Punishment is used in training to change behavior. It is a consequence the dog wishes to avoid.
Let’s use crooked sits as an example. If your dog performs a crooked sit in heel position, you’ll want to make that behavior less likely next time. How might you use punishment?
You can with-hold (-) something that the dog wants like cookies, toys or praise when your dog sits crooked. You are using -P.
Or you can add (+) some sort of physical discomfort technique for the same purpose (collar correction or other physical manipulation). Your dog won’t want you to do that in the future so straight sits become more likely (+P). Or you can verbally harass the dog (mental force) to make him uncomfortable, so next time he will try to avoid that by sitting straight (also +P),
If it makes you feel better you can actually straighten the dog (maybe give a hand signal/cue or pull on the collar to make the dog “fix it”) but that’s not training – the dog already did the behavior wrong (coming into a crooked sit out of heeling), so it’s somewhat irrelevant. Go ahead and fix your dog if you want, but recognize that at best you’re teaching your dog how to correct a crooked sit (useless in the ring) and at worst you’re teaching your dog not to want to sit near you at all.
I’m not a fan of punishment in either the + or – form, so I try to avoid it. Instead I prefer to isolate problem areas when they crop up and work on them away from my “flow” training. More often than not I don’t say much to the dog to let them know when they’ve made an error, and if I do punish I normally use a cheerful interrupter along with withholding reinforcement (-P). Whether I continue with an incorrectly performed chain depends on the dog. I consider temperament, hardness, experience level, and what I think caused the error. Lack of attention or effort is treated differently than a startle, which is treated differently than a proofing exercise that I set up on purpose, and each of those is considered within the context of the dog’s temperament.
I don’t dwell too much on quadrants – they make my head hurt. I just ask myself if the dog likes what I’m doing or does not like what I am doing. I try to do things that my dog likes so that he wants to be working with me and I try to avoid things that my dog does not like so he won’t find training stressful or unpleasant.
What happens if you punish a dog for an error, and you attempt the behavior again with no additional training?
It will depend on your dog’s understanding of the exercise. If they understood the punishment and are capable of performing correctly, then hopefully they will change their ways. You would then want to “mark” this correct attempt in some manner to cement their understanding; possibly using praise, a reward, or the continuation of the chain.
What if your dog fails again?
Two in a row suggests a training problem; your dog does not know how to win. And since training problems are your responsibility….
Stop testing and start training.
If you insist on allowing your dog to fail while you continue to punish, one of three things will happen.
1. Nothing. If your dog is generally stable and doesn’t care too much about what you’re doing, he’ll just stick it out and maybe figure it out or maybe not (dogs often learn in spite of us, not because of us). In this case your dog compensated for you – super! Stick with stable, easily motivated, clear headed dogs and you have a lifetime strategy, but somewhere in your head tuck away the fact that your dog is doing more than his fair share of the thinking.
2. Your dog will become distressed. What your dog does next is a function of temperament. Low energy or lower drive dogs tend to give up – they don’t know how to win so they opt out. That might look like wondering or sniffing or just working so slowly that you can’t stand it anymore. High energy or higher drive dogs that aren’t winning show frantic or frustration behaviors, such as whining or barking in work, or showing hectic, hyper, or zooming energy. Sometimes those behaviors are misunderstood, so I want to be clear; those hyped up behaviors are not your happy dog delighted to try harder after being punished (that’s not logical), that’s how your dog looks when agitated from either +P or -P. Like the first group, these dogs may or may not figure out what you want, but if you manage to condition your dog to either stressing down or stressing up in work or under unsureness, you’re going to have a heck of a time getting rid of that reaction any time your dog realizes that work is going to be hard or stressful. If you create an unpleasant CER to work (even if it looks “happy”), you’ll pay the price in competition when your dog’s brain freezes up, and it can be a very hard problem to solve. Been there, done that.
3. Your dog will become careful and methodical. If this is the case, you may well get very good scores in competition, but then you’ll want to know why your dog shows so little energy in work. Because they can’t; they might make an error and it’s just not worth the punishment. So they will give you exactly what you have trained for and not an ounce more.
Here’s a better plan.
If your dog fails, pull the behavior out of any flow training or chains that you might be working on. Whether you tell your dog that they have made an error and how you communicate that is up to you – depends on the dog.
I don’t even bother to work on the error at that moment if I don’t want to – I just make a mental note to work on it at some point.
Now figure out a way to communicate what you want so that your dog can practice perfect. This is the hard part because you might have to do some creative thinking to find another way to explain what you want that is more clear to your dog. Sometimes that even involves teaching the same exercise many ways until you find one that makes sense.
Now, at some point put it back in a chain or flow training and see what you have. Don’t be in any hurry – the more you focus on a problem area the more you’ll sensitize your dog, and soon they really will be neurotic. Just add it in a bit here and a bit there….
Looks good? Great! Reward it at the end of a very short chain a couple of times to make your dog even more enthusiastic about their competence and then reintroduce it into your regular training behavior chains.
I try hard not use intentional punishment very often because there are side effects. This includes punishment that is withholding an expected reinforcer or withdrawing my attention (-P) as well as punishment that is adding an unpleasant physical or mental correction (+P). In all cases punishers work because the dog wants to avoid them – that is the definition of punishment. The more things your dog finds that they want to avoid, the less fun they’re going to have in training and dogs that aren’t having much fun with the training experience have a way of running into endless challenges as a result of their low confidence levels.
I do use punishment in the form of -P but I try to make it minimal, and I consider the specific dog. I minimize the effects of punishment through the use of cheerful interrupters (see earlier blogs) so that I can separate out loss of reinforcement from loss of interaction. It is also extremely rare that I withdraw my personal attention (a form of -P) because I place ultimate value on keeping my dog’s attitude up, and withholding both a classic reinforcers (cookie/toy) AND my personality is simply too great a punishment for the majority of dogs.
I’ll post a blog soon to show this approach in action.