Schedule-Induced Behavior

Description of the Strategy

Schedule-induced or adjunctive behavior, first reported by John Falk in 1961, is generally defined as excessive behavior produced by a schedule of intermittent reinforcement but not required by the schedule. Although several types of schedule-induced behavior have been reported in a wide variety of organisms, including drinking and air licking in rats, aggression in pigeons, and movement and eating in humans, the prototypical example of schedule-induced behavior is schedule-induced polydipsia (excessive drinking) in rats. In a standard experimental preparation, fooddeprived rats are fed small, 45 mg food pellets approximately once per minute in a Skinner box with water freely available. Over several sessions, rats gradually develop a robust stereotyped pattern of postfood water consumption, in which a drinking bout of 5 to 30 seconds in length occurs a few seconds after the consumption of each pellet. Rats can drink 15 to 30 ml of water in 30 minutes and half their body weight in 3 to 4 hours. A domestic Norway rat normally drinks about 10 to 20 ml of water in a 24-hour period.

In most studies, only the total amount of induced responding is measured. Yet each adjunctive response has several distinct characteristics: magnitude, duration, probability, and postreinforcement latency. During acquisition, the magnitude and duration of the response generally do not change greatly, the probability increases to a plateau—sometimes occurring after 100% of food deliveries—and the latency decreases to just a few seconds. Food deprivation is not required, but greater deprivation generally produces faster acquisition and a higher response probability. Adjunctive responding is most likely when food pellets are delivered once each three minutes. Under some conditions, however, such as when interpellet interval lengths are varied within sessions, strong adjunctive drinking bouts, clearly distinct from ordinary drinking bouts, can occur under interpellet intervals as long as 16 minutes and can also be seen when sessions include only a single-pellet delivery trial and no schedule at all. The type of food can affect drinking, with powdered and most liquid food producing little adjunctive responding relative to dry food pellets. The size of the meal does not seem to have a definite relationship to probability of the adjunctive response. The fact that rats will lick an airstream in place of water suggests that the rats are not drinking due to the mouth-drying effects of food. Moreover, induced aggression in pigeons has the same general characteristics as drinking in rats and cannot be attributed to ingestive factors.

The origin and function of adjunctive behavior have not been resolved. It is not superstitious, maintained by contingent reinforcement, or conditioned by Pavlovian processes. Some suggest that adjunctive behavior serves timing functions or reduces anxiety or frustration due to interruption of eating. In terms of topography and development during acquisition, adjunctive behavior resembles sensitized elicited responding—behavior that is strengthened by repeated elicitations. This makes evolutionary sense. The reduction of drinking through habituation by repeated consumption of small meals would lead to dehydration. Sensitization would result in the extra hydration necessary to process food consumed in this way. Aggression in pigeons, [Page 500]similarly, might arise because smaller, spaced meals might imply a shortage of resources requiring strengthened defensive responses relative to situations with sufficient food. Excessiveness, therefore, might not be a fundamental feature of adjunctive behavior. Excessiveness is due to the repeated elicitation of discrete, strengthened reflexive responses.

...