Way back in high school science class, I learned an important lesson about balance. If you placed a iron weight on one side of the scale and a feather on the other, the scale did not balance. It wasn’t that you placed equal numbers of things on either side of the scale that mattered but the mass of the things you placed there. One heavy thing could weigh as much as several lighter things. Finding balance was about finding the right proportions of the items involved.
Today many dog trainers talk about balance. As I have written in a previous article here, the term “balance” seems to mean different things to different people depending on who is using the term. Even the science of behaviour can provide a deceptively simple and distorted view of how to modify behaviour. Looking at a chart of Operant Conditioning, you could get the impression that you should deliver “corrections” as often as you deliver “cookies.” But like that science lab experiment from my high school days, I’m not sure that this one-to-one ratio (or even close to it) is the optimal balance when training our dogs.
Balancing it out
Balancing positive and negative experiences for their dog is something over that dog owners have a lot of control over. Rewards come in many shapes and sizes from yummy food treats to a kind word or a game of ball. The same is true of punishments. From the simple “Ah ah!” to interrupt a behaviour to that fear inducing wrath-of-god correction, we have many ways to communicate that something is “not ok” for our dog to do. The value of a particular reward or punishment may not be as great as another. Do you know which kind of rewards and punishments are the iron weights and which are the feathers? And the value of one correction does not necessarily equal the value of one reward. It would be difficult to provide them in equal measure if their values differ.
The operant conditioning model put forward by B.F. Skinner and others provides four basic types of consequence that can happen as a result of a behaviour. You can add (give) something rewarding, remove something rewarding, you can add (give) something punishing, or remove something punishing. Providing something good or removing something bad improves the targeted behaviour. Providing something bad or removing something good diminishes the targeted behaviour. But do you need equal amounts of rewards and punishments in order to be “balanced?” Our experience of life and living would suggest that is not true. Not by a long shot.
If each moment brought an equal chance of something crappy or something cool happening, I think most of us would be reluctant to leave the house in the morning. Those odds really aren’t that good. The truth of it is that most of life is pretty good. I don’t know about you but things in my life go along pretty well most of the time (75-80% as an estimate). Do bad things happen? Sure. And not always really big, bad, awful things. Sometimes I just hit more traffic than I expected or I run out of popcorn on the wrong night. Bad things are aversive. They are things we will seek to avoid but they will happen over the course of our lives.
Learning to cope
These aversive events in our lives, particularly when we are growing up, are important to shaping how we will learn to cope with difficulties throughout our lives. They are necessary. Studies in other species such as rats and monkeys have shown that confronting the animal with an aversive situation and providing them a behavioural way to “turn off” the aversive can actually change the “wiring” in the brain that handles stress. Animals that have success in finding ways to escape or avoid aversives develop nervous systems that are better able to cope when confronted by unpleasant situations in the future; even if they cannot find a way to escape or avoid it.
So if some unpleasant situations can be beneficial, more punishment is even better for you, right? Wrong. Other experiments on animals exposed to chronic unpleasant situations has shown symptoms of depression and even something that looks like Post Traumatic Stress Disorder. Apparently this “balance” thing is trickier than just providing one punishment for every reward. It seems that there should be enough adversity in our dogs’ lives to teach coping skills but not so much that it kills the dogs spirit or motivation.
Discipline for it’s own sake
All dogs will need to deal with “no” in their lives. No you may not jump up on guests for attention. No you may not eliminate in the house. No you may not whine at me for some of what I’m eating. I have heard many trainers talk about the need to show a dog that they just can’t do those things and that they must be corrected when they show unwanted behaviours. In fact, some trainers even make a practice of setting the dog up in a situation where they do the unwanted behaviour just so they can correct them to teach them what is not acceptable. Some call it teaching “impulse control” or “self control” but it amounts to setting the dog up to fail on the premise of teaching it what not to do.
But there is a flip side to “no you may not.” If you may not jump up on guest for attention, what is the correct way to get attention from guests? If you may not eliminate in the house, how is it that you can ask to be let outside? This is where “corrections” fail to accomplish what their name implies. If I “correct” my dog for jumping up on a guest by knocking her off and yelling at her, in what way does this show her what is acceptable in this situation? It doesn’t. Should my dog be interrupted for this unwanted and annoying behaviour? Absolutely. But I also have an obligation to teach her what to do instead of jumping up for attention so I don’t have to repeat the interruption.
Mark & Reward or Test & Punish?
Let’s get back to that question of balance. Whether we are talking about our dogs or ourselves, I think we would have to agree that a balanced life would be a mostly pleasant experience. Yes, there will be disappointments and things we just can’t do, but on the whole the majority of our time will be pleasant and enjoyable. As a positive trainer, I use a Mark and Reward system to teach my dogs the behaviours I want. They offer me a behaviour and I will give them a marker signal and then reward them in some way to encourage them to repeat the behaviour in that situation again. It works great for teaching my dogs what I want from them. But what about the things I don’t want?
In nature, we could say that animals learn what not to do by a “test and punish” method. You eat the wrong berries and you get sick. You try to take food from the bigger lion and you get a beating for your trouble. You test to see if it’s possible and if it goes badly, lesson learned. It shouldn’t come as any surprise that the earliest forms of dog training took on this “test and punish” model. In order to be more efficient, trainers learned to set the dog up to perform the unwanted behaviour so we could punish them for it. No one wanted to make their dog sick or beat them badly enough to injure them so milder forms of punishment were used. Unfortunately, that meant that more repetitions of the test would be necessary to get the message across.
So how do you spend your time with your dog? Are you marking and rewarding for good behaviour or are you testing them and punishing any time they step out of line? Maybe you do both. What kind of “balance” have you established between the two approaches? Recently a reader of my column commented that he only uses punishments or “corrections” 10% of the time and that 90% of his training involves positive reinforcement for desired behaviours. Should this individual be considered a “balanced” trainer? I think so, yes. But the real test is what his dogs think about it. If then are happy and well adjusted, that’s great.
The imbalance of words
To me, there is something insidious behind all of this talk of “balance” in dog training. It is the implication that if you use positive reinforcement, rewarding the dog for good behaviour, then somehow you will not discourage or interrupt unwanted behaviours. And that’s just ridiculous. A positive trainer is not going to wait until the puppy finishes chewing on her hand – she will interrupt the puppy. It’s as ridiculous as suggesting that a traditional trainer would electrocute their dog for not sitting fast enough.
Trainers have proven time and again that using rewards such as food in training does NOT make a dog’s performance dependent on paying the dog every time they perform the behaviour. The same can be said of clickers or other markers, behaviour doesn’t break down without a marker once the behaviour is trained. Trainers who prefer different methods can’t legitimately claim that they are more “balanced” in their approach compared to positive trainers. Not without first defining what they believe that preferred balance to be and showing some hard data that their approach provides demonstrable benefits for the dog.
We are all trying to help our dogs live a balanced life. Balanced in the sense that they understand their world. Balanced in the sense that they learn to cope with things that are strange and different without fear. Balanced in the sense that they learn to cope with the frustration of not getting things when they want them or how they want them. And balanced in the sense that they have the confidence to work with us without fear of constant testing or punishment.
Until next time, have fun with your dogs.
The first Canine Nation ebook is now available –
“Dogs: As They Are”
Photo credits –
Sheepish – thart2009 2009 From Flickr