Positive punishment

A consequence in which something unpleasant is added following a behaviour, with the result that the behaviour decreases. The “positive” refers to addition (something is added), not to the behaviour being good or the punishment being mild.

A dog jumping up on someone and being met with a sharp verbal correction, a horse nipping and receiving a swift growl, a cat scratching furniture and being startled with a hiss from the owner, are all examples of positive punishment. The behaviour produces something the animal finds unpleasant, and the behaviour becomes less likely in future as a result.

Positive punishment is one of the four quadrants of operant conditioning and was historically widely used across species in training contexts. Modern applied animal behaviour practice has moved away from it for several reasons established by the research literature: positive punishment can suppress behaviour without teaching the animal what to do instead; it can produce fear and aggression as side effects; it can damage the relationship between the animal and the handler; and it requires precise timing and intensity to be effective at all (when applied poorly, it produces confusion rather than learning).

Where positive punishment is used in modern practice, it is typically very mild, very brief, paired with reinforcement for the desired alternative behaviour, and used sparingly. The behavioural science consensus is that differential reinforcement of an incompatible alternative behaviour is almost always a better choice than positive punishment, and that the cases where positive punishment is the only viable option are rare.