Consequences and Control
Thanks to the behaviourist Edward Thorndike and his propensity for putting small animals in mazes, we have discovered that behaviours are driven by their consequences. That's good news for dog trainers because our job is to change behaviours!
Edward Thorndike's Law of Effect states that behaviours that are met with happy consequences are likely to be repeated and behaviours that are met with icky consequences are likely to diminish. This law is apparent in almost all dog (and human) behaviour and is the driving force behind Learning Theory and behind modern dog training.
If you take a good look at Thorndike's Law of Effect, you can begin to really understand behaviour and more importantly, how to change your dog's behaviour by manipulating the consequences to his behaviours. In a nutshell, if you can control the consequences, you can control the dog!
Let's discuss the four consequences that can follow your dog's behaviour:
| SOMETHING FUN OR YUMMY | SOMETHING ICKY | |
| ADDED | Positive Reinforcement (good thing starts - satisfaction) |
Positive Punishment (icky thing starts - anxiety) |
| TAKEN AWAY | Negative Punishment (good thing ends - frustration) |
Negative Reinforcement (icky thing ends - relief) |
Clicker trainers focus almost exlusively on Positive Reinforcement (something fun or yummy is given to the dog). The use of Positive Reinforcement is the striking of a deal with your dog. You are saying to your dog "If you give me what I want, I will give you what you want". But the dog must hold up his end of the deal first and then you must give him something he actually wants.
Like us, dogs have their own needs and their own agendas and unlucky for us, many of their agendas are in direct conflict with what we want! By using Positive Reinforcement, you will find that you and your dog will be working toward a common goal instead of being in constant conflict. Traditional methods of training are based very much on conflict and the concept that you can train your dog that he has no choice other than to do what you wish. Your dog will always have choices in life and the goal of Clicker Training is to increase the chances of him making us happy by making him happy. By using Positive Reinforcement and the Training Deal (you give me what I want and I will give you what you want), you will find that your dog will begin to work hard to do things for you because doing so will provide him with access to fun and yummy things. Not only will he make good choices, but he will make them all on his own. It's a win-win situation!!
Since you are going to focus on Positive Reinforcement, it is important to know what it is. Positive Reinforcement according to Karen Pryor, is "something the dog will work to achieve". That can be cookies, tug games, access to other dogs, playing, patting, anything the dog will work for. It is important to remember that the dog gets to pick what his reinforcers are, not the owner. When I worked at a large training complex, there was an agility ring next to the pet class ring. The pet dog owners in class would be very discouraged as their dogs were so fixated on the lucky agility dogs who were running through tunnels and flying over A-frames, while the poor pet dogs were doing sits and downs. When I saw the dogs who were fixated on the agility dogs, I would walk over to see what the problem was, and I would see the owner waving a Cheerio in front of the dog's nose in a desperate effort to regain attention. They would give up and say to me "my dog is distracted by the agility dogs" to which I would reply, "he is completely focused on the agility dogs, it is the Cheerio waving in front of his face that is a distraction"! In this case, the Cheerio was not viewed by the dog as a reinforcer. This was obvious as the dog would not work to earn it. At this point, I might pull a liver brownie out of my pocket and invariably, the dog would suddenly rediscover his motivation and work would resume. The Cheerio might have been something the dog would work for in the owner's kitchen. But the Cheerio did not have enough value in the face of distractions so was not a reinforcer at that moment. Food and fun are only reinforcing if they cause a change in your dog's behaviour, regardless of the value that you assign to them. If you do not see a change in your dog's behaviour when you wave the Cheerio in front of his face, the Cheerio is not a Reinforcer!
Reinforcers and punishers can not only be positive and negative, they can also be Primary and Conditioned.
A very simple definition of a Primary Reinforcer is something the dog naturally wants. Food, play and interaction are all Primary Reinforcers: things that dogs naturally value and seek out. A Conditioned Reinforcer is something the dog learns to like because it has been consistently paired with a Primary Reinforcer. For example, dogs do not innately like riding in cars. After a few trips to the dog park however, the dog learns that rides in the car often precede the opportunity to engage in fun, so car rides will so carry as much value as the dog park itself.
Conditioned Reinforcers are important to Clicker Trainers for two reasons. First, when you teach your dog with Positive Reinforcement, you will build value for the behaviours you are teaching. If every time you say Come to your dog and he complies, he earns food and fun, you are reinforcing coming when called. Your dog is learning that every time you say Come, food and fun are very likely to follow. You are making coming when called a Conditioned Reinforcer. This means your dog will assign a lot of value to recalls and will be much more reliable at coming when called as the behaviour itself is a Conditioned Reinforcer!! This reinforcing capacity is also explained as the Premack Principle.
Conditioned Reinforcers are also important to Clicker Trainers because the clicker itself becomes a Conditioned Reinforcer for our dogs. The repeated pairing of the click and the treat means that the dog will learn to assign a lot of value to the clicker and will work very hard to earn a click. Think of the clicker as money. Both money and the click are Conditioned Reinforcers. By themselves, they have no value at all. Babies don't care about money and dogs who are new to clicker training don't care about the clicker. Through experience we learn that money can buy us access to Primary Reinforcers such as food and fun, so we learn to really value money. Your dog will also learn that clicks will earn him food and fun, and he will be eager to exchange behaviour for clicks.
Premack's Principle states that more probable behaviours will reinforce less probable behaviours. Every child understands the Premack Principle as they learn that eating their vegetables (a low probability behaviour) is reinforced by getting to eat ice cream (a high probability behaviour). If high probability behaviours are made contingent on low probability behaviours, then the lower probability behaviours are more likely to occur. Returning the ball to your hand after fetching it is a low probability behaviour for your dog. If doing so earns him the reinforcement of you throwing the ball again, he gets to engage in the high probability behaviour of ball chasing. Ball chasing reinforces the lower probability behaviour of returning the ball to hand.