Using Compulsion in Training: A Primer

Excerpeted from: Controlled Aggression in Theory & Practice

by Jerry Bradshaw

Every behavior must have a consequence, either a reinforcing consequence or a punishing consequence. “Punishment” here refers to punishment in the behavioral sense. Punishment is any consequence of a behavior that reduces the likelihood of that behavior. Reinforcement is any consequence of a behavior that increases the likelihood of that behavior. Both reinforcement and punishment come in two varieties, positive and negative: Positive in the sense of providing a consequence, and negative in the sense of withholding a consequence. This gives us the following four consequences:

·         Positive punishment means delivering an undesirable consequence that reduces the likelihood of a given behavior.

·         Negative punishment means withholding a desirable consequence to reduce the likelihood of a given behavior.

·         Positive reinforcement means providing a desirable consequence to increase the likelihood of a behavior.

·         Negative reinforcement means removing an undesirable consequence to increase the likelihood of a behavior. 

Training proceeds through three phases: acquisition, fluency, and generalization. Training is a process through which we introduce and vary the consequences of behavior to teach the dog associations (acquisition), and then teach limits to behavior when employing these associations (fluency). Finally, training involves requiring these associations to be performed in all different situations and contexts (generalization). This is commonly referred to as proceeding through the learning stage, to the correction stage, and then the proofing stage.

Note that I do not usually find any use for force or compulsion in the acquisition phase of training. Providing a foundation of proper behavior, and never allowing behaviors you don’t want to see, can be accomplished in this phase of training, because the trainer can exercise complete control over the training environment. In other words, we can set the dog up for success. However,  in order to proof responses, we must present the dog with situations that may induce non-compliance, anticipate the dog’s non-compliance, and positively punish him (corrections) for disobedience.  In my opinion, it is virtually impossible to de-condition a dog to every possible distraction using only negative punishment. Limits must be set in a general way with compulsion, and this is done in the fluency and generalization phases of training. In the fluency stage we introduce the concept of corrections or positive punishment, and continue to apply those corrections (as well as apply positive reinforcement) in the generalization or proofing stage of training.

Force or compulsion is applied to the dog through the use of devices intended to cause discomfort to the dog. This discomfort should be minimal, and just enough to get the point across. This minimum force rule should always be applied when using compulsion. At the end of this book there is a guide to introducing the e-collar in obedience, with some applications to protection training. Most people who criticize e-collars know little about their proper use.  Those trainers, who employ their use properly, know how the relationship with their partners grows tremendously through these described effects.  The trainer is the center of all good things, and offers relief from any punishment that is applied, through the use of positive positive reinforcement. Thus we concentrate on rewarding good behavior that is shown, once limits are set. Correction becomes unemotional when the physical component of it is removed.

Tarheel Canine will be providing a trained explosives dog to the Baltimore City Police Department on December 14 2009. The funds for this purchase were graciously provided by the Quarterback for the Pittsburgh Steelers, Ben Roethlisberger through his charitable foundation.

It just so happens that I have been a fan of the Steelers since first getting cable TV at my childhood home in Monticello, NY which was just an hour or so away from the NY/PA state line. We were fortunate to get the cable station out of Scranton PA, and so I began watching the Steelers every sunday. It is a huge thrill to be able to provide ths dog to Baltimore City with a grant from the quarterback of my favorite team!

Visit the foundation website for more information on grants for police dogs through the Roethlisberger foudation:

http://www.bigben7.com/foundation.aspx

Jerry Bradshaw, President, Tarheel Canine

Question: I just certified my new dog, he is the third dog I have worked in my 16 years as a handler. The dog is a 3 year old imported German Sheperd that was sold to us a titled dog. I really do not know what he was titled in but he obviously had prior training. Overall he is doing well but he has an annoying trait I cannot seem to correct. When he heals he wraps himself around my leg and looks up at me, to the point it is hard to walk without tripping over him. I know in some sport dog circles this is desired, but I think a police dog should heel next to the handler paying attention to what’s in front of him. I have had him 7 months and have not been able to break this. Any ideas?

 Answer: This is one of those behaviors that is heavily ingrained in an IPO or Schutzhund dog, and is called an attention or focused heel and is part of the rules of that sport. Police officers don’t tend to like it because of the crowding, or because they are not used to it. I submit to you, however, that in many circumstances you can probably work around distractions much better than many of your counterparts who don’t have an attention heel. Example: You arrive on a scene that has a number of people making loud gestures and screaming and crying. Having an attention heel, when your dog gets out of the car you can keep him from loading on these people. They may be loud and seemingly aggressive, although non threatening, but you can focus him on you and keep him from alerting unnecessarily. In my estimation it is a very handy behavior. Since you have your dog’s attention, if you want him to focus outwardly, all you really have to do is hold his line or collar, and give his alert command, right?

