Monday, July 11, 2011

Behavior Chains

Alright, I am about to become a dog training nerd here and talk about something called behavior chains.

Last weekend my family and I visited my husband's extended family in Oregon for a memorial service. The day before the service, most family members were all staying at the same house (as were we), preparing for the service and visiting with one and another. A few of them had brought along their dogs.

Try as I might, I could not seem to turn off my "dog training thinking cap." This was difficult as it is my profession! I did however, try to not meddle or attempt to give too much advice on some of the naughty dog behavior I witnessed.

This brings us to behavior chains. There are two types of behavior chains: ones that we create on purpose, and ones that were created on accident. The ones that are created on purpose usually arise from teaching a dog a command or trick that involves several steps. A behavior chain is when there is a reinforcer after multiple behaviors. Just having a dog sit, then getting a treat, isn't a behavior chain. But having a dog do an agility course, is a behavior chain. He is doing several taught behaviors beofre the end reward. If we link undesired behaviors without breaking the chain, we create an accidental behavior chain.

For example, you want to teach your dog that when he hears the doorbell, that he should run to his dog bed, lie-down, be quiet and stay until you release him. You have to teach him each step individually, then put them together until the doorbell is it's own cue and laying on the bed becomes rewarding as he knows after the whole sequence he gets to get up and say hello.

The behavior chains we create on accident usually aren't behaviors we want to enforce. However, many owners that train at home or use novice trainers, will fall into the behavior chain trap.

Last weekend I saw my husband's uncle's hound dog run off repeatedly. She would literally take off the property and run down the country road to see the neighbor's livestock and start baying and making hound dog noises. He would call and call and call, send his border collie to go get her and eventually get up, and run down the road within about 10 feet of her, then she would come. At his point, he would bring her back to where he was sitting before, turn her loose again and the whole ordeal would start again. He was essentially creating a behavior chain. The dog runs off, he calls her over and over, she doesn't come until he is super close, he gets her back to his chair and she does it again. To her, it is a great big game! She loves running off and she gets to continue to run off till her heart is content. Now I did suggest to him that if she doesn't come that she needs to be tied up for a while before let loose again OR not have the gate open so she can't get loose. He replied with what a lot of owners are confused about, "I don't want to punish her for coming back." But did she come back really? How many times did he call her? How close did he get before she finally did come? Did she learn her lesson and not run off again?

A behavior chain I see with novice owners/trainers is people greetings. In my classes, I like to teach the dogs to sit before getting pet by a stranger. This avoids them jumping on the stranger or other unwanted behaviors like crotch sniffing or leash wrapping. The accidental behavior chain that can be created is owner approaches stranger, dog jumps on stranger, owner asks dog to sit, he does, then gets rewarded with a treat and pet from stranger. Now some of you may not see the flaw in this, but what was just created is a dog that thinks he should jump up, THEN sit, then get rewarded. He thinks the jumping is part of the behavior he got rewarded for. How do we prevent this? Multiple ways. First, we can be proactive and ask the dog to sit as the stranger is approaching and reward then, thus distracting the dog from jumping, and reward again for remaining seated once the stranger has left. Another approach is that if the dog does jump, the stranger leaves (taking away that reinforcer), the owner turns around and re-sets to try again (thus not enforcing a jump/sit).

Another behavior chain that makes me just cringe is something I witness at the dog park. Dogs are happily playing and then suddenly there is a fight. Now if this were me, I would break the fight up, apologize, and promptly leash my dog and leave. I am not going to reward my dog for fighting by letting the play continue after the fight has been broken up. What I see happen is fight gets separated (usually in a dangerous way), dogs are then pushed into a down position NEXT to each other, then drug over to the benches and forced to sit near each other for about 2 minutes, then are let loose again. Sometimes the fight starts again, sometimes not. But I have seen many of the same fighters, fighting again on different visits. Why do the owners do this? One owner at the park told me why those people do it (and sadly, he thought it was the correct way to handle a fight). He said the dogs were being forced to submit to each other because they were fighting over dominance. Sitting near each other later, the owners were reinforcing that the owners were in charge and that they WILL like each other. They then release their dogs to play again because they drove all the way to the park and their dogs aren't tired yet, so why go home?

Forcing a dog to lie-down next to another is not making them submit to the other dog. They aren't doing it voluntarily, so how can that even be possible? It will only make your dog think YOU are mean and scary! Having the dogs then sit next to each other doesn't MAKE them like each other. Letting them loose to play again can create a behavior chain. I fight, I lie-down, I go again! Letting the dog loose to play again is reinforcing. I don't get why the average dog owner doesn't see this.

So that is a behavior chain. Have you created any by accident with your dog that you wish to fix?

No comments:

Post a Comment