I recently stumbled across this article: "What makes an aggressive dog, and how you can spot one." It's not the best piece of science journalism, since the substance of the article gives you basically no information about how to "spot" an aggressive dog. However, it does summarize the findings of this study: "Human directed aggression in domestic dogs (Canis familiaris): Occurrence in different contexts and risk factors." Since everything but the abstract of that study is behind a paywall, it's nice to have an article summarize some of the findings.
I'm not one to take one study as gospel, particularly a study that's based on surveying dog owners rather than on direct research, but it does confirm a number of things that are found throughout the contemporary literature on dog behavior. The article and study have a number of really interesting things to say about dog aggression that are worth reading about, including a lack of correlation between breed and aggression, as well as a finding of significantly less aggression in dogs who've been through puppy classes. However, for right now, I want to use it to highlight the relationship between punishment and aggressive behavior.
This study, like several before it, found yet another strong correlation between dogs who are trained with aversives and dogs who show aggressive behavior. Contemporary dog trainers push owners to avoid yelling, hitting, leash corrections, e-collars, and other dog training tools that involve pain, discomfort, or intimidation, not just because we love dogs and those things aren't nice. We also push people to avoid those things because we see the very real consequences when dogs sometimes become reactive and aggressive. There are many causes of aggression in dogs, and obviously not all dogs trained with aversive tools become aggressive, but there is abundant research demonstrating an increased risk of aggression when dogs are trained with intimidation, discomfort, or pain.
I can't tell you how often I have to encourage a client not to yank up on the leash and yell "no" when their dog stares at another dog and then barks. They want him to stop it—and fair enough, he ought to—so they punish him for it by using a stern voice (intimidation) and a tug on the collar (an unpleasant sensation). However, while some dogs might learn to stop the behavior when subjected to leash and voice corrections, the handler isn't really addressing the fact that the dog is nervous, energized, or possibly even afraid when he's barking. In fact, more often than not, the behavior actually becomes worse, because the dog is barking out of a kind of anxiety, and the owner is showing the dog he is absolutely right to be anxious. After all, if you're anxious already, wouldn't it make you more anxious if somebody yanked on your neck and yelled at you?
Instead, we work on building a relationship where you learn to catch the dog when he looks back up at you, and you reward him. He begins to learn that if he gets nervous, he can look at you, and you'll do something calm and nice. You work to create a positive feedback loop so your dog learns that you are a source of safety and trustworthiness. Over time, through repetition and reinforcement, this becomes a habit that runs deep enough to overcome nearly any situation.
As trainers, we have to get out of the habit of doing things that are uncomfortable, intimidating, or even painful to dogs. I've said before that it's not sporting to punish a dog in that manner for misbehaving when it's often a lack of understanding of context that makes the dog misbehave, but it goes beyond the fact that it's not fair. It's also counterproductive and even dangerous, and it stands to reason that the more you rely on intimidation and discomfort, the more you run the risk of creating fear, aggression, and other truly serious problems.
So while it might seem to make sense to be "old school" and use collar corrections, prongs, chokes, e-collars, and rolled up newspapers to punish behaviors you don't like, the more research that's done, the more we learn how counterproductive or even dangerous those tools are. Modern trainers don't just recommend dog-friendly training because we're froofy hippies who think dogs are people. In fact, many of us used those more old-fashioned methods back when they were the best tools we knew of at the time.
But we know better now, and we have methods available to us that are kinder and carry lower risks of side effects like aggression or a reduced trust. This study is just one in a long line that confirms that intimidating a dog or using discomfort to shape behavior can double—yes, double in this study—a dog's chances of showing aggressive behavior towards strangers, and triple a dog's chances of showing aggressive behavior towards a family member. This study does not prove that these punishments cause aggression, but they underscore the concern in a growing body of scientific literature that dogs who are punished in this fashion are more likely to develop reactive and aggressive behaviors than dogs who are trained through more progressive, dog-friendly means.