The best training philosophy I ever heard was from a friend who prefers to remain anonymous. She said, “I believe in training that is fun for both the teacher and the learner.” She said a lot in a very small package.
Here are some of the details about my own approach, and I hope they boil down to the same thing.
I call myself a positive reinforcement-based trainer. I am a LIMA trainer. That stands for Least Intrusive Minimally Aversive. I follow the Humane Hierarchy as described by Susan Friedman. The Humane Hierarchy shows how to start with behavior interventions that are the most empowering and least aversive to the individual animal.
Now here is the part that trips some people up. Because the Humane Hierarchy goes up to and includes negative reinforcement and positive punishment, some people say that following it equates to condoning shock, prong, and choke collars and other training aids. Nothing could be further from the truth, in my case and that of many others. The Humane Hierarchy is not a permission slip. Neither is it a slip-and-slide. I’m an amateur, but I’ve never made a training plan that included positive punishment. I haven’t followed one with negative reinforcement for several years. I would not put a shock (or other “training” collar) on my dog. I DO draw a line. Drawing a line about certain practices and following the Humane Hierarchy are not mutually exclusive. In fact, I think they are implicit in the warning and stop signs on the graphic itself.
The interesting part of the Humane Hierarchy is at the bottom, anyway!
The idea of the Least Intrusive Minimally Aversive approach is that aversive methods are used only when other methods are insufficient or fail. And I would add that it is important that the methods are insufficient or fail according to an experienced trainer and after having been tried over a reasonable time period. Otherwise, you are just a cheater who is abusing a really awesome tool.
Murray Sidman wrote,
Many are willing to accept restrictions on the use of coercive therapy, agreeing, for example, to use coercion only when no positive procedure solves the problem. In principle, I cannot dispute that well-meant and sensible condition. In fact, I believe that the prerequisite—nothing else works—is rarely met. I would go so far as to say to anyone who claims to have tried everything else, “Tell me what you did,. I will then suggest a procedure you did not try.” Undoubtedly, I would sometimes be unable to do this—but not very often. –Reflections on Behavior Analysis and Coercion, Behavior and Social Issues, Double Issue 1993, Volume 3, No. 1 and 2
The Humane Hierarchy is not perfect. There are antecedent arrangements (such as deprivation) that are intrusive. There is the whole problem with negative reinforcement being on the same level of negative punishment and extinction. Though in the end, I have to agree with Dr. Friedman’s response. Negative punishment and extinction can both be pretty brutal.
LIMA and the Humane Hierarchy are embraced by two of the big animal training organizations. To me, having a hierarchy of interventions is really beyond criticism, even though it’s not perfect. How could we possibly be better without one?
Intervening in Behavior
I acknowledge that intervening in the behavior of another being is a serious thing to do, not to be taken lightly. I have learned to commence that task by asking: “What does the animal want?”. Not “What do I want?”. Think about the animal first, then figure out if there is a way it can have what it wants and still live in harmony with you and other creatures in the household or environment.
Sometimes the animal can’t have what it wants in the context in question. In that case, I try to change the behavior with something different but more reinforcing to the animal than what the original problem behavior got them.
I do not use the term, “force-free,” but there are people in my community who do so with the purest of intentions. Not always with success, though. Permit me one snark when I say that I have seen several trainers who claim to use only positive reinforcement build a training plan around negative reinforcement.
I am not “all positive” or “purely positive” and am not fond of the terms. They are generally thrown out as straw men by trainers who defend aversives, as a way to “prove” that humane training is impossible, ridiculous, or even in some cases, secretly no different from what they do.
As has been pointed out by many trainers, since the operant learning nomenclature uses “positive” in a special sense and it can apply both to reinforcement and punishment, the waters get muddy immediately. In addition, I have not witnessed it to be possible to train with positive reinforcement only in real life. (By the way, positive reinforcement is not the first suggested intervention on the Humane Hierarchy.)
That being said, there are some great trainers who are coming closer and closer to a purely positive reinforcement approach. For myself, it’s very important to think about the ways I teach things and get creative about training behaviors without negative punishment or negative reinforcement (or positive punishment of course) when possible. I know for me it is easy to get set in my ways, and something really nice can emerge when I apply myself to thinking about a familiar behavior as if it is the first time I’ve taught it, and not fall back immediately on negative punishment for instance.
I applaud anyone who learns about the processes of learning and takes a deliberate step “down” the hierarchy to a less intrusive, less coercive method. Keep going!
Finally, a special word about negative reinforcement. I have made several videos that demonstrate possible fallout from negative reinforcement and/or show it in a critical way. They are “Negative Reinforcement vs Positive Reinforcement,” “Teaching a Dog to Back Up Without Body Pressure,” and “How Shock Collar Training Works.” I did this because negative reinforcement (R-) is often misunderstood and is also quite insidious, entering into our training and in our lives without our knowing it. It can have fallout, just like punishment. One of the reasons I write and make videos about it is that R- happens so easily in life all the time, and it comes so naturally to us. I think it should be handled with care. That being said, I have used it deliberately in a few different training situations in the past, and use it accidentally now and then in life with my dogs (but not as part of a protocol). What is important to me is to identify when I am using it, be mindful of it, be honest about it, keep watching the dog to observe the response, and continue to improve my own training and behavior.
Addendum October 2014
My answers to the World Dog Trainers’ Motivation Transparency Challenge: