Operant conditioning
Learning by consequences: behaviour that is rewarded becomes more frequent, and behaviour that is punished becomes less so.
Operant conditioning is learning shaped by consequences. An animal, or a person, does something; if the result is pleasant, the behaviour becomes more likely, and if it is unpleasant, less likely. Edward Thorndike set out the basic idea as the "law of effect", and the American psychologist B. F. Skinner developed it in detail, studying rats and pigeons in a chamber now known as the Skinner box.
Psychologists use precise terms. Reinforcement makes a behaviour more frequent; punishment makes it less frequent. "Positive" means something is added and "negative" means something is taken away. A treat for a dog that sits is positive reinforcement; an alarm that stops when you get out of bed is negative reinforcement. The common mistake is to think negative reinforcement means punishment. It does not: it still encourages the behaviour.
By rewarding small steps towards a target behaviour, trainers can build up surprisingly complex actions. During the Second World War, Skinner's Project Pigeon trained pigeons to peck at an image of a target on a screen, rewarded with grain, so that their pecks could steer a guided bomb. One bird pecked more than 10,000 times in 45 minutes. The project received $25,000 in government funding but was cancelled in 1944; a revived version ran from 1948 until 1953.
The same principles are used today in animal training and teaching, and in the design of games and apps that reward frequent use.
Coming up in future stories
- The Missile Steered by Pigeons 1943 · Minneapolis, USA