Lately, there seems to be a battle raging on the dog training front between trainers who believe that you absolutely must be the alpha in the relationship with your dog, and those who insist that coercion never works and you are far better off with positive reinforcement.
Personally, I have seen coercive training work. I’ve even used it myself. I’m not advocating it as the first course of action, but when all else fails, there really is nothing like a good, old-fashioned alpha roll, and I don’t think that makes me an abusive dog owner. Sure, if you’re flipping your dog on his back and making him submit all the time, then you’re getting something wrong. But in serious situations (for instance, if your dog snaps at you when you try to take something away from him), it can be the best, most immediate way of communicating very clearly “You do not get to do that, and you had better not ever do it again.” Your dog is not going to fear you forever because of one roll, is not going to run and hide every time you come near him, and is not going to stop loving you.
That said, if you’re always relying on coercion, you’re getting it wrong. And you’re running the risk of a dog that is at the top end of the assertiveness scale fighting back as a matter of course. It’s a bad pattern to set. Think of it this way – right now, you might be in a position to overpower your assertive dog, but what happens if you become ill or injured? Who’s going to have the advantage then? By the same token, if your dog is naturally submissive, you could be running the risk of “fear biting” or even your dog shutting down emotionally.
So, where do you find the balance? Let’s talk about the most common mistakes you can make when training your dog, and how they relate to the coercion model of training versus positive reinforcement. Some of these mistakes relate directly to coercion versus positive reinforcement. Others are simple common sense.
Trainers who insist that you always have to be the alpha will often suggest than any bad behavior on the part of the dog should be punished, or else he will never know the difference between right and wrong. The reality, though, is that a dog can learn without being punished. Instead of punishing the dog for what he did wrong, teach him that if he doesn’t behave, something good will be taken away from him.
Consider this scenario – you have your dog at the park, and he wants to go over and be friends with a person. He becomes overly enthusiastic, and jumps on the person. You can handle this one of two ways. First, you can yank on the leash, pulling him away from the person, and that way you will be telling him that something bad will happen if he jumps. Second, you can ask the person that your dog is approaching to wait, put your dog into a sit, and have him wait to be petted. In the first approach, you are teaching the dog that a bad thing will happen if he jumps. In the second approach, you teach him that if he waits and is polite, something good will happen – he will get petted. Positive doesn’t mean permissive – the dog has to wait.
The key here is not to punish the bad behavior, but to reward the good. It’s just as easy, and usually more effective.
I’ve talked in other articles, and in detail in Dog Training Made Easy, about the usefulness of treats when it comes to training your dog. Some trainers refer to this as “luring.” Basically, you show your dog what you want him to do, and when he gets it right, you give him a treat. This is perfectly acceptable, but you have to structure it so that your dog only gets treats sometimes. Otherwise, you can end up with a dog who sees no reason at all to behave if there’s nothing in it for him.
So again, you don’t want to punish your dog if he doesn’t do what you want, but treats shouldn’t dictate behavior. There could come a time when you want your dog to come to you immediately in order to avoid danger, drop something that he shouldn’t have, and so on. Therefore, what you want to do is “fade out” the lure. An example on this would be working on “down.” If your dog is accustomed to lying down on command and then getting a treat, then hold the treat in your hand, make him lie down, and then withdraw the treat.
This isn’t coercion or abuse. It’s simply a way of getting your dog to do what you want him to do without having to constantly bribe him. Release him from the down, put him in a sit, and then give him the treat. That way he doesn’t connect the reward with the “down” command all the time. Mix it up a bit using different commands.
You can use this procedure with any behavior that you have taught your dog. As soon as he does what you want without the need for luring, move on to another exercise. And of course, tell him that he is a good boy. Your praise is as much a reward as a treat.
Clicker training is certainly useful, but your dog can end up just as dependent on the clicker as on treats. You want your dog to respond to commands whether or not they are reinforced by means of treats or clicking. What happens if you leave the house without the clicker? If your dog runs into traffic and doesn’t know that he has to come to you whether or not he hears the clicker, you can bet that there will be no good outcome.
