Positive Reinforcement To Stop An Unwanted Behaviour?

Why consequences matter in learning

Evidence and terminology reviewed 11 October 2026. This article distinguishes learning theory from NAPET's professional perspective. The original categorical claim that reinforcement cannot reduce unwanted behaviour has been corrected.

Positive reinforcement increases the future likelihood of the behaviour that earns a reward. That definition does not mean that a reinforcement based programme cannot reduce a different, unwanted behaviour. What matters is the behaviour being reinforced, its function, the environment and whether the plan transfers to everyday life.

Teaching an alternative behaviour

Reinforcing an alternative or incompatible behaviour can help reduce unwanted behaviour. Teaching a dog to return to its handler, settle or engage with a different activity may compete with another response. Success is not automatic: competing rewards, fear, arousal and the dog's learning history matter.

Waite and Kodak's 2022 study used owner implemented, function based interventions and reduced mouthing in three dogs. It demonstrates that reinforcement based treatment can reduce an unwanted behaviour. It does not establish a universal solution for aggression, predation or livestock chasing. Read the original study and its scope.

A professionally assessed plan may combine teaching, reinforcement, preventing rehearsal, appropriate physical management and changes to the environment. No method label alone guarantees safety or success.

Extinction and withholding rewards

Extinction means that a behaviour no longer produces the reinforcement that previously maintained it. Withholding a reward is therefore not automatically punishment. Removing an available reinforcer contingent on behaviour may be negative punishment if it reduces that behaviour in future. The classification depends on the contingency and measured effect.

Behaviour may persist or temporarily escalate when a previously available reward stops. This is a risk to assess, not a reason to assume every extinction procedure must escalate or that aversive intervention is always required. Allowing a dog to rehearse dangerous behaviour is not an acceptable way to test a theory. Management must protect dogs, people and other animals while a suitable treatment is developed.

Reinforcement is also a consequence

Operant conditioning describes relationships between behaviour and consequences. Reinforcement and punishment are both consequences. It is inaccurate to contrast reward based training with learning through consequences as though only correction provides information.

A change of tone, body blocking or physical restraint cannot be called operant punishment solely because it interrupts an action. Punishment is defined by a reduction in the future likelihood of the behaviour. Immediately stopping movement and teaching a lasting response are different outcomes.

Safety in the real world

A trained sit does not make a dog safe beside a busy road. A lead, secure restraint and suitable distance from danger are still needed. An emergency intervention may prevent immediate harm, but it does not by itself demonstrate learning or justify a particular training tool.

Near livestock, prevention is essential. Use a lead and appropriate boundaries rather than relying on a claim of perfect recall. Training should help owners manage their actual dog in their actual circumstances, without promises of guaranteed control.

Assessing welfare, not just behaviour

A stopped behaviour is not sufficient evidence of good welfare. Assessment should include the dog's emotional state, stress related behaviour, physical health, recovery, ability to make choices and the owner's ability to maintain the plan.

Cortisol is one indicator whose meaning depends on context and timing. It should not be treated as a verdict on abuse on its own, but welfare research cannot be dismissed as relying only on cortisol.

Cooper and colleagues' 2014 comparison of 63 dogs used behavioural observations and physiological measures. Even with experienced, industry nominated trainers, it reported greater welfare concerns for the electronic collar group and no consistent efficacy advantage. Read the 2014 study.

Vieira de Castro and colleagues' 2020 study of 92 dogs examined behaviour, cortisol and a cognitive bias task. More aversive training was associated with poorer welfare measures. Dogs were recruited through existing training schools rather than randomly allocated, so the design has limits; those limits do not make the findings irrelevant. Read the 2020 study.

When progress is difficult

Prolonged failure deserves reassessment, not blame directed at the owner or an assumption that one tool must be the answer. Pain, health, fear, environment, reinforcement history and owner capability can affect a case. Seek veterinary advice where health or significant behaviour concerns are involved.

Medication prescribed by a veterinary professional can be part of appropriate care. It should not be described as proof that reward based training has failed, nor should individual stories about avoiding euthanasia be treated as evidence that a method caused that outcome.

Defra's dog welfare code provides the wider welfare context. This article is not a set of instructions for using an electronic collar or handling a dangerous dog.

Why this matters to NAPET

NAPET's stated position is to support ethical, proportionate practice, professional competence, owner education and welfare focused accountability. It opposes punishment only training and indiscriminate use of equipment.

NAPET advocates proportionate regulation rather than a blanket prohibition on electronic collars. That is its policy position. It should not be presented as proof that aversive methods are necessary for every difficult behaviour, that reward based approaches cannot work, or that competent use eliminates welfare risks.

Professional judgement should be open to scrutiny. Practitioners should explain the evidence, alternatives, risks, intended outcomes, monitoring and stopping criteria, and work within their competence and the law.

Clear communication matters, but clarity is not a substitute for evidence. Safety and welfare require an individual assessment, realistic expectations and appropriate ongoing management.