Wednesday, June 16, 2010

Punishment vs. Correction

In the course of a recent discussion about dogs pulling on leash (and what role corrections might play) a colleague queried me about the following:
“I change directions frequently, especially if I sense she's about to course ahead. If she moves to the end of the leash and there is pressure, I'll get her attention with a sound, and signal her to follow my new direction, and reinforce her for doing so. Of course any pressure is released for her doing so, but I did not deliver it. I keep my hands as still as stones. Now in your opinion, are there any leash corrections involved here?”

Without a doubt there are corrections in that example. The subject of the discussion is “learning theory,” not teaching theory. The intervention, as experienced by the dog, was effective and therefore “corrective.” Learning theory does not address who supplies the stimulus (pos. or neg.) only what its effect was on the subject. If, either by pulling the leash, simply stopping or changing direction, the dog experienced a stimulus that made him stop what he was doing, the effect was the same and the handler caused the correction to happen. The dog is not stupid, he knows from whence the correction came. If the correction was effective, the next thing the dog did was to look at you and move into a position beside you – he knows. He doesn’t stare off into the sky and wonder, “Lord was that you? Why me lord?”

If we were to all gather together in a large park and began training our dog, many of you would see that what I do probably isn’t a whole lot different than what you do – even those of you who say you use only “positive methods.” I am not a punishment based trainer, yet I do correct my dogs. The correction can be anything including redirection, removing a reward or applying an unpleasant (physical) consequence.

The first, (and most obvious) problem is deciding what the words themselves mean. Secondly, (and most important) we must begin to appreciate how the average layperson understands those words. JQP is seldom a scientist and yet these words have been part of the common vocabulary since long before there were “behaviorists” attempting to give new meanings to everyday words.

Terms such as “positive punishment,” “negative punishment,” “aversive” etc. for a variety of reasons, have all become a part of the ‘behaviorspeak’ movement. In the laboratories where they were studying behavior, people like B. F. Skinner began using many common words (such as the terms already mentioned) to describe their observations, philosophies, actions and beliefs. It likely was not their intent to change the meanings of many of these words but what they ended up doing was to add a somewhat different meaning to some of them. The problem comes when we don’t all sing from the same book (nor should we feel we must).

The meanings given these words by behaviorists are not what I (as a part of JQP) mean when I use them and herein lies the problem. Take the word “punish,” the Oxford dictionary definition is: “Cause (offender) to suffer for offence; chastise; inflict penalty on (offender); inflict penalty for (offence); (colloq.) inflict severe blows on (opponent)….” Punishment is defined as “punishing penalty.” Nowhere in the definition is there any mention that a punishment is to correct or change behavior; it is simply the penalty handed out, after the fact, for unacceptable behavior. These definitions are in line with what most people think when they use the word.

The word “Correct” does have as one of its meanings, to punish but it is with the intent of making right. Other meanings are: Set right; substitute right thing (for wrong one); counteract; bring into accordance with standard; eliminate aberration etc.

The “behaviorspeak” definition of punishment is any action that is used to stop a behavior or reduce its occurrence (and the definition is based on the outcome). According to those that subscribe to this definition, corrections would be included under the heading of punishment. When they talk in terms of positive or negative punishment, they mean to add something to punish (positive = to add) or taking something away to punish (negative = take away). This gets further complicated by the use of the term “aversive” by which they usually mean the “punisher” which was added to make the behavior stop.

The interesting thing is that some of these “opposed to punishment” types, who would label any “aversive correction” as punishment, have a different standard for themselves. A few years ago, I was at a seminar being put on by a rather well known PROC (Positive Reinforcement Operant Conditioning) trainer. He decried the use of any form of “punishment.” One of the participants wanted help with her dog that pulled on leash. He took the dog and began to walk it. As soon as the dog charged away from his side, he stopped, dropped all slack and held on to the handle tightly. The dog hit the end of the line rather hard (I’d be hard pressed to leash correct any more effectively). This was repeated two more times and then the dog “miraculously” began to walk beside him. Someone asked him about his use of punishment and he denied it saying what he had shown us was not punishment but rather the dog correcting himself. (Note: there is suddenly a distinction being made between punishment and correction?)

The aversion some trainers feel (to acknowledge an action as a correction) is an interesting phenomenon and it speaks to the person’s personal issues more than it does to dog training philosophy. Two possible reasons come to mind (there may be more) why someone would be reluctant to acknowledge a correction:
1. Economic – they want their training philosophy to seem politically correct so that clients will want to train with them.
2. Emotional – they may feel caught on the horns of their own “moral dilemma” and want to assuage their feelings of guilt.

The fact that JQP does make a distinction between correction and punishment is important. Those in the “positive behaviorspeak” camp know and exploit the fact that there is such a strong emotional reaction to the term “punishment” (as evidenced by this very discussion and many more just like it). If they can convince potential clients that they are “punishing” their dog (rather than correcting him) they can then make the argument to both the new client and to themselves, that theirs is the more “humane” way to train.

So to get back to that gathering in the park, I would expect to see a variety of techniques with some very good training going on. Pushing aside the different philosophies and terms others might use, I would expect to see an interaction between each dog and handler that connected consequences to behaviors. Beyond this I’d also expect to see (especially with the really good trainers) a level of communication that transcends anything the theorists could have imagined.

Tsuro Dog Training
Visit us at

No comments:

Post a Comment