Myths About Dog Aggression: Part VII
Yesterday, we discussed the popular misconception that aggression from the handler directed at the dog is the only or best way to treat aggression in domestic dogs (or any other animal, for that matter). Today, we'll discuss another popular myth about training solutions for aggression and reactivity.
AGGRESSION MYTH #12: OBEDIENCE CURES AGGRESSION, AGGRESSION SIGNIFIES A LACK OF OBEDIENCE
Many people say, "why don't you just train your dog to sit instead of barking?" At a quick glance, this seems like a reasonable solution. Why can't we clicker train a "sit" and use it to combat aggression? After all, don't clicker trainers advocate the teaching of alternative, incompatible behaviors? I frequently tell my students that I like to focus on solutions and not problems - "don't tell me what you want your dog to stop doing, tell me what you'd like him to do instead. We can train for that!"
Aggression and reactivity are a bit tricky in this sense because I do not believe they are "obedience" issues. In fact, I know of a number of dogs that compete successfully in obedience and a number of other performance quite successfully despite the fact that their dog otherwise fits the criteria for aggression and reactivity. The dog performs beautifully in the ring, but will snap at or bite a child in a heartbeat if presented with the opportunity.
Clients with reactive or aggressive dogs frequently say, "It's like a switch went off. He acted like he didn't even know I was there. He wouldn't respond to his name or listen to any cues, even though he's normally very obedient." For the record, it's possible (and not uncommon) to have a very obedience dog who is also very aggressive or reactive.
The barking, lunging, and biting associated with reactivity and aggression the outward manifestations of an emotional (and sometimes, neurochemical or genetic) response, generally anger or fear. You may be able to "get rid of" the external appearance of the emotion through
A) punishing the behavior - if you punish the dog for growling, he may stop growling. He probably won't, however, be less afraid of or angry at the trigger.
B) reinforcing an alternative, incompatible behavior - you may train a dog to "sit," and use that to combat the outward signs of aggression and reactivity, and often, this has the appearance of working quite well. When it does appear to succeed, it is because this method usually combines some elements of desensitization and classical conditioning - you will not be able to get a "sit" response under threshold so you must gradually work the dog through increasing exposure levels. There are certainly elements of building a positive association here if you are using positive training to teach the behavior. The problem here is one that Jean Donaldson calls a "slick operant" in her fantastic (albeit expensive) six-disc video seminar "Canine Fear, Aggression, and Play." The dog appears to be under control in the presence of the trigger, but his emotions about the trigger have not changed. The cue, rather than exposure to the trigger, is the reliable predictor of reinforcement.
Another nasty side effect of using incompatible, alternative behaviors in the early stages of reactivity and aggression intervention is the potential for poisoned cues. I've seen this a lot, actually. Owners practice "sit" and "down" heavily when the dog is exposed to another dog. Often, if the dog can comply with the cue, he is still uncomfortable about responding - especially when asked to lie down which places him in a vulnerable position. The dog is in conflict, he has to choose between complying with his owner and placing himself in a situation where he feels vulnerable or at a diminished ability to flee the trigger he's being exposed to or defend himself if need be. Imagine being asked to "lie down, close your eyes, take a deep breath and relax" while a stranger crept up behind you in a dark alley and slid his fingers around your throat. Could you? Would it make you feel better? More or less vulnerable?
This conflict leads to ambiguity and confusion - sometimes responding to "down" makes good things happen for him, but other times it makes him feel threatened and as though he can't escape/is powerless. The dog becomes less reliable responding to the "down" cue rather quickly and the trainer is left scratching her head; her and the dog are both frustrated.
Often, treating the signs of aggression or reactivity with operant conditioning techniques, be they what we perceive as "dog-friendly" or "balanced" totally fails to address the issue. You are treating the symptoms but not the cause. It's kind of like giving a cough drop or suppressant to a lung cancer patient - sure, you may temporarily decrease the discomfort or coughing, but the fact is, the cancer is still there, raging and spreading throughout the body. Immediately, it looks like you have solved the problem. Really, the cause has gone unaddressed.
The key to making a dog feel less afraid or upset about exposure to the stimulus is by making the dog feel better about the stimulus. That's the goal of classical counter conditioning and desensitization protocols which treat the root of the problem. These techniques are counter-intuitive - the presence of the trigger, rather than the dog's behavior, must be the reliable predictor of reinforcement. Essentially, we're giving the dog what he wants (toys, food, etc.) at the instant he sees the other dog or person, continually until they're out of sight, at least in the initial stages of training. It doesn't matter what my dog is doing - I don't care if he's sitting, lying down, standing, or if his hackles are fully raised. Yep, I'm going to feed him even if he's noticing the dog and not giving a full-blown reaction, maybe if it's a soft growl or muted "woof," I'm still feeding him.
People HATE this. They get really upset with me, "aren't you reinforcing aggression?"
No. Fear and anger are emotions. "Sit" and "down" are behaviors. When working with fear and aggression, if you treat the cause (emotions) you see the symptoms (behaviors like barking, lunging, snapping, growling, etc.) dissipate. You have treated the cancer instead of the cough.
A final word on this topic is that I absolutely think the teaching of alternative, incompatible behaviors should be the final step in this training. Later this week, we'll talk about whether reactive and aggressive dogs can ever be "cured." Once you have worked through systematic counter conditioning and desensitization, you should absolutely focus on DRI (differential reinforcement of incompatible behavior), which is a fancy way of saying - reward your dog for not being an idiot! Teaching your dog life skills on what "to do" if a situation ever makes her uncomfortable is a valuable part of the training process, but that comes at the finish line rather than the starting line. Teaching your dog DRI and proofing that well will give her skills to survive with reduced stress should she ever find her exposure boundaries stressed in a social situation involving her trigger. I'll come back to this is another entry, but I like to work dogs separately on operant thresholds and classical conditioning/desensitization thresholds - the operant threshold is usually much farther back than the D/CC threshold and is always introduced at a later stage in the rehab process.
Stay tuned for tomorrow, we've got more myths to discuss!