“I wish I knew then what I know now.” I think that just about everyone who has been training dogs for any length of time has had this thought.
I consider myself fortunate to have been exposed early on to the science of Operant Conditioning in general, and positive reinforcement (reward-based training) in particular. Someone recommended that I read the book “Don’t Shoot the Dog” by Karen Pryor, and the description of the basic scientific principles was very appealing to my brain, which loves logic, reasoning, and guidelines. I set about going to classes and happily training my first dog, Sunny, using lots of treats in the process. I heard many trainers sing the praises of positive reinforcement, and how much better this was than the old fashioned methods of training dogs which relies on jerking on the dog’s collar with a leash every time it gets out of position and then praising the dog because it is now in the right place. I felt lucky to know a better way. Although I understood the mechanisms at play in “jerk & praise” training, the emotional side of me just couldn’t make sense of acting happy about having to do something to the dog that it didn’t like, and the logical side of me didn’t think it made sense to praise the dog for something it didn’t choose to do.
So I trained with treats, giving a reward every time Sunny got something right. My dog loved training and so did I. I couldn’t wait to get into the ring to show off what we could do. And that’s when I discovered that the true challenge of my chosen sport isn’t so much in teaching my dog to do the tricks. It’s in getting him to perform the tricks in new places, without any food or toys, and while I was nervous as heck and my dog knew it! Right from the start I realized my dog wasn’t the same in the ring as he was in training. He showed signs of stress and didn’t perform some things as accurately or eagerly as he did when he knew I had a treat in my pocket, and that detracted a bit from my happiness with our results. Even so, we had lots of success in the grand scheme of things (Novice A placements, scores in the 190s, a high qualifying rate) and I became hooked on the sport.
By the time I got to my second dog and saw many of the same problems, I was starting to listen more to the naysayers of “positive only” training that advocated “balance” instead. According to this philosophy, the problems I was seeing in the ring are due to the fact that you can’t take food (or toys) into the ring with you, and after the dog is shown enough it figures that out. At this point, when just “want to” isn’t an option any more, you need to balance that out with “have to”. Furthermore, the argument goes, Operant Conditioning as a science describes forms of learning that include “have to”, and those are just as valid and effective as positive reinforcement. I still didn’t want to just jerk my dog around, but the idea of “balancing” the rewards and the physical corrections started to become more appealing as I yearned for results in the ring that were accurate and reliable enough to allow me to compete for top scores, placements in the “B” classes, and the coveted OTCH (Obedience Trial Championship). So I decided to dip my toe into the waters of teaching my dogs that they have to do what I command, even if the reward is not immediately apparent. After all, I was only asking for a measly 5-10 minutes of work at a time in exchange for giving my dogs a life of canine luxury. I was determined, though, to use as little compulsion as I could get away with. I didn’t want my dogs to turn into workers that do everything accurately and reliably but with no spark in their eyes.
I spent the next many years honing my version of balanced training. I completed OTCHs on 3 dogs, two Shelties that belonged to me and one Beagle that had been started by my best friend before she passed away from cancer. My dogs placed nationally each year in a variety of ranking systems. I became one of the voices that argued for balanced training, always making sure though that I taught the dogs *how* to do what I was asking with positive reinforcement, and only adding as much compulsion as I thought I needed any time the dog thought he had a choice.
Still, at times I could see negative effects that physical correction created in my training. I managed to create two dogs that could be slow to front because they were afraid to be wrong. Bright, the first one, never did completely get over it. I realized what was happening sooner with Flare and managed to reverse enough of the damage to avoid deductions. Some days my dogs worked with a fair amount of good attitude, but other times I felt that I struggled to maintain it, especially in the ring. I found myself going to matches so that I could “get in a correction” on some skill that had been sub-par in the ring. At first these corrections seemed to help, but over time my dogs came to know I wouldn’t correct them in the ring any more than I would give them a treat, and too many times I ended up with a dog that just looked stressed when doing that skill in the ring with the same flaw, or worse, that I had been trying to fix.
At times I looked longingly at people training like I did with my first dog, Pez®-dispensing treats to a dog with super bright eyes and I longed for that again. But I thought I knew better, and what results they would get in the ring once they couldn’t provide a steady stream of rewards. Periodically I would take a quick look around the internet to see if anyone had come along claiming to use mostly positive reinforcement to train for a situation like competition obedience where the animal has to perform for periods of time without primary reinforcers, especially in a place where it doesn’t have a previous history of reinforcement. But the people advocating positive reinforcement only, or even just primarily, were usually folks like marine mammal trainers, who can toss as many fish to the dolphins and whales during a show as they wish. Many of the “pure positive” trainers tended to be evangelical about their beliefs, which is a turnoff for me, and lessened my view of their credibility. Some of the things they said didn’t make sense to me. At the same time I didn’t know of anyone in competition obedience who had achieved results sufficient to earn an OTCH or national rankings without some degree of physical compulsion.
