Ever wondered how trainers teach their dogs to play chess, open the fridge, or even walk blindfolded down a tightrope? What may seem to be the result of magic or just getting lucky with a smart dog, comes down to the techniques used by the trainer to properly shape the behavior.

What Is Shaping

The textbook definition of shaping is "reinforcing successive approximations of a behavior," but what does that actually mean?

Well, it means getting closer and closer to a desired behavior and providing reinforcement every time you get a little bit closer.

Let's use the example of the dog playing chess to illustrate what shaping is and how it can be used to teach behaviors.

1. Look at the Peon (chess piece). Whenever the dog looks at the chess piece, he is reinforced with food, causing the frequency of that behavior to increase. In other words, the dog will look at the chess piece more often.

2. Touch the Peon. After the dog understands that he needs to look at the chess piece in order to receive reinforcement, he then only receives reinforcement if he touches the piece with his nose or mouth.

3. Bite the Peon. Reinforcement is then withheld until the dog bites the chess piece.

4. Pick up the Peon. Similar to biting the chess piece, the dog transitions from just biting, to biting and picking up the chess piece before he can receive a treat.

5. Drop peon in box. Once the peon has been picked up, the trainer slides a box under the dog and provides reinforcement when the dog drops the chess piece in the box. Dropping the piece outside of the box does not provide reinforcement.

And I'll let you watch the video for the rest of the steps.

The "successive approximations" are the small changes in the behavior that are required to receive the reinforcement. After the dog is able to touch the chess piece, he is no longer rewarded for simply looking at it. After he has learned to pick it up, he is no longer rewarded for just touching it. As the behavior evolves, so does the contingency for reinforcement, making it more likely that the dog will perform the more recent tasks. The speed at which you evolve the behavior is up to you and your dog.

Why Shaping Works

The key to behavior change in shaping is the small steps that are taken during the shaping process. You'll notice that the trainer did not jump from the dog looking at the chess piece, straight to dropping it in the box. Without the steps in between, the dog would never connect the dots between those actions and understand what you expect of them.

Karen Pryor, a famous animal behaviorist and author of the book "Don't Shoot the Dog," coined 10 rules of shaping that need to be followed in order for your shaping procedure to be effective. Her shaping rules, now accepted by many animal behaviorists are as follows:

1. Raise criteria in increments small enough so that the subject always has a realistic chance of reinforcement. This is what we discussed above. The animal needs to connect the behaviors together if they are to perform the way that you intend.

   
   

2. Train one aspect of any particular behavior at a time. Don’t try to shape for two criteria simultaneously. If the dog is working on a down-stay, shape the down command first, before attempting the stay. If you attempt to shape both criteria at the same time, the animal may become confused and not offer either of the behaviors.

   
   

3. During shaping, put the current level of response on a variable ratio schedule of reinforcement before adding or raising the criteria. All this means is that if the dog has understood a particular step in your shaping process, stop offering rewards every time they perform that step. Instead, only reinforce that behavior variably, or occasionally, before moving on to the next step in the shaping process.

   
   

4. When introducing a new criterion, or aspect of the behavioral skill, temporarily relax the old ones. Similar to step 3, if moving on to the next step of a shaping process, Don't maintain high standards for the previous part of the behavior unless it directly coincides and is essential for the next behavior.

   
   

5. Stay ahead of your subject: Plan your shaping program completely so that if the subject makes sudden progress, you are aware of what to reinforce next. If your dog begins picking up the peon early, have the box ready so that you can move onto to the next step.

   
   

6. Don’t change trainers in midstream. You can have several trainers per trainee, but stick to one shaper per behavior. Stay as consistent as possible. The way that you say sit and the way that your brother says sit are different. If the dog is learning a new behavior, don't switch things up on them.

   
   

7. If one shaping procedure is not eliciting progress, find another. There are as many ways to get behavior as there are trainers to think them up. This concept is fairly self explanatory, but don't get bogged down on one training process if it doesn't work. Think of a different route to the same goal.

   
   

8. Don’t interrupt a training session gratuitously; that constitutes a punishment. If you stop the flow of reinforcement, especially if the dog is still performing the behavior, then the frequency of that behavior will decrease and you could even accidentally put that behavior into extinction. This could make the process of shaping that behavior more difficult in the future.

   
   

9. If behavior deteriorates, “Go back to kindergarten.” Quickly review the whole shaping process with a series of easily earned reinforcers. Going back to kindergarten simply means to make the process easier. Go back to an earlier stage in the shaping process, get the dog to have some successful behaviors to earn rewards and begin making progress from there. You will find that although you had to back track, you may make progress faster then if you would have continued to try pushing through the struggle at the harder stage.

   
   

10. End each session on a high note, if possible, but in any case quit while you’re ahead. Ending the training session with reinforcement, maybe even a jackpot, will help your next training session start off on a good note. This will also reduce frustration in the dog.

These rules are designed to help us effectively accomplish our shaping goals without unnecessary struggles. They help us slow down, and really understand the training that we are doing so that we can effectively train and make changes as needed. Thinking about rule 7, there are some things that might look good on paper, that don't actually work in practice. Knowing this helps us to be flexible and adjust our training plan as needed, instead of getting bogged down and blaming the trainee.

