Premack gave a group of rats free access to drinking water and a running wheel (Premack, 1959). He recorded their behavior across several minutes. Initially (at baseline), each rat spent some time drinking and some time running, but on average rats spent more time running than drinking: about 250 seconds running and only about 50 seconds drinking. Then, Premack restricted the rats’ access to the wheel: they were allowed to run only after they had drunk a certain amount of water. The rats soon learned the contingency and started drinking water in order to gain access to the running wheel. Unsurprisingly, the total amount of running decreased, because the rats now had to work to obtain access to the wheel. But the total amount of drinking increased to more than 100 seconds, as the rats now performed this behavior more often in order to gain access to the wheel. In effect, the activity of running was acting as a reinforce, and it was increasing the probability of an otherwise infrequent behavior, drinking:
S (running restricted) → R (drinking) → C (access to wheel)
Premack went on to show a similar pattern in human children (Premack, 1959). He put the children in a room that contained a pinball machine and a bowl of candy, and he recorded how much time each child spent playing pinball and eating candy. Some of the children spent more time playing pinball. Premack then restricted access to the pinball machine, allowing these children to play only after they had eaten some candy. Candy eating increased, showing that access to the preferred activity (pinball) could reinforce the less-preferred activity (candy eating). Conversely, children who preferred eating candy in the first place could be trained to play more pinball, by making access to the candy contingent on playing pinball.
Thus, in both rats and children, the opportunity to perform a highly frequent behavior can reinforce a less-frequent behavior. This idea came to be known as the Premack principle. Examples of the Premack principle abound in human life. For example, left to their own devices, most children will spend more time watching television than doing their homework. Thus, watching television is a preferred activity, and it can be used to reinforce the less-preferred activity of homework. The parent restricts television time, making it contingent on homework. As a consequence, the child spends more time doing homework than he would have done if television had not been restricted. A later extension of the Premack principle, the response deprivation hypothesis, suggests that the critical variable is not which response is normally more frequent but merely which response has been restricted: by restricting the ability to execute almost any response, you can make the opportunity to perform that response reinforcing (Allison, 1993; Timberlake & Allison, 1974). For example, perhaps you have a chore, like cleaning your room or doing laundry, that you normally detest. But if access to this activity is restricted, it can become reinforcing. If you have been studying for several hours straight, the idea of “taking a break” to clean your room or do the laundry could begin to look downright attractive. If so, you’ve experienced the Premack principle at work.
Learning and Memory: From Brain to Behavior New York: Worth Publishers, 2007