So I continued on my path, thinking that I was doing the best that could be done, given the restrictions of our sport, and the fact that there are parts of obedience, such as heeling, fronts, and finishes, that don’t have much instinctual enjoyment for the dog like agility does with its running, jumping and high adrenaline nature. I had a lot to be proud of too. In 2009, my partner at the time, Flare, started the year off with a series of nice placements that put her in first place nationally in the OTCH point rankings. By mid-year I decided to go for it, and we held onto the placement to earn the prestigious award of 2009 Obedience Dog of the Year. This was beyond my wildest dreams, and seemed to validate my training methods and philosophies.
Then little more than a year later, Flare and I were at a match and she just wasn’t putting in effort on her finishes, moving slowly and not lining up straight. If ever there was a dog that should have known how to do what she was being asked to do, this would be a good example! So using the philosophy of correcting lack of effort, I put a leash on her and gave her a firm correction. To my surprise I had to give her a few corrections before I started getting the finishes I wanted again. I’ll admit that the need to do this irritated me and I was not as light about it as I could or should have been.
To make a long story short, from there I noticed that she seemed a little “off” in front, and then it progressed to thinking she had a pinched nerve in her neck, and less than a month after that match I found myself sitting in a veterinary neurologist’s office after she had an MRI hearing the nightmare words “your dog has a brain tumor and likely only a few weeks or months to live”. She was just 8 years old. Later, when I got home and the numbness of shock had worn off, I found myself sobbing – gut wrenching, heaving sobs that brought me to my knees. Yes, I was mourning a lot of things – most importantly the impending loss of my beloved companion with whom I had done so much, but the pain was compounded by the recent memory of having corrected her when I now realized that she was giving her best. As it turned out, a new kind of radiation therapy ended up giving us years more together, and for that I am extremely grateful. But the agony of realizing the unfairness of my corrections at that match stayed with me.
After the shock of the sudden end of Flare’s obedience career, and subsequent decision making and focus on her treatment, my heart just wasn’t into training for a while. When I did get back to it with my youngest dog, Laser, I found myself wanting to mostly train new stuff, where I could use a lot of treats, and stay very patient with mistakes. He was pretty early in his training anyway. It was about the time that I needed to start putting things together that I heard about a two day seminar called “Out of the Lab and Into the Field”, being put on by Bob Bailey and Parvene Farhoody.
I had heard of Bob before, and was aware that he is a huge advocate of positive reinforcement training. He has a reputation for having trained a lot of different species of animals to do a lot of things in real world scenarios, and I had heard of the “chicken camps” that he and his late wife put on to teach people the ins and outs of Operant Conditioning. I had even thought about going to one a few years earlier, but ultimately decided that it probably wouldn’t offer me any new info for the needs of competition obedience trainers than what I had already found from the writings of other positive reinforcement trainers. I assumed that even in the military, animals could be treated as often as needed. But when I heard about this new seminar, the title of it intrigued me, and my heart was in a place that I decided I owed it to myself to at least go and hear what Bob had to say. I figured this would probably be my last attempt to explore the possibilities of positive reinforcement without compulsion to achieve great obedience performances. I did not have high hopes of success.
To my surprise, Bob Bailey was not evangelical at all. He didn’t talk like positive reinforcement is a religion or that those who use compulsion are bad people. He just explained in a straightforward and commonsense way what he believes works the best to get maximum performance and reliability when training animals, which is skillful, knowledgable application of positive reinforcement. He stated that he seldom uses punishment, and then only to stop an unwanted behavior. He didn’t even spend much time talking about the disadvantages of aversives, and for me he didn’t need to because I already was very aware of what those are. But what really got me to listen is his extensive experience training birds, cats, dolphins, and yes dogs too, to work for long periods of time in completely novel environments before they had the opportunity to be given food or other type of reward. They were all trained without the use of compulsive techniques that are so common in dog training, not even the milder ones. These were mostly military applications, the details of which he still can’t always talk about, but he said enough to make me realize that this is someone who *has* been there, and done that, and knows what it takes. He offers the information based on a depth of personal experience unmatched by anyone else I have ever known or heard about, and you can do with it as you please. Furthermore, Bob said some things that seemed to be the opposite of some of the popular ideas about positive reinforcement, and more than one was exactly what I believed from my own observations and experience.
