Monday, October 19, 2020

Why I don't use punishment or aversives in training.

One of the many things that people will read about when looking for a trainer or behaviourist is what method of training the dog trainer will use. Most modern trainers are taught to use "positive" or "reward based" methods. Some trainers will use "balanced" methods which utilise punishment and aversives, while others are completely punishment based. Thankfully the completely punishment based are very few and far between, the balanced are a dying breed, and positive is the way forward.

"What the fuck are you on about?" I hear you ask. Well here goes.

For many years, people were taught that dogs need to be shown "who is boss", "who is dominant in  the pack" and that their dog should be "submissive" to their human. These ideas were based on a flawed study involving captive, unrelated wolves and their behaviour around each other and resources such as food, mates and so on. A bit like "Big Brother" on tv, but with wolves. 

These wolves did not know each other from Adam, and being confined only had access to limited resources as they were unable to move out from the area and go hunting and seeking the things they wanted. So they fought over the things that were available to them. Naturally, the bigger, healthier and stronger wolves were able to be more competitive than their smaller counterparts and received a bigger share of the resources. So, the humans decided that these wolves were dominant and "Alpha", and this is how a pack would behave, and so to make the dogs act like the seemingly "lower ranking" wolves who were less aggressive, we must bully them into being submissive. 

The trouble is, wolves do not behave like that in a natural setting. They are more like a family, with mum and dad, taking care of the rest of the family, older siblings that can go off and start their own families and they rarely fight over resources at all. The pecking order as to who gets fed first varies from day to day, and the wolves jobs within the pack is also fluid. Then there is the fact that although they have a common ancestor, dogs and wolves are two entirely different species. One evolved to stay away from humans, the other to hang around with us instead, and their brains think and react differently as a result.

So the study of how wolves behave was wrong. The scientists who observed these wolves then spent a bloody long time trying to tell the world of their mistake, but by this point the damage was done. Trainers in village halls, and TV trainers were telling people to yank their dogs with prong collars, zap them with shock collars, do "alpha rolls" which involve pinning the dog to the floor until they stop wriggling, biting the dog's ear and so on, and because the dogs were afraid, for a temporary period the treatment works. Punishment, by definition, means to reduce a behaviour. Punishment must work, or it would not be punishment.

However... with using punishment in training there are flaws. Not just that it is not very nice to zap, bite, hit or yell at your "best friend" while you try to get them to "submit" to you - to me that sounds like domestic violence, and if a human did that to another human, it would be viewed incredibly dimly by other humans and the law, but also, when the unwanted behaviour stops, so does the punishment. 

Which means training has stopped. If a dog has practiced something, they have gotten good at it, even if that thing is an unwanted behaviour. Now, the dog has probably reached a level of tolerance to the punishment, has stopped being trained not to do the unwanted behaviour, and so the unwanted behaviour can creep back in and need a harsher level of punishment before the behaviour stops again.

Punishment does not tell the one whose behaviour is being punished what you DO want instead. Try asking your partner to do the washing up, by saying nothing but "no" from the moment you see them, until they get right what you want them to do. Now just simply say "it would be really nice if you could wash up, please" and see how much quicker they understand.

Think of the dog who has been told "no" countless times when it comes to chasing the cat. The dog, might stop chasing the cat if told off, often enough. But he loves chasing the cat, plus he is also getting used to being told off, so when he goes to chase the cat and being told off no longer cuts it, what next? Should he be shouted at more harshly? That might work quickly at first, but when that stops working, what next? Hit? What when that stops working? Kicked? Zapped with a shock collar? The punishment is spiralling and even though there are interludes of the dog deciding to behave appropriately and not chase the cat, eventually the chase of the cat becomes more valuable to him than the fear of the punishment, especially if the behaviour has not been punished for a while so the dog is less bothered by it, and so the cycle begins again. He may even begin to blame the cat for the bad thing that happens to him and then you have a whole new set of problems. 

Now imagine, if instead of chasing the cat, the dog was firstly managed with stair gates, or a long line, so he could not actually physically get to the cat (after all - positive does NOT mean just let the dog do as he pleases). The dog could be taught calmness in his daily life so as to not be so aroused to feel the need to chase so often. The dog could be taught to disengage from the cat. The dog who is really, really, reeeeaaaally adamant on chasing the cat as he has a really strong prey drive and can't help himself could be rewarded for doing something else still - he might very well enjoy a chaser tuggy (a fluffy toy on a rope for tugging and chasing) or a flirt pole (like a cat's toy on a string but much bigger and the string is attached to a horse lunging whip so he can run around very fast, chasing it) for example, and so they can indulge their desire to chase and catch something fluffy or squeaky, and actually be rewarded for doing so. If the reward is reinforcing enough, the dog's behaviour will change.

By practicing behaviour in an appropriate manner, the dog is getting good at an appropriate behaviour and both human and dog are winning - and as a bonus no cats are eaten in the process. The better the dog behaves, the more they get rewarded. In the meantime, management means he is not practicing chasing the cat and so that behaviour is becoming slowly extinct. Training continues for the dog, with every praise. How much better is that for the dog and human relationship? How much nicer and fairer is that for the dog? After all - it is us who want the dog to behave in a certain way. The dog couldn't give a fuck. He just wants to chase something. So it is up to us to make it worth his while.

Positive training has never been about letting the dog do as they please, or being soft, despite popular thinking - if anything it takes a bigger skill set than punishment, as the trainer has to work out how to lead the dog forward in an appropriate way that works. They need to work out what makes that dog tick, what reinforces his behaviours. A good trainer with good knowledge and a good skill set will be able to do this - if you need help with your own dog, ask a qualified professional (like me ;-) ). In truth it might take a little longer at first than simply zapping, yanking or yelling at the dog each time they do something you do not like, though with adequate management most behaviours can be instantly stopped. 

The effects of positive training are much, much longer lasting than punishment and mean relationship building, and help to teach your dog that the world is a good place, you are a good person, how to behave appropriately and be happy about doing so - and is that not what we want for our best friends?