How Dogs Learn Classical vs. Operant Conditioning

How Dogs Learn: Classical vs. Operant Conditioning

By LukEli K-9, Dog Training and Behavior Modification Professionals

The first book they had us read at Highland Canine Training’s school for dog trainers was Excel-erated Learning by Pamela J. Reid. We had to read a book every week or so and after tackling this book I thought my brain might explode. It is, however, chock full of insight into how animals learn, and I’ve borrowed heavily from it today to do my best to begin to explain how B.F. Skinner caused dog trainers-in-training everywhere to curse his name as they struggle to grasp the concept: operant conditioning. Before I can tell you about operant conditioning, I should explain another type of learning to you - classical conditioning.

CLASSICAL CONDITIONING

Classical conditioning, Pavlovian conditioning, associative learning - these are all terms for what happens when an animal learns associations among things. Learning associations means learning that things go together: when one thing happens (you burn your dinner), another thing will follow shortly (the smoke detector will go off). You say “Let’s go for a walk”, and your dog jumps up excitedly because he has learned that this particular phrase always signals going outside and chasing squirrels. Your cat runs to the kitchen every time it hears the can opener because it has learned that this sound precedes feeding time. In each case there is a predictable relationship among the events and the animal learns to respond to the first event in anticipation of the second. That’s what classical conditioning is all about: anticipation.

You’ve probably heard of Pavlov’s bell. Ivan Pavlov was interested in how reflexes work and set out to study the salivary reflex in dogs. In his experiments, he would ring a bell before presenting a dog with food. The dog eventually began to salivate at the sound of the bell, even though no food was present, because it had learned that the bell’s ring always preceded feeding time. Pavlov’s experiments embody the basics of associative learning; an initially neutral or meaningless stimulus (the bell) is presented prior to the delivery of an unconditioned stimulus (the food). It’s called an unconditioned stimulus because presentation of the UCS (food) always elicits an unconditioned response(UCR)(salivating). For instance: Food always elicits salivation. The dog doesn’t have to learn to salivate, it just happens. Once the animal has learned the association between the neutral stimulus (bell) and the unconditioned stimulus (food), the neutral stimulus (bell) itself elicits a response which is called the conditioned response (CR).

Keep in mind there is nothing that requires the animal to perform the conditioned response. The dog doesn’t need to salivate when the bell sounds. The food is delivered no matter what the animal does. The CR (salivating) just happens; it’s involuntary. When you attempt to classically condition your dog, you are trying to get them to have an automatic, involuntary reaction to a stimulus.

OPERANT CONDITIONING

Operant conditioning, Skinnerian conditioning, instrumental learning - these are all terms for what happens when an animal learns that it’s behavior has consequences. It’s called operant conditioning because the behavior operates on, or has an effect on, the animal’s world. This applies to us humans as well. We do all sorts of things every day that lead to consequences that wouldn’t happen if we hadn’t done something. If you don’t work, you don’t get paid. If you don’t cook, you don’t eat. If I ask my dog to sit and he doesn’t sit, he doesn’t get a treat. Animals learn the relationships between the things they do and the consequences of those actions just as people do.

A popular way to remember the formula for the process of operant conditioning is to think of it as the ABC’s of learning:

Antecedent(a cue like sit) --> Behavior (the animal sits) --> Consequence (you give them a treat)

Operant conditioning is very different than classical conditioning in that the animal must perform a behavior in order for the stimulus to be presented. Aside from the antecedent, the events that occur are under the animals control.

In the operant sequence there are four possible scenarios: two of which increase the likelihood that a behavior will occur again and two that decrease the likelihood of the behavior reoccurring. All learning also involves stress, and each of these induce a differing amount of stress while learning.

POSITIVE REINFORCEMENT

Positive reinforcement (P+) involves the presentation of a good consequence when the behavior is performed. For instance, you say “sit”, your dog sits, and you give him a treat. This serves to increase the likelihood that your dog will perform the behavior again. Human behavior is also regulated by positive reinforcement. Child does homework, child is awarded a good grade. Child does chores, child gets an allowance. Person works hard for several months, person goes on vacation. You get the idea. Much of dog training is based on ideas that incorporate R+. Quick learning takes place with low stress when using this quadrant for learning.

NEGATIVE REINFORCEMENT

Negative reinforcement (R-) involves the removal of a bad consequence when the behavior is performed. As with R+, the reinforcer must be contingent on the behavior; you cease “prodding” once the behavior has been performed. For example, you apply pressure with a choke collar and say “sit”, your dog sits, and you release the pressure from the choke collar, removing the unwanted action. This also serves to increase the likelihood of the response in the future. Negative reinforcement is a very common method of controlling behavior in people as well. Child does homework to avoid mom’s nagging. Person goes to work to earn money to avoid being homeless. You say uncle to get your brother to stop twisting your arm. You drive the speed limit to avoid tickets and fines. Some examples for dogs: Applying pressure on the rear end of the dog while saying sit until they sit and then you stop pushing down. “Escape training” such as pinching the dog’s ear until they open their mouth to receive a dumbbell for forced fetch retrieving is an often used R- method. Behavior Adjustment Training utilizes this quadrant heavily for rehabilitating aggressive dogs by applying a negative (whatever stimulus causes the dog to be reactive, i. e. another dog) and then taking the dog away when they’ve given the correct response (like a look away or paw raise) to the aversive. Negative reinforcement can be a powerful training tool when utilized correctly and humanely. When used inhumanely, R- can seem like nothing less than torture. Negative reinforcement is the fastest way to learn and is also the most stressful.

POSITIVE PUNISHMENT

Positive punishment (P+) involves the presentation of a bad consequence when the response is performed. For instance, you say “sit”, your dog lies down instead, you jerk him back on to his feet with the leash. This serves to decrease the likelihood of the behavior in the future. Positive Punishment happens a lot: child hits brother, father spanks child. Person drives drunk, person ends up in jail. Puppy pees on floor, puppy gets hit with a rolled up newspaper. Dog continues to bark after being cued to stop, dog gets squirted in the face with water to startle it. I’m pretty sure we’re all fairly familiar with P+, though we may not have known that’s what it was. This quadrant is probably the most widely used. Animals learn pretty quickly when using this quadrant, but it’s also pretty stressful.

NEGATIVE PUNISHMENT

Negative punishment (P-) involves the removal of a bad consequence when the response is performed. For instance, you say “sit”, your dog lies down, so you eat the treat instead of giving it to your dog. This serves to decrease the likelihood of the response in the future. Child calls friend a booger head, friend goes home and won’t play anymore. Person drives drunk, person loses license. Dog plays too rough, you stop playing with them. Puppy jumps up on you, you turn your back to puppy. Animals learn fairly quickly this way, and the stress level is pretty low.

All four of these scenarios will cause your dog to learn. You must choose which combination of techniques you feel is right for you. This knowledge will aid you when searching for a dog trainer as well. Remember not to hesitate to ask questions about a trainers methods so you can decide if that program is right for you. There are many books on the subject a well should you decide to read more about learning theory and behaviorism.

Contact a trainer at LukEli K-9 today to learn more about the methods we utilize and how we can help you reach your training goals today!