Learning Theory

Imprinting - Non-associative learning - Classical conditioning - Operant conditioning

Extinction - Positive reinforcement - Negative reinforcement - Punishment - Shaping

Negative reinforcement

Negative reinforcement, which is all too often confused with punishment, occurs when an animal learns to behave in a certain way to avoid a negative stimulus or reduce its unpleasantness. This is central to the traditional training of horses since the application of the conditioning stimulus precedes both the desired behaviour and the potentially unpleasant reinforcer. In order to be effective, removal of the unpleasant sensation must immediately follow the correct behaviour. For example, tapping the flank with a whip to prompt sidewards movement is effective if the tapping ceases when the horse moves laterally. Similarly, in dog training, choke chains that are intended to act as a portent of pain can only be effective if their characteristic sound acts as a warning to the dog. Sadly very few people can use these dangerous devices properly by releasing the tension on the leads when their dogs stop pulling. Instead they tend to hang on and launch into the first bouts of a tug-of-war tournament that is endless for both parties, frustrating for the humans and painful for the dogs. Meanwhile many more owners fit choke chains incorrectly, forming a uni-directional ratchet device, eliminating the possibility of reinforcement by relief from neck pain. Giving a verbal warning is the only means of overcoming this problem; hence the 'Heel'-choke approach popularised by Barbara Woodhouse.

Let's suppose that after our begging dog has been taken away from the table, he is taken to the park and trained to walk to heel using a choke chain. Now we have an opportunity to consider, in the table below, how both reinforcement (a titbit of food) and punishment (a yank on the choke chain) can be positive and negative. Be it positive or negative, reinforcement will always make a response more likely in future. Conversely, positive or negative punishment will always make a response less likely in future.

Effect of the treatment

Punishment versus negative reinforcement (and examples using food and choke chains).

Both punishment and negative reinforcement are consequences of behaviour and so are central to operant conditioning. Many trainers claim not to use negative reinforcement but are instead simply confused by the term that may have unpleasant connotations. It is almost somehow politically incorrect. Unfortunately, it seems the fear of prejudice has only served to muddy the waters of animal training. In this context, negative refers to the removal of something from the animal's world while positive refers to an addition. So, when trainers reinforce a behaviour with the removal of something unpleasant, they make the behaviour more likely in the future. The response has been negatively reinforced. Punishment and negative reinforcement are interrelated. By definition an animal must know that a stimulus is aversive in order for its removal to be reinforcing. So, in order to use negative reinforcement a trainer has to have used positive punishment as well.

Negative punishment or omission forms an important part of our attempts to improve or modify responses. Most readers will agree that a dog being encouraged to perform a new behaviour will first attempt to use an established response. The absence of reinforcement at that point makes repetition of the unwanted established response less likely. Reinforcement has been omitted and therefore the dog has been negatively punished. This moves the animal on to try new solutions to its problem. The trial and error process continues. The use in dogs of training discs, as developed by the late John Fisher, relies on omission or non-reward. The discs make a characteristic sound when rattled or thrown to the ground, and are introduced to the dog in association with the removal of a food reward that the dog is expecting to consume. The frustration that may be experienced is compounded by the same consequence on three or four occasions, after which the dog seems to stop expecting to be rewarded. The discs can next be presented at the time of unwelcome behaviour. Perhaps by associating the sound with learned frustration, the dog stops the response for a brief period. This allows the trainer to reward him for stopping.

In the interests of clarity all trainers and animal educators should consider carefully their use of these terms. Punishment is not in itself a dirty word. Nor is negative. Both negative punishment and positive punishment can be extremely mild. The degree to which one relies on either reinforcers or punishers and the consistency and tact with which one applies them are what matters to the animal.

next section