Purely Positive, Force Free,
These days there are so many people that give advice on dogs that I feel compelled to give a brief background on where I am coming from. I started training dogs for performance titles in 1990. I worked at obedience, agility, protection, tracking, herding, and flyball (some of these more than others). I had good success and began a business as a part time Canine Consultant in 1992 to help folks that were having behavior problems with their dogs. From 1995-2005, I worked part-time for PAWS with A Cause training hearing and mobility-assist dog teams. I spent a summer handling one of my dogs in a co-starring role in a movie (NoNames, 2010). In short, I am a dog trainer and have been for a long time.
Perhaps of greater importance for this article is that I am a scientist who has spent an almost 45-year career studying, researching, and teaching about animal learning and behavior. I have published articles in peer reviewed scientific joarnals with titles such as Animal Learning & Behavior and the Journal of Experimental Psychology: Animal Behavior Processes. I have taught college level courses such as the Psychology of Motivation and Emotion, the Psychology of Learning, Canine Behavior and others dozens of times. I retired last spring. I now have the title, Professor Emeritus, which basically means that I still get access to the library and scholarly resources of the university.
After 10 years of college, I received my Ph.D. in Experimental Biopsychology in 1984. I am especially interested in the areas where biology and psychology overlap. During my career, the focus of my research was on employing animal models of human behavioral problems. I was known as a "rat runner", that is, I ran rats through mazes (or operant chambers, Skinner boxes, puzzle boxes and such) to study how they learned, how well they were able to learn after receiving drugs, or after their mother received drugs while she was pregnant, etc. I have also worked with various other species.
Given my background as a dog trainer and as a scientist who studies animal learning, I believe I am in a particularly good position to talk about what science says about how dogs learn. To me the terms "force free" and "purely positive" are buzz words that trainers (or would be trainers) use for marketing purposes. Nobody (reasonable folks anyhow) wants to hurt their dog. So, typically, when a trainer says they do not use force and never do anything bad or negative to the dog, it sounds good. It is what people like and want to hear. I seriously applaud the notion of not using force and not using negative or (what scientists call) "aversive" events in training. On the other hand, there are both pleasant as well as aversive events in the world and we learn from both whether we like it or not. Thus, in training my own dogs, I try to achieve a balance. I strongly emphasize the use of the good stuff and use the aversive stuff in a more targeted way to decrease specific undesirable behaviors because science shows it to be the most effective way to do so. I should note that when I say, "science shows", I am referring to a whole body of literature and not just a few cherry-picked studies.
I find the term "force free" to be confusing at best. Is the use of a leash and a flat-buckle collar considered force? If it is not, it should be. I have seen people use the leash and collar in ways that I would like to erase from my memory if you know what I mean. Ironically, one of these incidents was at a "purely positive" seminar that I attended some years ago. To reduce pulling by the dog, the presenter of the seminar (a well-known, highly-successful, & respected trainer) yanked the leash in such a way and with such force that the dog flipped over and landed on its back. It was a young, but large, Rottweiler. You could hear the thud. As a paid member of the audience, I was truly shocked (no pun intended) to see this.
The term "purely positive" is also confusing. That is because science views positive and negative as mathematical terms and pleasant or aversive as terms that indicate whether the animal is attracted to or repulsed by the event in the environment (or stimulus, as scientists call it). I talk about this more in Confusing Consequences: A Brief Introduction to Operant Conditioning. For even more detail, see Operant Conditioning Paradigms.
In physics, there is the Law of Gravity. It was first put forth by Isaac Newton in 1686. The law basically says that there is a force pulling objects toward the center of the earth. Scientists continue to study gravity, but gravity is something we can count on (that's why it's called a law). Similarly, in psychology we have the Law of Effect, first advanced by Edward Thorndike in 1898. It basically says that if an animal does something and it leads to a pleasant outcome, the animal is more likely to do it again. Conversely, if an animal does something that leads to an aversive outcome, it is less likely to do it again. We continue to study the Law of Effect, but it is something we can count on. The famous behaviorist B. F. Skinner reformulated the law in 1972, when he said that, "behavior is shaped and maintained by its consequences." Those "consequences" are biologically relevant events, or in other words, the pleasant and aversive events we have been talking about. Thus, animals (dogs, people, etc.) learn from biologically relevant events whether they are pleasant or aversive.
