ClickerSolutions Training Articles

On Punishment

This article was originally posted on the Click-L mailing list. It is reprinted here with the author's permission.

If you go back over the years, in my posts I talk primarily about positive reinforcement. There are occasions where I have used other procedures. Like most others, I rather routinely apply time outs, but I always ask myself "Did I create this problem by not enough, or even ill timed, reinforcement?"

I accept that most behavioral problems are the creation of the trainer, and I am perfectly willing to go to the wall and bang my head three times when I make a mistake. I have NEVER taken out my frustration on the animals I am working with, and I demanded the same of my people. Trainers must accept responsibility for what they do or don't do.

When I have really punished, and that means REAL positive punishment, it was always to save lives, property, or some other such reason. I would never think of using positive punishment for anything like standard obedience training. First, the need is not great enough, and there is a better way to achieve the level of behavior, including reliability.

If we are talking about dogs risking their lives and the lives of the handler under wartime conditions, I would not hesitate to take that extra step to punish an error, rather than to rely on positive reinforcement alone. Now, I initially got the wanted behavior using R+ and reduced to negligible levels unwanted behavior using extinction, BUT, for the dog's and the handler's sakes, at the end I add that final, and VERY extreme use of punishment, to be absolutely sure that the animal will not, under any circumstance, commit error.

When I punished, the maximum I EVER had to apply it was three times, and usually once was enough. Mind you, out of the many thousands of animals we have trained during many hundreds of training programs, we punished about a dozen times. The fact that I would do this at all disqualifies me from the ranks of what people call "clicker trainers." I don't mind. I have a clear conscience. I believe I did the best for animal and human kind.

You will not find anywhere in our workshop manuals, or in our posts, detailed instructions on punishment. I believe punishment is greatly overused. When it is applied, I think it is misapplied (strangely, my second-most objection usually is that it is not severe enough. Of course, my first-most objection is that it was used at all in the first place. I said it sounds strange). I won't instruct people in punishment techniques because I do not want to be a party to misapplication of a perfectly good technique. If someone misapplied positive reinforcement, the animal is likely not to be directly hurt too much. The same can't be said for punishment.

In nature, animals face aversive situations frequently, possibly more often than they have the opportunity for reinforcement. It is often pointed out that animals learn quickly to avoid aversive stimuli, and that aversives tend to generalize more than reinforcements - likely true.

But, having said that, and even accepting the likelihood that it is true, animals do come equipped with a certain level of built-in resistance to the influence of aversive stimuli. A pup that is punished for nipping at a momma's nipple too hard does not abandon the nipple. The pup might be a bit tentative while nursing for a bit, but gets over it, and there is usually less nipping at the teat that feeds it. I watched a coyote pup fall into a cold spring and come out whining and shivering. However, the pup was soon back trying to catch the minnows that attracted it there in the first place.

Over a wild animal's life, it meets lots of aversives in the process of trying to earn a living and find a mate, and do all of the other things that that wild animal normally do. Trainers did not invent aversives. It just so happens that when we, or another organism, applies an aversive, we call it punishment.

There are a lot of philosophical issues that arise when we address punishment. When one of those coyotes finally caught a fish in that spring, it is likely that the fish felt some pain before being ingested by the coyote. Would we call the coyote's crunching the fish a punishment? What if the coyote could "suck up" the fish, thus swallow it whole. Well, I would think that being digested slowly would not be much of an improvement over being crushed.

Some might say the coyote was not trying to change the fish's behavior, or attempt to "penalize" the fish, so it, the aversive, is not a punishment. This would imply that the intent of the applied aversive is the difference between an aversive and a punishment.

Perhaps more significantly, some might say that humans have ethics and should never apply punishment to any organism. Well, there we have a problem; our society, and every society, has those member humans who are lawbreakers. The lawless members of society often do not respond appropriately to what the rest of us call reinforcement; we are left with punishment.

We can get into the ethics of our penal system, the fairness of it all, but that is not my point. What is my point is that aversives, and punishment, plays a role both in nature and in our society. If, as a society, we are bent on the punishment of our own kind, why the ethical demand to eliminate punishing animals? There is not a lot said about punishment in nature. There is quite a debate about punishment within our society. Gradually, over many centuries, our group perception of the role of punishment of animals and humans has evolved to where we are today. As a society, we are still debating the issues, and likely we will continue debating for centuries to come. As long as we continue open debate, we are likely to arrive at a consensus that we all can live with.

As animal trainers, whether professional or amateur, we seek to change animal behavior. There are those in our culture who believe that animal ownership, and especially ownership that involves confinement and the changing of behavior, is tantamount to punishment. Some of you must be aware of the arguments of Peter Singer, and others with similar beliefs; most of this group are not so much for open debate, just change. While most of us may believe that Singer's position is extreme, the question may be asked, at what point in the continuum of moderate to extreme is animal ownership OK? When are we allowed to confine or to punish animals? For that matter, what are the rights of a society to confine and punish its citizenry. These are ethical issues that societies around the world are facing.

Again, as animal trainers, we face a microcosm of the general issues of the definition and ethics of punishment. Should we punish, or should we not? If we say absolutely that there should be no punishment, there remains the issue of what is punishing. I suggest that we should be very careful about stating that there should be an absolute ban on punishment of any kind and under all circumstances. Be very careful about what you ask for; you just might get it.

My personal view is that virtually all animal training would profit from the use of positive reinforcement. There are a very few circumstances in animal training where the addition of punishment is, in my opinion, extremely worthwhile, and possibly essential. There are a few more situations where adding punishment would likely be very useful. In those rare circumstances where punishment offers potential benefit, it is always to stop some behavior that could cause harm to the animal, a human, or damage or destroy property. In my opinion, and just as examples, this would exclude the use of positive punishment in the training of sport or obedience training and severely restrict punishment in the training of service dogs.

This sometimes philosophical and broad brush discussion of punishment is my attempt to point out that there could be political fallout from a society's decision to totally ban punishment. Further, I suggest that we follow our own creed of "think positive," and spend less time attacking those who believe that they have the right and justification to punish. We will win the war of ideas if we prove our point with training successes. We may lose the right for future trainers to train at all if we convince society that it should condemn, and disallow, what we call traditional trainers and training methods.

Forgive the lengthy discourse. While I wrote this in a hurry, I have given the issue considerable thought. As you might guess, Marian and I discussed these, and many more related issues, at great length. We recognized that these are emotional issues. What we usually concluded was that most trainers, including ourselves, are political babes-in-the-woods. If we try to convert our gut feelings to action, whether within professional organizations or in the political arena, we could hurt ourselves in the long run. We concluded that time and technology is on our side, but only if trainers learn how to correctly apply behavioral methods and apply them ethically. SO- go out and train, folks, and let the opposition alone. To fight them is to, in the long run, fight yourself and your profession.

Bob Bailey
behavior@hsnp.com
copyright 2002 Robert Bailey

 

| Training Articles Contents || Site Home |


Copyright of all posts is the property of the original author. Please obtain permission from the original author before copying, quoting, or forwarding.

List and Site Owner: Melissa Alexander, mca @ clickersolutions.com