All of this got my attention. I wanted to hear more. When I found out that Parvene had convinced Bob to join with her in a venture to start offering the “chicken camps” again, I jumped at the chance. Over the last three years, I have attended a total of 9 weeks of chicken camps, learning from Bob and Parvene both the scientific principles to a depth I never understood before, and how to apply it using a species, i.e. chickens, with which any kind of “hands on” or compulsion just doesn’t work. I have always wanted to achieve as much as possible with positive reinforcement training, and thanks to Bob and Parvene my belief about what is possible has expanded enormously. I now believe that the best, most effective training methods to meet my goal for an enthusiastic and reliable obedience partner are also the kindest. Operant conditioning was described by B. F. Skinner way back in the 1930s, and it does describe all the ways animals learn, including reinforcement and punishment of various types. However, there is only one part of operant conditioning that has the power to create the "willingness and enjoyment" on the part of the dog that is held as the ideal in the AKC Obedience regulations, and that is positive reinforcement, i.e. adding something to the dog's environment that it wants as a consequence when it emits the behaviors we want to train. Other methods of getting behavior may work, but they will not add to the dog's favorable emotions about the activity, and indeed many of them risk detracting from it due to stress or discomfort. I also know now when I am having performance problems to look first for errors in how I am applying positive reinforcement, rather than fall back onto techniques, no matter how "mild", that erode my dog's enjoyment of our work together and erase the benefits of that history of positive reinforcement that I worked so hard to create in the first place.
It turns out that I was doing a lot of things right. I broke down the teaching process into incremental steps. I avoided accepting “good enough” and instead maintained high standards of performance. I trained and proofed my dogs thoroughly and didn’t enter them until they were fluent in the necessary skills. And all that time spent Pez®-dispensing in the early stages of training created a real enjoyment in my dogs' minds for the training we did together.
What I didn’t understand was how to go from rewarding every little correct behavior to asking for long sequences of behavior before giving my dogs enough reward relative to the amount and difficulty of the behaviors that I am asking for at one stretch, making sure they understand and trust that it will be worth their while. I was too haphazard in the process, and when attitude and thus performance subsequently declined because my dogs thought that they weren’t being rewarded any more, I blamed the lack of food, and then compounded the problem by adding compulsion. The fact that they continued to give me so much in the ring once they realized that neither food or other primary rewards nor physical correction would happen there is a testament to the strength of their early conditioning with positive reinforcement. I never realized how much I was hurting my goals by associating the skills with the stress of compulsion, nor how powerful positive reinforcement can be, until I stopped viewing aversives as anything but a last resort and a sign of weakness in my skills as a trainer.
Laser, my most recent Obedience Trial Champion, earned his title without the deliberate use of aversives in his competition training, and it was the easiest and most joyful for both me and him of the 4 OTCHs I have achieved. I say without deliberate use of aversives because I know I annoyed, discouraged, or confused my dog more than once with bad timing of reinforcement, clumsy manipulation of criteria, and/or too low a rate of reinforcement. But the better I get at these things, the better my results, and I now realize that most, if not all, of my previous perceived need for physically making my dogs do things was created by my lack of understanding of the fine points in applying positive reinforcement. For me it is no longer a matter of balancing "want to" vs "have to". If the "want to" is there, the "have to" is a moot point. The only balance I worry about now in my training is maintaining the individual skills with high rates of small reinforcers vs keeping the dog used to having to perform many skills in a row before it gets a sufficiently valuable payoff. The good news is that the amount of time I spend on training is actually less now, because I don't waste time fixing the damage caused by the "have to".
Does all of this mean that I now use positive reinforcement only, or that my training is free of punishment? That my dogs are allowed to run around willy-nilly doing whatever they please until they hit upon something that I will reward? No, it doesn't. If you go by the strict definition of punishment as defined by Skinner, punishment is "the procedure of providing consequences for behavior that decrease that behavior" (taken directly from one of the slides at the Bailey-Farhoody workshops). So anything I do to cause a particular behavior to decrease is punishment by the technical meaning of the word. This includes, for example, saying "oops" and calling him back (no matter how cheerfully) when my dog starts to stray on a go out, or putting my foot over a treat on the floor when a dog goes to the treat instead of to retrieve the dumbbell, or breaking off heeling and stepping away from my dog (and thus signaling no reward) when he looks at a distraction. I may do any of these things or similar in response to behaviors that I don't want.
However, the words spoken to me by Bob Bailey during a lunchtime conversation at the very first workshop I attended are always in the back of my mind: "If you are having to punish something very often, then there is likely something else that you need to be reinforcing more." I have found these words to be oh so true, especially when it comes to training for behaviors that I want my dog to love doing. I have found that when a dog is willing to choose behaviors other than the one I want, it is usually a sign that the dog either doesn't really understand what I want, doesn't have a strong enough history of reinforcement for it, I am not providing enough value relative to the alternatives or the dog doesn’t believe that I will.
I still wish that I had learned these things sooner, mostly for the sake of my previous dogs. I just lost Flare to that brain tumor a few days ago, 5 years after her original diagnosis. I spent those years trying to pay her back for all the joy she brought into my life, and for helping me to realize where my heart was trying to send me. I am, however, comforted by something that Maya Angelou said – “Do the best you can until you know better. Then when you know better, do better.”
Laura Romanik, trying to do better