This is where you need to start working on what is known as “intermittent reinforcement. Once you’re sure that your dog will respond each and every time to the clicker, skip a click or two and offer praise instead. Eventually, your dog will respond to your commands with or without the clicker.
Most people start off using clicker training along with treats. So when you start weaning your dog off the clicker, use treats along with praise. Then, from time to time, use praise alone. Ultimately, your dog will respond simply to your voice.
Keep in mind the concept of intermittent reinforcement, though. What you want, over the long term, is for your dog to do what is asked of him without the need of any type of reward. So switch it up from time to time, offering treats, praise, nothing at all, and even rewards that don’t involve treats or kind words. For instance, if your dog rushes the door when people come to visit, a reward could be letting him go out in the yard after he has sat and waited for your visitors to greet him. Think about what your dog wants, and from time to time, let him have it. You’re not being permissive; you’re being kind and reasonable.
Of course, other times, he’s going to have to stay inside. Your dog will learn to behave because there is a possibility, not a guarantee, of a reward. You could think of him as a gambler who might win, might break even, or might get nothing at all. Either way, he’s probably going to do what you want because he feels that he just can’t take the chance.
You have probably heard of the “three second rule,” and no, this isn’t the one that suggests that if you drop a cookie on the floor, and pick it up within three seconds, it’s safe to eat because germs won’t have had enough time to get onto it. This is the rule that many trainers suggest when it comes to correcting your dog – that if you don’t correct the behavior within three seconds, it’s too late, so you might as well move on and wait for the next instance of bad behavior. The fact is there are no hard and fast rules, although I would suggest that if you wait more than five seconds before offering a correction, that is probably too long given a dog’s very short attention span.
I try to be Speedy Gonzales on this, and correct my dogs the second they misbehave, but of course I know that doesn’t always happen. Between the time I go “What the heck have you just done?” and the time I correct my dog, I suspect that I’m probably nearer the three or four second mark. Regardless, the point is that you’ll get the best results when you offer the correction (or, for that matter, the positive reinforcement), mere seconds after the event.
This is where you have to make the decision to punish or not. Coercive trainers will tell you that any infraction of the rules should result in a scolding, or in severe cases, an alpha roll. Positive reinforcement trainers will tell you to simply remove the dog from the scene of the offence, and move on.
I’m not sure that either approach is invariably correct. Much, I think, depends on the dog. Simple mischievousness is probably best handled with positive reinforcement. If it’s a serious dominance issue where your dog is really challenging you, though, think about whether you can live with this down the road, and about the seriousness of the situation. Real aggression issues may sometimes call for more extreme measures.
As to timing, immediate verbal cues or a quick click can work wonders. You still have to do it the minute the bad behavior happens, but it gets the dog’s attention and gives you a bit of breathing room to decide what you’re going to do next.
Poor timing, on the other hand, can actually reinforce the behavior that you’re trying to correct. This confuses your dog, and is going to end up frustrating both of you. Going back, for instance, to the idea of greeting guests, let’s imagine that you’ve worked with your dog and he pretty much knows that he’s supposed to sit and wait before greeting guests. Your guests arrive, and your dog sits at the door. Then he decides, for whatever reason, that he would prefer not to sit, and he jumps on a guest while you are fumbling for the clicker. By the time you click, he has already jumped, and what you have actually told him is that he is a good boy for jumping!
This is why you always need to accompany clicker training with verbal cues – you might not always have the clicker at hand, but you can always speak. So get the timing right, and never rely on the clicker alone.
This sounds harsh, doesn’t it? And yet, it’s so important when it comes to effective dog training. You can’t be lazy. You have to take an approach to training, and you have to stick with it.
Now, let me say that this doesn’t mean that you necessarily have to be alpha rolling your dog all the time, or yanking on his collar. Certainly there will be people who will tell you that you’ll never get your dog to behave if you aren’t the boss, but that’s not necessarily true. You don’t have to be on top all of the time. You do, however, have to let your dog know that you are not his littermate. You are his friend, his companion, and also his leader. You do not necessarily have to punish him, but you do have to let him know when he is engaging in behaviors that are not acceptable. You want to deliver cues that tell him what is expected of him, and you do not want to “poison” the cues. It’s your job to deliver cues in a kind, consistent fashion.