In summary, the Law of Effect says that we learn from both pleasant and aversive events. The folks that argue for the force free/purely positive perspective seem to want to have nothing to do with that which is aversive. I understand and am sympathetic to this point of view. But science shows that people and animals do learn from aversive events. Such events are especially good at teaching us what not to do. I believe that whether we want to use aversive events in modifying the behavior of dogs (or humans) is an ethical question.
As an animal researcher, I spent over 15 years on my university's Institutional Animal Care and Use Committee (or IACUC as they are known). This committee decides what research the members of the university can and cannot do with animals. By the way, there is a parallel committee that decides what research can and cannot be done with humans. These committees evaluate the protocols (proposals for procedures and/or research to be performed) individually using a cost-benefit analysis. This analysis looks at the cost to animals involved compared to the potential benefit the research could bring to humankind. As summarized in the table below, in cases where the cost to the animals was low and the potential benefit to humankind was high, we felt it best to approve the protocol. In cases where the cost was high and the benefit was low, we did not approve the protocol. In cases where the costs and benefits were high, tough decisions had to be made.
|Benefit||Low||Maybe||Don’t do it|
|High||Do it||Tough decision|
Early in my career, I was particularly influenced by some tough decisions that were made in human studies. One case report (Lang & Melamed, 1969) involved a 12 pound, 9-month-old baby who was going to die because of persistent vomiting (which explains his low body weight). A few brief sessions that employed electric shock as punishment of the vomiting response saved this baby's life. I believe that in this case, the cost to the baby was high (the electric shock it received was aversive). I also believe the potential benefit was high (saving the baby's life). A decision to approve the protocol was made and the baby's life was saved. I should point out that this is just one example of many such reports. Punishment has been successfully used to treat other life-threatening behaviors in humans (e.g., self-mutilation).
My take on these reports (and that which is taught in your typical college level Psychology of Learning textbook) is that aversive events can be an effective tool in modifying behavior of humans and other animals when used wisely. In Steven Lindsay's scholarly three volume series entitled, "Handbook of Applied Dog Behavior and Training", he takes a similar view. He goes on to provide 20 guidelines for the wise use of punishment. I published a brief summary on the web a few years earlier entitled, "Punishment: Problems & Principles for Effective Use".
In my own training of dogs, I try to keep the use of aversive events to a minimum. A bait bag (containing food treats) and toys are my main training tools. But there are times where I believe the potential benefit to the dog in the long run is worth the cost of using an aversive in the short run. I realize that this is a personal decision. This emphasis on the use of pleasant events in training while reserving aversive events for more targeted use to decrease specific undesirable behaviors is what I am referring to as a balance. When working with clients, I discuss the various options for dealing with a problem behavior in detail and only do what everyone is comfortable with.
In conclusion, science does not say that animals learn better when using a force free/purely positive approach. Science (the Law of Effect) says that we learn from both pleasant as well as aversive events. The use of aversive events is an ethical question which should be analyzed using a cost-benefit analysis. I believe a balanced approach to dog training will place heavy emphasis on the use of pleasant events in training and use aversive events when a cost-benefit analysis suggests that it will be worthwhile to do so.
It is noteworthy that Steven Lindsay also advocates for a balanced approach. In Volume 1 (p. 301), he notes that since the turn of the 1900 century, "the pendulum has swung from a stubborn reliance on punishment and negative reinforcement to an equally unnatural extreme on which the use of punishment and negative reinforcement (in some quarters) is shunned to embrace a so-called "positive" approach to training and behavioral control. Extreme positions, whether based on good intentions or not, are typically based on irrational beliefs and assumptions--not scientific knowledge and experience. . ..Instead of extreme positions, accusatory inuendo, moralizing, and half-truths, what is needed is a balanced and informed attitude regarding the practical use, misuse, and abuse of punishment."
Various professional organizations including the Association of Professional Dog Trainers (APDT), the Certification Council for Professional Dog Trainers (CCPDT), and the International Association of Animal Behavior Consultants (IAABC) encourage the use of the LIMA or the Least Intrusive, Minimally Aversive approach to behavior modification and training. It should be clear that I strongly agree with this approach. The American Veterinary Society of Animal Behavior (AVSAB) has published a position statement on the use of punishment in the behavior modification of animals. While the LIMA approach as well as the position statement of the AVSAB on punishment strongly encourage the use of pleasant events in training, niether excludes the use of aversives. Thus, it seems that the balanced approach that I am advocating is not at all unique or uncommon.