Good Dogs Bite, Too: Why You Need to Understand the Bite Threshold Model

There is a strange belief in our culture that says only bad dogs bite. Of course, most people believe that their dogs are “good,” and therefore, that must mean that their dog won't bite. While I really do believe that most dogs are good, I also believe that all dogs are capable of biting. It's just that it takes a lot to push most dogs to the point where they might bite.

In her book The Culture Clash, author Jean Donaldson introduces something called the Bite Threshold Model to help explain why a “good dog” might bite. This model states that each dog has a threshold at which he will bite. It looks something like this:

 Each line indicates how much stress a dog can endure before he displays silent stress signals (like freezing, licking his lips, tucking his tail, etc.), before he growls, before he snaps, and ultimately, before he bites. Of course, each dog is different, so the lines on each dog's graph will be at a different height. For example, as a reactive dog, Maisy's growl threshold is pretty low in general, and especially if we were to compare her to a more stable dog. Interestingly, Maisy's snap threshold is quite high (it takes a lot of stress before she will snap at someone), and while I've drawn a line for where I think Maisy's bite threshold might be, I honestly don't know where it actually is since she's never bitten anyone.

Donaldson also says that each dog has “risk factors” that will cause the dog to feel uncomfortable or stressed. These risk factors, taken individually, may or may not cause the dog to display any symptoms. The so-called “good dogs” will usually have risk factors that are quite low- they may not even reach the first line. Others do, but whether the owner notices that their dog is feeling uncomfortable is another matter. The symptoms are so subtle that they are often easy to miss. Here's Maisy's graph, with some of her risk factors added:

As you can see, while crowds of strangers might make her feel uncomfortable, she probably won't show any signs to indicate that. Meanwhile, she will definitely display stress if her head is touched, and the presence of children pushes her right up against her growl threshold- she doesn't always growl at them, but she might.

Most dogs- like most people- don't really want to hurt others. But given the right situation, any dog- and any person- is capable of defending himself or his loved ones. That situation happens when multiple “risk factors” converge into one single incident.

This recently happened with Maisy. Although she's quick to display stress, and even quick to growl, she very rarely snaps. Still, it happens when her triggers “stack” into one unfortunate situation. In this case, I had taken her to a group gathering. What I didn't know was that she wasn't feeling so hot (she had a bit of an ear infection starting), nor did I know that there would be a child there. So when the little one reached out to pet Maisy's head...

Yup, she snapped at the little girl. It all happened so fast that I didn't get a chance to stop the child that had approached. I felt horrible, of course. I'm always sad when I can't protect Maisy, and I hate that others saw her as a “bad dog” that night. I also want to be a good dog owner and keep others safe, and while the child (and her parents) weren't upset by the incident, I felt bad that the girl had a bad experience with my dog.

It's important to note that while the graph presents a logical, orderly progression of behaviors, that's not always the way things play out in real life. Although Maisy did stiffen briefly as the child reached for her, she did not growl. There wasn't time; the triggers stacked too quickly. Behavior also isn't linear. Our dogs are individuals, and so is the way they respond. Not only are each of their thresholds at different levels, but sometimes they will be missing a threshold entirely. For example, people sometimes punish growling, which can result in the loss of that particular line. That means that a dog could go from freezing slightly to biting without ever growling- the so-called bite without provocation.

Perhaps Maisy isn't the best example- although she's never bitten anyone, I'm not entirely sure that others would call her a “good dog.” (I, of course, think she's amazing.) The point remains, though: a series of relatively small things can stack up to the point that even the best of dogs will bite. In fact, sometimes I think these “good dogs” are the most dangerous. Because Maisy has been “bad” so frequently, I have a pretty good idea where her thresholds are, and so I can pretty accurately predict when she's more likely to snap or bite. People with “good dogs” don't have this advantage, and if and when things stack up in just the right way, they will be surprised by their dog's behavior.

If you have a good dog, I urge you to think about situations where your dog's behavior has surprised you. Think about how high each of his thresholds might be, and what types of things are likely to push him to each one. Do thunderstorms unnerve him? What about someone reaching for his rawhide or food bowl? How does he feel about children, other dogs, or men in hats? Pay attention to the times he stiffens up, freezes, tucks his tail, ducks his head, or licks his lips. These are all signs that he is feeling uncomfortable, and that situation should be put on your graph.

I hope that your dog never encounters a situation where all of those risk factors pile up into one big, scary stack, but if he does, the bite threshold model will help you understand why your good dog ended up biting. If you want to avoid such an incident- and who wouldn't?- I encourage you to be proactive. Being aware of your dog's stressors is a great start. If you see things beginning to stack,  avoid a possible incident by simply putting your dog somewhere safe, like in a crate or another room.

More ambitious dog owners might consult with an experienced dog trainer or undertake a counter-conditioning program in order to help reduce each stressor. If you lower the height of each risk factor, then when they begin to stack, the overall effect will be lower as well. Perhaps the dog will only snap instead of bite. And while a snap is still scary and something I take seriously, it is far preferable to having your dog bite.

So don't believe the cultural lie that good dogs don't bite. As the bite threshold model demonstrates, it's not bad dogs that bite, it's stressed ones... and good dogs get stressed, too.