This post is part of the series in response to Dunbar’s 2012 Australian seminars. See index.
Ian Dunbar chooses to look at punishment in a different way to most modern trainers.
When we look at things that are punishing (or are supposed to be punishing) then we can ask two questions:
- Is it punishing (does it reduce the behaviour’s frequency)? and
- Is it painful or scary?
From here, there are four possible scenarios:
- Punishment that is painful or scary, and changes behaviour, is effective but undesirable (that is, it is not desirable to hurt or scare dogs).
- Punishment that is painful or scary but doesn’t change behaviour is abusive.
- Punishment that is not painful or scary and doesn’t change behaviour is nagging. It’s annoying, a nuisance.
- Punishment that is not painful and not scary, but is effective, is desirable. This is where Ian Dunbar chooses to work.
In this way, Ian Dunbar advocates verbal punishment. Verbal punishments can be used to enforce trained behaviours, and is more about tone than volume.
But that doesn’t mean that Dunbar is actually that into punishments. Firstly, the dog needs to be taught the appropriate behaviour (he can’t be punished for something he didn’t know to (not) do). Secondly, the punishment should be instructive and redirective (i.e. propose a solution to the dog – more on this in another post). Dunbar also advocates repetitive reinstruction of instructive punishment (more on this later, too!).
But, importantly, Dunbar defines 8 rules to punishment. These are listed below.
8 Rules of Punishment
1. Punishment has to be effective. Dunbar believe if a punishment needs to be delivered more than once, then the punishment isn’t effective and is more abusive. He suggests that leash jerks, ineffective yelling, and dogs ‘still wearing’ a shock collar are dogs that suffering ineffective, abusive punishment.
2. Punishment has to be immediate. Dunbar said, “To delay is inhumane”, and a delay is anything less than 0.5 seconds. Most owners struggle with delivering punishment within a fraction of a second.
3. Punishment has to be instructive. Dunbar categorized leash jerks as not-instructive. This includes providing the dog with opportunities to not be punished.
4. Punishment must be appropriate. That is, the punishment should fit the crime, with ‘harsher’ corrections for more heinous behaviour. (These harsh corrections should occur within a ‘praise sandwhich’, with lots of praise for the behaviours before and after that were good.)
5. Punishment should only be delivered when appropriate behaviour is known.
6. Punishment should only follow a warning. Dogs should be given the opportunity to avoid the punishment.
7. Rewards are always preferred. It is much easier to reward the one ‘right’ behaviour, than punish a bunch of less desirable behaviours.
8. Punishments must be consistent. A punishment must occur each and every time the dog misbehaves, as inconsistent punishments act as random negative reinforcers (a strong form of reinforcement).
If you’re interested in more insights on punishment, you might like to read Paul McGreevy’s thoughts on punishing dogs.
Pingback: Lure Reward Training | Some Thoughts About Dogs
Pingback: Repetitive Reinstruction (Dunbar) | Some Thoughts About Dogs
Pingback: We have lost words from dog training | Some Thoughts About Dogs