One of the biggest reasons dogs make such great domestic animals is that they like to play games, even games they don't get to invent the rules for. What is herding, for dogs, but an enormous game, a game a good stockdog will play just as long as they are permitted?

Games are a fascinating thing to think about. When you try to define what makes a game a game, it becomes very slippery indeed. Lack of seriousness isn't one of the defining characteristics, for example; it takes no effort to call to mind games which people take enormously seriously. And dogs are the same way. It seems the only criterion that is true for all games is that they have rules, and everyone must agree to play by them, otherwise the game falls apart.

Having enrolled Ty in an obedience class, I re-experienced how much dog training has advanced in the past decades. The training methods of the club which sponsors this particular class have apparently fossilized. They are saying exactly what I heard when I first started training dogs when I was, oh, about eleven or twelve. Jerk that chain collar to get their attention. Wait for them to make a mistake and then punish them for it. Roll them over and hold them down so they know who is boss. Most of the class is consumed by explaining exact techniques for physically forcing your dog into the various positions while saying the command word.

I watched as one lady tried fruitlessly to get her beagle to stop sniffing the ground. Jerk, jerk, jerk, no effect. A man with a nervous Weimaraner couldn't get his dog to relax enough to pay any attention to him. In fact, there were only two dogs who were paying attention to anything but the other dogs: a Sheltie, and Ty, who was the only puppy. Ty was focused on me completely, knew all the commands and executed them instantly. Ty has never had a choke chain on him, and I never picked up the slack in his leash. We hadn't done any of the homework, not even once. He'd had about six lessons before classes had ever started, each about five minutes long. In those lessons we covered lie down, heel, sit at heel, stay, come and sit in front, and stand.

It would be a happy story if Ty and I were noticed and praised for our excellence, but the opposite was true. Because I had refused to put a choke chain on Ty, because I never bothered to buy a leather leash (it's easier on your hands when you jerk your dog around, so they are insisted upon), because I offer my dog treats and play with him throughout the class, but I imagine mainly because my dog was uncriticizable while I was being completely disobedient myself, the teacher glared at me every time I caught her attention. I was nothing but a thorn in her side, even though I had never spoken a word.

It is so easy to teach dogs to do these simple things (compared to, say, square inside flanks, or redirecting your dog on an outrun). I was appalled to see so many earnest dog owners struggling to master such outdated methods. I bet I could have picked that beagle's nose off the floor in seconds. Just give her something more rewarding to do than sniff. No agility trainer, no search and rescue trainer, no service dog trainer, uses those old force-training methods, not because they are abusive, but because they work so very poorly.

There is no question that positive reinforcement methods have limitations. They aren't particularly useful when training stockdogs, or sled dogs, or hunting dogs — any activity where true drives are involved. They are also pretty useless for eliminating behaviors, like counter surfing and bicycle chasing. And they stop short of completion, since a truly trained dog knows that "sorry, I don't feel like it right now " is not an acceptable response to a command, and this is simply not acquired through positive techniques. What they are incomparably superior for is eliciting quick and happy understanding of otherwise-meaningless-to-dogs behavior requests.

There is a bridge to cross, from understanding what is being asked, to understanding that it is required — that's a separate and subsequent development. The old method is to start out on the far side of the bridge, with forcing the completed behavior, and then hoping to work slowly back to the understanding part through dogged repetition. Imagine a person who spoke, say, only Latvian, was trying to get you to put your finger on your nose by grabbing your hand and hitting you in the face with it while saying "Slava! Slava!" How long would it take you to figure out what the heck was going on? How fun would it be?

Ty doesn't know he's being trained. He thinks it's a game, one he would play for a lot longer than I have patience for, perhaps called Guess For the Cookie. On a bet, I once trained Bonnie to go across the room and stand with her front feet on a phone book, in about ten minutes. All I did was keep rewarding her best guesses with a dog treat. It's a game people play sometimes play too, it's called Warmer Colder. Aussies learn this game very rapidly.

Raymond Coppinger, the great dog behaviorist, has noted that dogs initiate play with each other by 'play bowing', which signifies that everything which comes afterwards should not be taken at face value, but should be construed as a game. He goes on to say "Instead of threatening our dogs every time we want to train them, we need to perfect the human play bow which tells the dog the games are about to begin." (Barbara Petura interview on Working Dog Web).

I admit I can't resist playing with my dogs. I love the grins they get when they see I'm about to do something really pointless, like put a sock on their nose. I know that when I say, Bonnie! Bonnie! in that voice that means there's a job to do, she flies to me, not because she is the self-sacrificingly helpful Noble Collie, but because it's a game she loves playing: the serious game of stock work.