Basically you need to reward the behavior you want, which is positional correctness (by your leg) but looking forward, and reward him when he looks forward (variably) by sending him to bite. Don’t reward him for looking at you ever again with praise or touching, or looking in his eyes. If you really want to get rid of the behavior, here is what you can do:

Question: I think I have created a monster! I have a Dutch Shepherd single purpose narc dog that I have worked for the last 3 years. One of the things I liked about him in the selection test was how possessive he was with his toy. Now it has gotten to be a major battle, there are days he simply will not give up his toy and the problem is getting worse. I have tried the two toy system but it does not work, once he has one toy, he does not want to give it up for a new one. Do you have a good way to get him to release the toy to me on command?

Answer: You need to treat the situation like training a patrol dog to out. Just remember, though, he is not a patrol dog for a reason. He may not be as hard as you think. First, do this when you have a few days when you are NOT going to have to reward him for his drug finds. Take him someplace he never normally has to work, because you are going to use some force to accomplish the task, and you don’t want him to associate the compulsion with his work.  This is critical.

Back tie him to a sturdy fence post (better than a tree as he can’t wind himself up) on a flat collar or harness with about a 10 foot long line, and put a second line attached to your 6’ leash on a prong collar. (Yes, I said prong collar. If you have been conditioned to think this is only used as a last resort, you may not be inclined to use it, but contrary to what you have learned, it is safe, effective, and used by the rest of the world over choke collars almost universally when training working dogs. It will work for this exercise much better than a choke collar).

Use an object different from his normal drug reward, and a little less enticing, like PVC pipe. Have 2 of the same toy, and give the dog one of them, and let him hold it in his mouth for a few seconds. When you are ready to make him let go, say, “Out,” or whatever your new out command will be. Don’t use the command he has been ignoring for the last 3 years, pick a new one. As soon as the “t” in out is out of your mouth, correct him into your body fairly hard until he releases it. As soon as he drops it, give him the other toy, and let him enjoy it for a few seconds, then repeat until the dog is releasing quickly. If you do this right, he will let go after a few repetitions on the command. In behavioral terms, we are punishing his holding of the toy beyond hearing the out command, and then rewarding the act of him releasing with another toy. 

When he does this well on the back tie, introduce his drug toy in the back-tie context, and do the above exercise until he is releasing it on command. Then, when you are getting a good response on the back-tie on his normal reward toy, after a few short sessions (3-6 reps and then put the dog away, and do multiple sessions a day), put 2 lines on him as before, but do not back tie him. Use the 2 lines just as before, but the handler holds the correction line and another trainer holds the back line, and acts as a moveable back tie. Throw the dog’s toy for him, and then put him between the 2 lines, and out him as before, correcting if he doesn’t release against the posted back line. Then, reward the release with the second toy. Repeat in different contexts. As this goes well, eliminate the back line, and use only the correction line.

When the dog is doing well, bring the dog to a drug training session, and set up an easy hide, and we will test the training. Put the correction collar on the dog with a short tab line on it. Do not run the dog to search on the prong. When he hits the hide, reward him with the toy and play with him. Move him away from the search area, well away, as you are playing with him. Then grab the correction line and tell him out, and if he does, let him grab the toy again and let him play. If he doesn’t, correct as usual, and then after a successful out (even if you had to force it) let him get the toy again. If he still fights you over the toy on this test, go back to correcting away from the drug searching until you are again confident he will release on command, and put it together again in another test.

As the dog releases for you consistently, and the behavior becomes conditioned to release on command, you can remove the correction collar altogether. If he regresses, put the collar back on. Don’t make him collar wise. Put it on and off randomly at times other than your training sessions for short times when he is supervised, but don’t hook anything to it. The collar is only an aid in conditioning the behavior, so don’t remove it too quickly before the behavior is well conditioned. You want to keep the option to correct so that the dog knows he cannot get away with ignoring this command.

In doing this exercise, you should use the minimum force to get the job done, and be careful of making any unwanted associations since you are using some force. You should do this under the eye of an experienced trainer until you are comfortable. The hardest part is gauging the correction to stimulate the out without overcorrecting. You may be surprised how easy this is if you carry through it. If you need more guidance before giving it a try, you can e-mail me at Malinois_jb@mindspring.com.