Old School vs New School

If you haven’t studied animal training recently, your idea of dog training is most likely what’s referred to as “dominance” training. The concept of dominance training stems from the idea that dogs in the wild live in packs and that each pack has an alpha dog who acts as the pack’s leader. The alpha, as the logic goes, is the biggest and strongest dog in the pack. He earns his position through physical dominance and once established, the rest of the pack follows his commands because of that physical dominance.

Classic dog training programs teach you to seek obedience from your pet by establishing yourself as the alpha dog. They’ll advocate such things as having your dog watch you eat before he’s allowed to eat himself because in the wild, as the legend goes, the alpha dog eats first. They also often recommend forceful training techniques, such as physically forcing your dog into the sit position, and sometimes even advocate using physical punishment to stop a behavior you don’t want.

As you’ll see below, there is another school of thought that’s emerged in dog training, and while this school is not new, it’s really come into prominence over the last 5-10 years. It goes by a lot of names - Positive Reinforcement, No-Force Training, Operant Conditioning, and on, but the one thing they all have in common is they leverage the training with reinforcement.

Harper subscribes to this new school, not only because it’s effective, but because its principles apply even beyond the world of dog training. Whether you wish for your pet to understand that he shouldn’t jump on your friends or you want your colleague at work to stop spending 20 minutes at your desk every morning making small talk, the principles and even the methods are the same.

In other words, by using Harper to train your dog, you’ll also learn to train the world around you. That’s some Mr. Miyagi mind-vodoo for you right there.

Leadership is the art of getting someone else to do something you want done because he wants to do it. Dwight D. Eisenhower

No-Force Training

The term “no-force training” draws the clearest distinction between classic and new school dog training methods. No-force training means you do not force your dog into obedience. You don’t get him to sit by physically forcing him into the sit position. You don’t get him to stop jumping on your visitors by kneeing him in the stomach when he jumps. You don’t try to discourage him from eating your shoes by swatting him on the nose with a newspaper.

This does not mean you never have to get physical with your dog. There may come a time when he does jump on one of your visitors and you need to fix the situation by grabbing him by the collar and pulling him off. This however, is not training. It’s what you’re reduced to when your dog hasn’t been properly trained.

Harper likes no-force training for the following reasons:

  1. It’s effective - Have you ever wondered how dolphins are trained? You certainly can’t force a dolphin to jump through a hoop. That no-force training is used with animals that can’t be forced to do anything, like chickens and even whales, is strong proof of its effectiveness.
  2. It works in more situations - If you rely on force for training, what do you do when you can’t apply force? If you’re not there to apply the force? What if it’s not something that can be forced? (good luck forcing your dog to stay and not run out in traffic from 30 yards away)
  3. It can build a behavior as well as stop a behavior - Force is typically only useful as a punishment for stopping a behavior. It’s not as useful in trying to build a behavior. (again, good luck forcing your dog to catch a frisbee)
  4. It’s the kindest form of training - Yeah, there’s no denying that a major motivation for this form of training is it encourages being kind to animals but if that were the only reason, it’s doubtful it’d be this popular.
  5. The pet you love doesn’t fear you - You may think that fear is synonymous with obedience but it often just means fear. If you train the family pet with physical force you may find the pet prefers everyone in the family over you. He may listen to you when you’re around but promptly ignore anything you taught him when you’re not.

Reinforcement Theory

If you don’t use force to train your dog, what do you use? Reinforcers.

A reinforcer is anything that when presented along with an act that is being offered, increases or decreases the likelihood that act will be performed again.

There are positive reinforcers, which increase the likelihood of an act being repeated and there are negative reinforcers, which decrease the likelihood of an act being repeated. Both can be useful training tools, but most of Harper’s training relies on the use of positive reinforcers.

You can call them rewards and punishments, but don’t get caught up in that terminology and lose sight of the behavioral science behind the treat and the punishment. Similarly don’t mistakenly equate positive reinforcement with “good” and negative reinforcement with “bad.” They are behavior modification techniques with no inherent degree of morality or immorality and they both have a place in modern dog training.

There is a lot to learn about reinforcement theory. In fact there are whole departments at universities dedicated to it, although you may find it goes under another name such as “behavioral psychology.” You don’t need to crack open a textbook though. For the purposes of dog training keep these things in mind:

Reinforcers are most effective when they occur at the exact moment the behavior you’re trying to reinforce is taking place

If you like when your dog sits, then you want to reinforce him exactly when he sits, not 10 minutes later. Sometimes this isn’t possible though. For example, if you want your dog to stay while you walk 20 yards away, it’s hard to deliver the reinforcer (aka the treat) from that distance. That’s where conditioned reinforcers come into play. Harper uses the word “yes” to reinforce the dog’s behavior. The reason this works is because the dog is conditioned to associate “yes” with a “treat.” So when he hears “yes” he knows he’s done something that will earn him a treat, even if that treat comes a little later.

You can’t reinforce a behavior that’s not occurring

If your dog never sits, you can never give him a treat for sitting; thus, you can’t show him that sitting earns him treats. To get around this, trainers use something called shaping where they break a complex behavior up into very simple behaviors. In Harper, you’ll find games that use the a technique called “luring.” Luring is essentially a primitive form of shaping. If you want your dog to sit, you first use a treat to get him to point his head up and lean back which causes his butt to go on the floor. You shaped the sit by using his natural instinct to follow the treat. This broke the complex behavior of sitting up into the smaller behaviors of head up, head back, butt on floor.

The reinforcer must be strong enough for the behavior being asked for

This is a simple way of saying you need to find a reinforcer that motivates your dog. Many dogs will be overjoyed to receive a piece of kibble. You might find others that require a more enticing treat like a piece of cooked chicken. Some dogs will even refuse to work for treats at all and instead are only motivated by a toy. Furthermore, what motivates your dog to sit, might not motivate him to leave the squirrel he’s just chased up a tree to come to you because you called. If you find your dog is not motivated, try to find a stronger reinforcer.

Constant reinforcement is only needed during the learning stages

Beginners need to be reinforced every time they do what you want them to do but once a behavior is learned it’s important that you do not reinforce every time. Instead, you want to randomize the reinforcement. By randomizing the reinforcement, for example, sometimes giving it on the first successful rep, sometimes requiring four successful reps, you’ll condition the trainee to keep attempting the behavior, and sometimes even trying to perfect the behavior (e.g. jump higher) in order to get the treat. In Harper, you’ll see this technique employed with games where you jackpot and withdraw treats.

It’s never the dog’s fault

If your dog does not give you the behavior you want, it’s not useful to ask what’s wrong with the dog. Instead, you need to examine what’s wrong with your training. Are you accidentally reinforcing the wrong thing? (maybe you think you’re reinforcing sitting when visitors come, but if you’re not careful with how you offer reinforcement your dog might think you want him to first jump on the visitor and then sit.) Is your reinforcer too weak? Is your reinforcement too infrequent? As noted above, beginners require constant reinforcement. If you switch to randomizing reinforcement too early, it may create confusion. Are your dog’s behavior self-reinforcing? To a dog, jumping on a visitor is fun and its own form of reinforcement. In other words, he’s used reinforcement theory to train himself.