Thank you! Have a Chocolate: Operant Conditioning at Work

Any psychology professor who played The Office clip about classical conditioning probably shared this scene from The Big Bang Theory:

While Jim conditions Dwight to want a mint when he hears a computer chime (learning by association), Sheldon reinforces Penny's behavior choices by giving her a chocolate (learning by consequences).

There are four possible consequences in operant conditioning: positive reinforcement, negative reinforcement, positive punishment, and negative punishment. Two of these descriptions probably sound redundant to you while the other two may seem contradictory. How can punishment be positive? How can reinforcement be negative?

In most contexts we equate the word positive with good and the word negative with bad. But in operant conditioning we think in mathematical terms: positive means adding something: Penny chose to take her phone call outside, so Sheldon gave her a chocolate. Negative, on the other hand, means subtracting something: Sheldon might move the chocolate away from Penny if she remained on the couch chatting on her phone.

The other two key terms to understand are reinforcement and punishment. Reinforcement increases the likelihood that a behavior will happen again. Punishment decreases the likelihood that a behavior will occur again. We also have to remember that reinforcement and punishment are ALWAYS from the learner's perspective. If Penny didn't like chocolate, she wouldn't find it reinforcing.

Positive reinforcement is the most familiar section of the grid and the one that I use when working with dogs and their people. Treats, toys and the opportunity to play, attention, and praise can all be reinforcing depending on what the dog likes best--some dogs find cheese more reinforcing than a dry liver treat while others live for a quick game of tug. I'll dive deeper into positive reinforcement next month.

With positive punishment you add something to discourage a behavior. Punishments can range from tools like shock collars and choke chains to spray bottles and raised voices. Just as with reinforcement, the learner decides what is punishing. Some dogs balk at a raised voice whereas others seem not to notice a choke chain tightening around their necks. Next month I'll explain more about why I choose not to use punishment in training.

In the case of negative reinforcement, you remove something aversive or unpleasant in order to encourage a behavior. We often apply negative reinforcement after positive punishment. For example, a dog wearing a prong collar pulls on leash and the collar pinches her neck (positive punishment). When she stops pulling (the goal behavior), the pinching stops. Eventually a dog may stop pulling in order to avoid the pain, but she won't learn to walk well on a leash without the tool.

Sometimes we see negative punishment at work with parents of small children who stand by while the child has a temper tantrum in public. Withholding attention is a form of negative punishment that you may also see people use with dogs who jump on them. Instead of pushing the dog down (which reinforces the behavior with your attention), you ignore him. Negative punishment seems benign, but it can lead to stress and frustration in a dog (and a child). They may try harder to get your attention or they may shut down.

While operant and classical conditioning are not the only tools in a trainer's kit, they provide a foundation for understanding how learning works in all kinds of animals, including our dogs. So discover what your dog finds reinforcing and use that for teaching her new skills. Just don't give her any chocolate!