Reinforcing what you don’t want: Elegant solutions for tough problems

The suggestion of reinforcing an unwanted behavior often shocks people. Nevertheless, it’s a very useful tool to solve complex problems. That is because it is the practical way to get the dog to recognize the behavior as a unique behavior. Meaning your first goal is often to make the dog recognize the sensations that accompany a behavior. This is critical for behaviors like house training. The dog must recognize the subtle sensation of a full bladder as meaning something – just like an infant in diapers. Unless the animal knows that those sensations are connected to specific consequences it will never acquire behaviors that solve the problem. i.e. You can’t know that it’s beneficial to go outside to use the bathroom if you don’t know that a full bowel has anything to do with it.

For practical purposes, positive reinforcement gives you more options for making these very low-level associations. i.e. Punishment isn’t going to help inhibit a behavior unless the animal can identify the behavior as a unique action or set of actions. For instance, if you punish a dog at the instant he lifts his leg on a chair, he’s as likely to think it was the chair and switch to the drapes. He may think it’s the location as much as the act of scent marking. Worse, he may thing it is your presence that is the key. Here is why positive reinforcement for inappropriate behavior makes more sense – you don’t trigger avoidance. Cops at a stakeout don’t want to tip off the bad guys. Undercover cops often “buy” drugs several times before they nail the bad guys. If a dog is confused about what caused the punishment, he may continue the behavior secretly to avoid detection. This is the common response. When someone tells me that “I never see him pee in the house” I know the dog has been punished for inappropriate elimination. After the punishment, the dog’s senses will be finely tuned to avoid anything or anyone that is connected with the event. The bottom line is that you may never see the behavior again. You can’t strengthen what you can’t observe directly or indirectly.(Not completely true, but for the purposes of this paragra That means any punishment you attempt to apply will be hopelessly disconnected from the behavior. Disconnected punishment tends to trigger something akin to terror. The animal senses an aversive event but has no immediate way to avoid it or terminate it.

So, even if I want to inhibit the behavior, later, I may choose to start the process by positively reinforcing it. Yes. It does sound nuts. However, once the acceptable behavior is established it is easy to eliminate it in other contexts. i.e. Teach the dog to lift his leg outdoors in your presence. Strengthen it and add a cue. Then you can apply punishment indoors and the dog will flip to going outside to get the exact opposite response – treats for correct behavior. Without the initial positive reinforcement you are dead in the water.

One thought on “Reinforcing what you don’t want: Elegant solutions for tough problems

  1. Thanks, for all of the knowledge you freely pass on. It’s appreciated.

    Not to oversimplify this concept, but once a pup is solid in eliminating/leg lifting on cue in the appropriate place, a punishment (“NO” and an interruption or something along those lines) would be appropriate to then eliminate the unacceptable behavior if/when it occurs, correct?

    If that’s correct, then the most important (or one of the most) goal in house training a pup would be to keep a strict schedule to allow reinforcement of the behavior in the appropriate location, and trying to prevent the opportunities/necessities indoors with this same schedule. I hope I’m comprehending the idea.

Leave a Reply

Your email address will not be published. Required fields are marked *