How To Use Shaping In Practice

Over the past couple of weeks, I have been attempting to train my dog to yawn on command. As silly as that may be, I wanted to test myself in training a novel behavior in my dog. I imagine the scenario when one asks "can your dog do any tricks?" and a smile breaks across my face "watch this".

whenever I get home from work, or get out of bed in the morning, my Golden Retriever greets me with what many would call "a big stretch". Often times this is accompanied with a large yawn. Out of pure enjoyment and curiosity, I began to offer reinforcement every time my dog stretched, at anytime throughout the day that I was home. When available I could offer training treats, however in times where I didn't have treats on hand, I gave him a good scratch behind his ears and marked the behavior with "yes" and "good boy".

Because of the nature of positive reinforcement, the frequency of that behavior increased and my dog began to stretch much more than was normal for him. Because yawning was already paired with stretching, the frequency of yawning also increased. Soon, he would only be reinforced for stretches that were accompanied by a yawn. Stretches without yawns became less frequent and my dog was often found yawning throughout the day. The behavior was put on cue with the word "yawn" and now just saying that word elicits a big stretch and yawn in my dog.

As simple and silly as this may seem, this is how all animals learn behavior. When a specific behavior is rewarded, the chance of that behavior happening increases (see our blog post about operant conditioning). If you get a bonus for showing up early for work, you're likely to keep showing up early for work. Below are some examples of what you can accomplish with shaping procedures.

Examples Of Shaping In Action

Simple Examples

• A teacher only calling on students who raise their hands (students are more likely to raise their hands in the future).

• A trainer rewarding a dog for sitting on cue (the dog is more likely and willing to sit in the future.

• A personal trainer praising their client for working out (the client is more likely to workout in the future).

• Offering a roommate a piece of chocolate each time they do the dishes (the roommate will become more likely to do the dishes).

Complex Examples

• A parent teaching their child to say a word and rewarding them every time they get closer to the correct pronunciation.

• A trainer teaching a dog to use an "outside" button and rewarding them with time outside every time the dog presses the button.

• Teaching a rat to pick up a marble by offering rewards when the rat gets close to the marble, then touching the marble, then grabbing the marble, and finally picking it off the ground.

• An occupational therapist encouraging and rewarding small movements, then moderate movements, and then full movements.

You may have read through this list and thought "well I could probably do some of those". You can! The principles behind behavior shaping are very universal and can be applied to almost any living thing to warrant a behavior change.

Common Mistakes And How To Avoid Them

As straight forward as this process may seem, there are also many things that could slow or weaken your shaping procedures that may frustrate you with slow progress. It is important to note that although shaping procedures do take time, they will go faster if you can avoid these pitfalls.

Don't try to work too fast. Like Pryor's first rule of shaping, make sure you are not pushing the behavior too quickly. If you move on from a behavior that was not fully formed, the behavior may deteriorate and you will struggle with the rest of the behavior chain. When training a dog to use an "outdoor button", don't just stick it on the wall and expect the dog to understand. Work slowly by rewarding any attention to the button and slowly rewarding increases in attention until the dog begins to press the button, slowly move the button closer to the wall and continue to reward button pressing until the button is in its final spot and the dog continues to press it for a reward. The reward can then shift from a treat, or toy, to going outside.

Marking the behavior at the right time. Dogs will associate any reward to the behavior that was happening just prior to that reward. If you wait too long, the dog will not associate the reward with the behavior and it wont have a positive effect on that behavior. For example, when trying to strengthen a dogs recall, the reward should be given as soon as the dog gets to you. Having the dog sit or wait once they get to you are not bad practices, however if done in initial training sessions, then the reinforcement will be connected to either the stay or the sit, not the recall.

Not taking breaks. Just like how a child is unlikely to sit at a table and do homework for an hour straight, when training a dog for too long they will eventually lose focus and drive. It is important to take breaks from the training sessions and come back to it. It may be hard to stop a training session when you feel that you are making good progress, however if you end the session on a high note (Pryor's 10th law of shaping) and come back to it later, you will likely have more success than if you didn't take the break. Training sessions between 15-30 minutes often can keep the dog focused and allow enough training for learning. I often advise clients to practice for up to 30 minutes at a time, with at least a 15 minute break between sessions. You can perform that cycle as many times as you would like or have time for throughout the day, granted that the dog is still motivated by the chosen reinforcer.

Not generalizing behavior. Some behaviors, especially those that happen outside the home, need to go through generalization before it can be reliable. Dogs have a much harder time generalizing than people do. You may have noticed that your dog might sit really well in your living room, however as soon as you step outside it really struggles with the same command. It is important to practice commands and behavior in multiple contexts and environments for the dog to really understand the behavior. Most canine behavior is not tied to location, although some definitely may be. Practice in different locations. Practice with small distractions (or large distractions for an even more reliability). Practice with a longer duration of the command or more distance between yourself and your dog.

Takeaway

Many people practice shaping without ever thinking about it. In sports, you practice and work your way up to the demands of competition. Teachers will provide an education with a gradual increase in difficulty. Parents will praise their child for saying "Ba" before they ever say "Ball".

Because shaping is so natural, it is wildly effective in changing behavior. Children learn languages, dogs learn how to perform complex behavior chains, and athletes continue to break world records year after year. Shaping isn't just for animals or behaviorists, it gives us a framework for learning and growth.

Next time you are working towards a goal, whether that be training a dog, a personal goal, or a business goal, think about the effects that shaping could have. Understanding shaping and its procedures is great, but does nothing if we don't choose to apply it. When making changes with the help of shaping, you can expect to see success and progress that is very real and very permanent.