You Really, Really Don't Need to Use Force to Train Reactive Dogs

I’ve heard a lot of arguments from balanced trainers (and those who support them) regarding their use of aversive tools (including but not limited to citronella collars, e-collars, prong collars, choke chains, leash pops, etc.). But no argument breaks my heart quite like this one that showed up in my inbox after I questioned a rescue’s use of balanced training partners:

"[Positive training] worked great with obedience and general training, but I have not had great success with the behaviorally challenged dogs… Most dogs can be rehabilitated with positive based methods, but true behaviorally challenged dogs are often not among them."

I was devastated when I read this. Because not only does positive reinforcement absolutely work on behaviorally challenged dogs, but it is vital that force is not used on these dogs.

Using force, punishment, and aversive tools on a fearful, reactive, or aggressive dog is a bad idea for a few reasons. Among these are:

1. Risk of redirectionIf a dog is in the middle of an aggressive reaction to a trigger, and you punish the reaction with a prong collar pop or a shock from the e-collar, you are at risk of worsening the reaction and ending up on the receiving end of a bite. The only time I have ever been worried about a bite from JJ was during a reaction like this one, where our e-collar “trainer”* was present and instructing us. I forget about this sometimes, because it feels like a lifetime away. JJ was another dog (so much more scared and confused), and I another human (also so much more scared and confused). In her reactions since we crossed over to positive training, she has never come close to redirecting again.

2. Suppression of warning signals Behaviors like growling or air snaps are what dogs use to signal to other dogs (or humans) that they need distance. Using punishment on a reactive dog "works" by suppressing the outwardly visible aspects of a fear or anxiety response, like these warning signs. However, by punishing a growl out of a dog, you have effectively taught them that their warning signs do not work to get others to back off. So what is the next step? A more severe behavior, like biting. Do not punish your dog for barking or growling. These are important behaviors that they are using to give us information about their emotional state. Without them, you have a dog that will seemingly snap "on a whim", which is a lot scarier for everyone involved - including the dog.

3. Worsening of negative associations with the triggerWhen you punish a dog for their reaction to a trigger, you risk your dog interpreting the punishment not as "I bark and lunge, and this unpleasant/painful thing happens" but rather as "that stranger walked by, and this unpleasant/painful thing happens." Over many iterations, punishing a fearful or reactive dog can actually worsen their aggression because they begin to association the already scary thing with this new, painful stimulus. This is what happened to JJ when we started out with the e-collar "trainer."* Her aggression to dogs worsened, and we are still working through repairing that damage with positive reinforcement.

When you use positive reinforcement, you are addressing the emotional state of your dog.

This is not the same as “obedience.” This is not simply teaching your dog to sit in the presence of the scary thing. This is helping your dog to feel safe and confident around the scary thing by turning it into an actually-good-not-so-scary thing. So how can we go about this?

The goal is to pair something scary, like other dogs, with something incredible, like hot dogs. This method is known as classical counterconditioning, and is primarily what we use with JJ on walks. The key goals are to keep your dog under threshold - aka don't get so close to the scary thing that your dog freaks out - and always follow the scary thing with the amazing thing. On our walk with JJ this morning, this is what this looked like:
Another dog was coming toward us, so we ducked off the path to give JJ enough space. As we were moving away, I gave her treats each time she looked at the other dog (getting the treats out after she saw the dog). She was disengaging easily to look back at me, and not taking the treats roughly, so I knew she was still feeling okay about the dog being there. As soon as the dog was out of sight, the treats went away.

ALL dogs can learn this way. Classical conditioning is not a mode of learning that is unique to certain dogs or other species. If you're having issues seeing progress, you may be operating at a distance that is too stressful for your dog. Your reinforcer also may not be of a high enough value or motivating enough for your dog. Progress will be quicker if you use a reinforcer that is high value, as well as unique to that particular trigger. For example, we fill a squeeze tube with canned food that JJ pretty much only gets when she sees other dogs. If your dog is not eating (and medical problems have been ruled out) they are too stressed. Do not punish your dog for reacting. Instead, back up to a distance that your dog can handle. Or, go inside. Take a deep breath and try again tomorrow, making it easier for both you and your dog.

There are other positive-reinforcement based ways to work on reactivity and aggression, but as someone who transitioned from balanced training to force-free training, I have found that classical conditioning is a great starting point. And as always, I am happy to help troubleshoot or help you get started.

*Note that I have put "trainer" in quotations when referring to the people at the company we worked with that used e-collars. As I have learned about behavior and learning theory, it has become clear that these people did not have any understanding of how dogs learn or how to appropriately and safely address reactivity and anxiety. They also were not certified trainers through the CCPDT. I do not consider them to be deserving of the title of "trainer."