Friday, January 10, 2014

The Problem of Self-reinforcing Behavior


A little more of this, a little less of that.

That's what training is all about no matter what type of beast is being trained.

Most people know how to reward positive behavior, and to be honest it's not that tough to train a small child or a new puppy.

But what about bad behavior that is strongly self-reinforcing, and may have gone on for years without any opposition at all?

"Train a different behavior" we are told, but of course those are just words. The answer to a barking dog is not to "train it to bark on command and then never give the command," no matter how often that bit of nonsense is repeated.

For dogs, barking is a self-rewarding behavior, same as deer-chasing, and digging in trash cans for stray scraps of food.

But, of course, not all training is about dogs and small children is it?

We also train adult people, organizations, corporations, and even clients.

And is all that training done using a "click and treat" paradigm?

Of course not.

Nor could it ever be.

For example, what are we supposed to be doing with criminals who rob our houses and shoplift in our stores?

Are we to reward them every time they decide not to rob us?

Are we supposed to buy a meal and a movie ticket for a rapist every time they allow a pretty girl to walk by unmolested?

I think not!

And what about corporations that engage in illegal theft, profiteering, and bill-padding, or which manufacture products that maim, poison or kill?

Are we supposed to give tax breaks to all the companies that don't kill us and don't poison us and don't rip us off?
So what's the solution? 

The answer, of course, is that common sense lies somewhere between "shoot the dog" and "click and treat" when it comes to self-rewarding  behavior.

What's needed is solid consistency and excellent timing so that the right bad actors get the message that we need "a little less of that" early on in life. 

With a puppy or small child, it does not take much to send a strong "no" signal.  But if the behavior is long-established and very self-rewarding, then the initial signal may need to be a good deal stronger.  Nothing brutal, of course; timing and consistency are always more important than severity.  That said, the "no" signal has to be very real, and one that cannot be shrugged off as "the cost of doing business."

In short, there is a place for cheese and choke chains and there always has been and there always will be.

It's not a matter of one or the other -- it's a question of using the right tool on the right problem at the right time.  
 .

No comments:

Post a Comment