Notes on Dog Reactivity and Aggression from the Michael Ellis Seminar

I’m taking a 2-week Obedience Intensive seminar with Batman at the Michael Ellis School. ME is a fantastic trainer, and I’m learning a ton.

Today, he answered a student’s question about managing dog aggression and reactivity. I wanted to share my notes since we’ve had discussions on this particular problem on the blog.

ME said that, in an ideal world, you wouldn’t allow a dog to get worked up to the level where he would aggress (lunge/bark/bite) at the trigger. Dogs usually aggress as a fear response to something they’re not comfortable with (i.e. skateboarders), or something they perceive as a threat to their resources (i.e. other dogs). Ideally, you would spot the incoming trigger first, and would be able to get your dog to heel or do something for you as the trigger passed. Alternately, you could get your dog to get out of the way or block him, and avoid the thing altogether. (Side note: This all falls neatly into my plan of being a dog walking ninja.)

He said that you don’t want to let your dog “rehearse” the process of aggression because the process itself is addictive. Growling/barking, lunging, biting — the process is an adrenaline-fueled response that releases endorphins. If a dog gets used to doing this frequently, it becomes harder and harder to break the habit, and his trigger threshold could get lower and lower because the dog gets hooked on acting out this process. In addition, the process almost always works in the dog’s favor (scary thing runs away). So not only is it physically addictive due to the chemicals in the brain that are released, it’s reinforcing as an effective strategy for the dog.

For example, let’s say Batman sees a runner doing lunges outside our house. As we walk closer to the runner, Batman’s hackles go up, and he gets more and more agitated. If I don’t redirect him at this point, we will break the threshold distance, and when he’s about 6 feet away, Batman freaks out and lunges and barks at the man. The man startles and runs away. To Batman, what he did “worked.” There was a weird man he didn’t like, and when he aggressed at him, the man ran away. Victory! What do you think will happen the next time Batman sees a weird man he doesn’t like? Probably the exact same thing.

As an owner, I can’t let him rehearse this behavior. What I should have done in a situation like this is get Batman tuned in on me, and reward him with a game of tug or lots of treats. I could also have him heel with his ball past the man, or make Batman sit and watch me while the man passes. If Batman starts to look at the man, I would give him a verbal correction, like “Hey” or “No,” and tap him on the butt to get his attention.

ME recommends being very judicious with corrections when the dog is reacting out of fear. The last thing you want is to amp up a fearful dog because it could increase his anxiety. Then the dog may think, “OK. Bad things happen when weird men do lunges in front of the house.” You want him to think the opposite. “Good things happen when weird men do lunges in front of the house. Maybe this guy isn’t so bad after all.”

ME advises against using the “leave it” command for these situations because it isn’t exact enough. “Leave it” works if there is a precise object you want the dog to ignore, but it’s too vague for an emotional state. ME compared it to being deathly afraid of spiders, and having someone tell you, “Don’t look at that spider,” as a giant spider dangles down a few feet away from your head. It’s very difficult to “leave it” because you know it’s just right there. Instead, he recommends giving the dog an incompatible behavior (sit, down, heel, watch me, etc).

This has three benefits. First, it redirects the dog away from the trigger. Second, by giving him something incompatible to do, it takes the very option of reacting to the trigger off the table. You have now relieved the dog of having to make the choice to aggress, demonstrating that you are in control of the situation — don’t worry about that man, I’m telling you to sit — so the dog does not have to be in control. Third, it allows for a correction if the dog disobeys. You are not correcting him for looking at the man, you are correcting him for not doing what you asked.

If the dog fails to listen and doesn’t respond to a light correction (tap on the butt, verbal “No”), then you must correct him hard enough to get him to knock it off. For me, this is pop-the-collar time. The point of a correction is to create an interruption in the dog’s behavior that allows you to reinforce an alternate behavior. The point isn’t to dominate or hurt him, but to interrupt the harmful action, institute an incompatible alternative, then reward him for it. According to ME, if you use physical corrections (most commonly a prong collar), then you have to learn the mechanics of it so that you do it effectively, with the appropriate force, but also the right timing.

The ideal environment for a dog that is reactive to people is one where everyone just ignores him. ME demonstrated this first hand in class when one of the students’ dogs bit him (don’t worry, it wasn’t Batman — he thinks ME is the best thing since sliced bread). The dog had a slightly nervous temperament, and when ME leaned over to show us something, the dog barked at him and nipped him in the leg. ME just calmly stood back up and kept talking to us like nothing happened. In fact, it wasn’t until later that we learned the dog had made contact, that’s how nonchalant ME was about it. The dog, in response, was confused and didn’t really know what to make of it. He ran back to his owner, and the owner quickly re-engaged him for work.

If ME had freaked out and screamed, “OH MY GOD! THAT DOG BIT ME!” it would’ve made the dog worse. To the dog, ME was doing something weird and therefore was suspicious and deserved to be barked at and nipped. If ME had screamed and flailed about, it would have confirmed the dog’s fears, like, “I thought that man was weird so I bit him, and now he’s screaming and acting crazy. I was right. He IS weird. I knew it!”

That’s why it’s important to practice these exercises with trainers and fellow students if you have a reactive dog. Because you will very rarely get someone on the street who just calmly ignores your dog trying to bite him. Usually what happens is the person gets freaked out and runs away, and/or you have to pull your dog away from the trigger. This is counterproductive because the dog may interpret your restraining him as pack support, and that may motivate him to lunge harder. Also, he learns that by acting out, he gets what he wants. The thing he doesn’t like goes away.

Instead, what you should do is reshape the dog’s habits when confronted with something that makes him uncomfortable. You will never be able to eliminate the dog’s fear impulse or arousal. ME says it like an emotional response. You can’t hit the dog over the head and say, “Don’t be scared,” and expect him to stop being fearful or excited. You can only manage what he does when he’s feeling it. ME recommends rewarding the dog in the presence of the trigger. It needs to be a safe distance away, or else the dog will not accept food (again, imagine being deathly afraid of spiders, and being asked if you want a burger while a giant spider sits on your shoulder). Over time, you can decrease the distance, with the goal of having the dog actually look to you and expect a reward event to happen whenever the trigger appears.

Finally, ME said that there are some extreme cases where the dog has rehearsed bad behavior so frequently, it’s impossible to break it completely. The best option for those cases is to manage the dog’s environment as much as possible (take walks when no one is around, never let the dog loose outside of your own property, etc), so that you’re never in a position where your dog could hurt someone or another dog.

That’s all for now. Sorry this post has become a dissertation. I’m really loving this class, in case you couldn’t tell. I’ll post daily updates on Batmanimal for those interested.