Carolyn claims that herding is the hardest dog spot one can participate in. After Sunday, I’d have to say I agree with her.
Herding is unique in comparison with all other dog sports (with perhaps the exception of the bite work part of Schutzhund) in that you have to actually train and encourage your dog to bite another living creature. It is arguably more difficult than Schutzhund (but perhaps not protection work) in that sheep do not have the added protection of a bite sleeve or suit. As a matter of fact, sheep have extremely fragile skin that rips off easily. When bitten, sheep can develop fly strike, where flies lay their eggs in the open flesh of the sheep and slowly eat the sheep alive. Fly strike can kill an untreated sheep in 24 hours. Herding is arguably harder than protection work in that you are training your dog when to make his own decision on not only when to bite but how hard to bite and when not to bite but to bark instead. Granted, there are cues for the biting (in herding called “gripping”) aspect (“Grief” grip the sheep, “Aus” release the sheep, “Faß” bite the sheep full on “Paß auf” release the sheep from a bite and “Gib Laut” bark), but for the most part, you are leaving these decisions up to the dog, he is, after all, the one with the instinct.
We won’t be working on gripping or biting for many moons, but when Carolyn told us about it, my face drained. I never particularly wanted to train my puppy to bite anything. Then Carolyn told us a story about a 100 pound German shepherd who had been chased out of a herding trial by three very independent sheep. The GSD’s owner didn’t want him to bite anything either, so he (or she) did what is becoming increasingly more fashionable in herding, and put a shock collar on his (or her) dog (hey, who needs longer and longer sticks when we have such innovative new technology, right?) and every time his (or her) dog tried to bite, the dog got shocked. It worked. No more biting.
Unfortunately, there was no more tending either. See, when the dog no longer has gripping in his arsenal of “ways to get sheep back in line” the sheep see that, and the dog goes from predator to prey. Temple Grandin calls this response in dogs “trained helplessness”, where the dog has been so intensely punished that he can no longer act for himself, not even in defense of himself.
I guess sometimes it’s hard to imagine that the same dog who gives you kisses and snuggles with you when you’re sick and sleeps on your bed could easily ravage a 120 pound animal, eviscerate it and feed on its innards. Then again, after Shelby ate that two day old bunny, bones and all, I guess I can begin to picture her devouring a sheep.
Thankfully, with the right training, Shelby won’t be eating any sheep, and she won’t be gripping anything for a long while. It takes two years to get a good tending dog trial-ready and bite work shouldn’t be started before full maturity anyway. So I guess I can put that on the backburner for now and deal with it when it comes along.
What Shelby will be doing is learning to calmly patrol the border, what the border is, when to cross it and when not to. She will also be learning to stop and down immediately on cue, without turning to face me for direction, wherever and whenever. She will be learning directional cues, when to speed up and when to slow down and when to back away. Most importantly, she will be learning to think and make decisions on her own, with minimal interference from me.
I will be learning how to get the sheep to follow me (not an easy task even with good treats), to trust me, how to call them to me. It will take all my attention and multi-tasking skills to be cognizant of the sheep and their movements and emotional states and the equivalent in my dog. I will need to learn when the sheep have had enough of Shelby and when Shelby has had enough of the sheep. Most importantly, I will learn to trust my dog and trust that I have taught her what she needs to know to make the right decisions in every variety of highly stimulating situation.
As I become more and more immersed in the herding culture, I find it curious that herding trainers still cling so strongly to training using physical correction. Independence is especially important in a tending dog, as a tending dog typically works with large flocks of 100-200 sheep. That means that the dog can be working 100s of feet away from the handler, sometimes out of sight. Besides being well trained, the dog needs to know how to think and how to turn those thoughts into actions in a situation where the handler might not be able to give direction.
Typically, when you hear about training a dog to “think” or “learn”, you soon hear “positive reinforcement”. Positive reinforcement is known for creating “thinking dogs”. The easiest way to stifle learning is to punish mistakes. You see it in all living creatures (including people, especially children). For this reason, Carolyn is especially sensitive to us using any aversive when our dog is “on” the sheep (including the no reward markers used in clicker training). Carolyn’s basic concept is – if it isn’t going to cause $150 dollars of damage or more, ignore it. If it is, then by all means, do whatever you have to to stop it. Thus, if your dog is mauling a sheep, go get the dog and pull it away, but if your dog is running in the wrong direction, ignore him and wait for him to come back. Reward profusely when he does come back.
Okay, so here’s the question I know a lot of people are asking themselves – herding dogs have always been taught using punishment, right? And it obviously works, right? And herding dogs that ribbon at trials are trained using punishment, so what’s the big deal?
Well, as a matter of fact, herding was not traditionally taught using punishment, that’s actually a newer development (within the last 50 years or so, right around the time when people started believing that dogs were nothing but wolves in cuter and smaller bodies). The oldest texts on herding that we have describe how the shepherds would smear mutton on their boots to encourage the dogs to lie down at their feet (an extremely important skill for a herding dog to know). Sound like luring to anyone?
And, while many dogs which are trained using punishment do win ribbons, winning a ribbon and working sheep are two different things. For one thing, the sheep used in herding trials are different than the sheep found on a working farm. They’re desensitized to dogs, for one, and most of them already know the course, for another. Carolyn was telling us that the AKC judges are having a hard time with herding trials, because they have done test runs where a shepherd will put his dog in a down stay, and the sheep will walk through the trial all by themselves (other little known fact about sheep – they are smarter than people give them credit for. For example, they can recognize up to 50 different sheep in any given flock, it is the lack of muscles in their face that make them seem dumb to us, because they don’t have many different facial expressions, an attribute that people tend to equate with intelligence).
Dogs that are trained using positive reinforcement using food as a reward, but mainly just the joy of working sheep, are dogs that are being trained to do a job, not just to win ribbons. Dogs that have been trained using operant conditioning with an emphasis on positive reinforcement do better with new sheep, because they have learned to think on their own. They also do better working further from their handlers than the dogs trained using punishment, because they don’t need to rely on their handlers as much as dogs that have been punished and acquired “trained helplessness”.
Lucky for me, Shelby doesn’t have one helpless bone in her body, and Carolyn grins every time she sees Shelby work. On Sunday I heard her muttering to herself, “She’s just a fantastic dog”. I smiled and nodded, then tossed a cookie to Shelby and called “Good!”