molded (the Greek plastikos means precisely this). Behavior can be modified as long as the necessary schedules of

reinforcement can be identified-and, technically, they can be; though, in practice, they are not. Skinner, in particular, has been characterized as basing his theories on the view that the human is a passive receptor of stimuli rather than an active creator of his or her own environment. The tradition dates back to the seventeenth-century philosopher John Locke (1632-1704) and his tabula rasa conception of mind, as a ‘‘blank tablet’’ with no innate ideas. Yet, in Skinner’s theory and, for that matter, Locke’s, the human being is far from passive: in responding to stimuli in the environment, subjects operate actively, discriminating, comparing, and combining. For Skinner, understanding human and, indeed, any animal beha-

vior requires no reference to internal states, such as COGNITION, consciousness, motive, or PERCEPTION. All behavior is subject to the same principle of stimulus-response: As long as the pertinent stimuli and the appropriate schedules are known, any type of behavior is subject to operant conditioning. The emphasis for Skinner and for all those who subscribe to the explanatory power of reinforcement is on the environment-that which lies outside the organism. Applications in sport. Accordingly, applications of reinforcement

theory have focused on changing environmental contingencies, or circumstances, as a way of shaping and maintaining behavior. Desirable behavior is palpable; the chances of making it happen repeatedly can be enhanced by reinforcing it. The results are observable. Inner states, by contrast, are neither palpable nor observable, so followers of reinforcement theory are not interested. Instead, they concentrate on encouraging expedient and discouraging inexpedient behavior. Many coaches, perhaps intuitively, do this: Even something as simple as congratulating athletes after good moves and reprimanding them after errors is reinforcement. More extreme negative reinforcements include benching or not selecting players in team games; punishments might include fining athletes for misdemeanors on or off the province of play. Positive reinforcements are many, of course. Besides the obvious ones already mentioned, there are the less direct rewards that accrue from successful performance: endorsement deals, celebrity status, and the media exposure coveted by so many athletes. Suppose a successful basketball player is offered the opportunity to host

a television show due to start in eight weeks’ time at the season’s end. His contract with his club stipulates that he needs permission and the club allows him to do the television work, but only if he can finish with 3,000 points, a tough though not impossible task to which he rises splendidly, producing some of his best performances ever. Without realizing it, our player is demonstrating the PREMACK PRINCIPLE. Named after David

Premack (1925-), it states that reinforcement is a relative not absolute PHENOMENON. Chocolate will reinforce a child with a sweet tooth, but if they need to run around the block as fast as they can to get the chocolate, they will run hell for leather. Eating chocolate reinforces running. One activity works as reinforcement for another. The activity of having one’s own television show reinforces points scoring for the basketball player. Unlike this example, much of reinforcement’s application in sports

has been geared toward the learning of desirable behavior, such as SKILL, and the maintenance of habits that will either improve or preserve that skill. ‘‘When a particular goal-directed behavior is repeated frequently and consistently in a similar situation, with positive reinforcement, it eventually becomes automatic or habitual,’’ write Tracey Brickell and her colleagues, outlining the process through which skills become ‘‘second nature’’ and require little attention, only the appropriate contingencies. The skilled behavior ‘‘can be elicited by environmental cues without conscious guidance.’’ In 1997, D. Scott et al. disclosed how a laser beam that let off a

beep functioned as a positive reinforcer, signaling a successful pole vault, which eventually led to a modification of vaulting technique. Rewards were used to great effect in a study of golfers’ performances by T. Simek et al. (1994). In this study, the golfers behaved rather like satiated rats. After responding to rewards with good play for a while, they came to expect the rewards, and their performance deteriorated abruptly when they were withdrawn. Sometimes, reinforcement is not immediately apparent. For exam-

ple, Dave Clarke asks, ‘‘Why do problem gamblers continue to gamble when the rewards are few and the consequences such as lack of internal CONTROL are unfavorable?’’ He argues that excitement, the release of tension, and the approval of others are reinforcements, which, when combined with periodic winning provide variable-ratio schedules of reinforcement; that is, where the ratio between responses and reinforcement varies in some random fashion. While Skinner might not approve of the migration of reinforce-

ment, the term social reinforcement has been used in several contexts, perhaps most effectively in the study of EATING DISORDERS. Donald Williamson et al. for instance, found, ‘‘social reinforcement for the thin ideal from family, peers, and the media was correlated with the onset of bulimic symptoms.’’ The CONCEPT of vicarious reinforcement has also been offered to

describe the repetition of behaviors that people observe being rewarded when performed by others. This has clear relevance to the understanding of AGGRESSION, observational MODELING, and VICARIOUS

AGENCY, where stimuli and response are inferred. This involves cognitive processes and, as such, is far removed from the purer concept of reinforcement favored by Skinner.