What this means is simply that you do not punish your dog for failing to obey. If you ask him to lie down, for instance, and he fails to do so, and you punish him, you have poisoned the cue. He now has a negative association with that command. Rather than doing what you want, he will actually be less likely to obey. He will be stressed, and maybe even traumatized, and less willing to work with you. This can ruin the relationship that you have with your dog, and you might never be able to get things back on an even keel again.
When it comes to dealing with your dog, you always have choices. If one thing isn’t working, then you can try something else. It is always best to avoid force whenever you can, and try to use positive reinforcement to get the behavior you want. Constant use of force is the way of the lazy trainer, and it makes for a very bad relationship with your dog. There are almost always better ways than coercion when it comes to dog training.
Coercion if Necessary, But Not Necessarily Coercion
I talk about Neila a lot in my blogs. That’s because we’re not just friends at the dog park – we’ve been friends for years, because we share a passion for big dogs: me with my Boxers, and Neila with her Rottweilers. Over the years, we have walked our dogs together, babysat one another’s dogs, and shared training tips. I’ve learned a lot from the way that Neila handles her dogs, and I’d like to think that she’s learned from me and my gang as well.
Last week, I was over at Neila’s house for coffee. During the course of our visit, her youngest, Dallas (just about six months old at time of writing), somehow found a dead rat in the backyard and brought it into the house. Neila made a grab for it, as she’s accustomed to doing, since all her Rotts are trained from day one to give up anything they shouldn’t have. Apparently Dallas had other ideas, because when Neila made a grab for the corpse, Dallas snapped at her.
What happened next took mere seconds. Neila grabbed Dallas’s collar, flipped him on his back, scolded him, and extracted the dead rat from his mouth. Then we finished our coffee and conversation.
I really can’t assign any blame to Neila on this one. Dallas could have choked on tiny little rat bones, and besides, he was offering up a challenge that couldn’t reasonably be ignored. Neila’s babies are spoiled rotten, but they don’t get to snap at Mom. My perspective on what Dallas did, and the subsequent alpha roll, is simply this – he needed that.
I might point out, too, that Neila does not make a habit of alpha rolling. But she knows when it’s needed. She knew that if Dallas got away with snapping one time, he’d probably do it again. He’s getting his “big boy” teeth right now, too, so if he did manage to make contact, the outcome likely wouldn’t be all that good for Neila.
I have another friend who had a Golden Retriever. Note that I said “had.” The dog had very little training, because my friend worked a lot of hours and, by his own admission, didn’t have a lot of time for the dog. My friend, Max, always meant to do more with the dog, just as soon as he had the time. The upshot was that he did no training at all, whether coercive or positive reinforcement. One day, the dog bit a child who was trying to take a ball away from him. Max had to have his Golden put to sleep.
Max isn’t my friend anymore. He neglected training, and his dog paid the price. Any type of training would have been better than nothing at all. I suppose he was permissive in a sense, but in a way that shouldn’t have happened. Permissive training doesn’t mean NO training. He had no control whatsoever over the dog, and what happened was probably inevitable. I’d rather have seen him coerce the dog into behaving than do nothing at all, although of course the better course of action would have been to train in some way – in any way.
Simply stated, all dogs need to be effectively trained if they are to be good canine citizens. Most of the time, permissive training works best. But any training at all is better than no training whatsoever. An untrained dog is a disaster waiting to happen, both for the dog and for anyone that the dog encounters.
So, there are great divides between dog owners and trainers who advocate coercive training as the best means of control, and those who go to the other extreme, assuming that dogs should always be allowed to be dogs and express themselves accordingly. I’m a big believer in working with dog nature wherever possible, and avoiding punishment and coercion whenever possible. But I stand by what I said at the beginning of this blog – sometimes, there’s nothing quite as effective as a good alpha roll. Only as a last resort, though.
Positive reinforcement is always the better course of action. But there can be certain circumstances where you just have no choice but to step up to the plate and let your dog know that certain behaviors will not be tolerated. Firmness and abuse are two very different things, and sometimes, if you are firm, it can actually work to your dog’s benefit.
I am very much in favor of permissive training. But coercive training, even if it is the lax way out, is better than no training at all.