Improve Your Dog Training Skills with Secondary Reinforcement
If you are your dog’s trainer, this is for you.
Photo by Maksym Kaharlytskyi
When we train our dogs sometimes we question our efficacy and some other times the efficacy of whatever method it is we follow. While there are methods that suck, most of the times the fault of an unresponsive dog is ours.
When we meet a dead end path in dog training, it’s good to employ training devices and concepts that can unblock us. This is where secondary reinforcement enters.
Secondary reinforcement is a simple idea of using cues (e.g., sounds) for reinforcing behaviors. Those cues have no initial value to the dog but acquire power when linked with cues that have.
Let me give you some needed definitions as we are breaking things down.
What is a reinforcer?
Reinforcers are training devices used in dog training to encourage the dog to perform desired behaviors. Most notably (and recommended) they take the form of something pleasant, like praise or food, so that the dog is motivated to offer the requested behavior.
We add the reward immediately after the behavior occurs. For example, you give the command/signal ”sit”. The dog sits, and you immediately give a treat. The treat is the reinforcer. The way to encourage the successful response to our command.
Food and praise are a universal reinforcements. You don’t have to teach a dog to like food or praise. Dogs naturally want food and praise and take pleasure from it. That’s why we call them primary reinforcers.
What is a primary reinforcer?
A primary reinforcer is a reinforcer that is tied to a dog’s biological needs. Dogs are born needing, wanting, or desiring it. You don’t have to teach your dog to like it. Examples of primary reinforcers are food, shelter, play, touch. Primary reinforcers are inherently reinforcing.
And what is secondary reinforcement?
Now, introduce a sound into the equation, and a secondary reinforcer is born. In the example above, say ‘yes’ immediately after your dog sits, followed by the treat. With time and repetition, ‘yes’ will acquire the reinforcing properties of the treat and your dog will learn to respond to it in the same way she/he responds to the treat.
Secondary reinforcers are sounds or visuals that, although initially neutral to dogs, get their reinforcing properties only when linked with primary reinforcers. After the association is established, secondary reinforcers can replace the primary one, mark the exact moment that the rewarded behavior occurs with more precision, and/or links the behavior to the reward.
A widely used secondary reinforcer is the well-known clicker. It is powerful because of its distinctive sound. Clickers are a proven method to help dogs learn faster. There is a ton of work you can do with a clicker. The methodology is simple: Get the behavior, click the behavior, reward the behavior. The sound comes immediately after the desired behavior occurs and is followed by a treat. This way, the clicker becomes a powerful training tool that is tied to the desired treat.
Trainers call clickers ”event-markers”. Clickers mark the behavior precisely and tell the dog exactly which behavior is wanted. It is particularly useful for high performance dogs as it shortens the length of the signal and thus marks the exact instant the dog offers the wanted behavior. That’s why it is hard to find an equally effective alternative.
I personally find clickers annoying for two reasons: first, the sound is very penetrative for my sensitive right eardrum and second, I found it inconvenient that I had to carry one with me at all times when Aria was learning the basics. I prefer verbal secondary reinforcers.
Words come naturally when we interact with our dogs. What we want is a short, plain, soft but firm word to mark the behavior.
The methodology is the same. You give the command, your dog responds, you say ”Yes”, ”Bravo”, ”Nice”, or whatever comes naturally to you, and then you add the reward. As one of the pioneers in clicker training, Gary Wikes says, the shorter the better. Your words are the event markers, you are trying to mimic the more precise clicker. Remember to mark and then treat.
Real-life example. When Aria walks more than 50 meters ahead of me, or closer, depending on the surroundings, I call her to stop. The moment she responds, unable to deliver the reward, I say bravo. She knows she did the right thing because bravo is a well-known secondary reinforcer to her. She expects a treat but it doesn’t happen all the time. Treats happen on a variable schedule. In this example, the secondary reinforcer ”bravo” has been linked to the primary reinforcer, the treat. Bravo does a perfect job letting her know the exact time she did the right thing to stop, and wait there with the hope for the well-deserved treat.
Key-points and clarifications:
1. Why use secondary reinforcement?
You might wonder ”why should I use secondary reinforcers since a small bite of turkey does the work?”. A small bite of turkey does the work for me too. My dog is food driven.
There are some fierce advocates for the somewhat controversial clicker, like Linda P. Case, that convinced me once to use a clicker. I know they work. I didn’t find them essential and stopped using them. But, as I said earlier, I use the word ”bravo” to let Aria know she has my approval. ”Bravo” is regularly ‘charged’ with food, her biggest will power.
I use verbal secondary reinforcers for two simple reasons. First, I don’t want us to be dependent on treats. Second, at times when the treat can’t be immediately delivered, secondary reinforcers are very handy.
As Ken Ramirez from Karen Pryor Clicker Training explains: ”While it may not be necessary for learning to occur, a clicker is an excellent tool that, when used properly, can facilitate better training. Animals can and do learn without deliberate human-created markers; a marker may not be needed when reinforcement can be delivered immediately.”
Ken Ramirez, an renowned animal trainer, stretches the effectiveness of a secondary reinforces (the clicker) as an event marker, and that, as long as your primary reinforcer can be delivered immediately, secondary reinforcers are not essential.
As per research, the timing of reinforcement is crucial for your dog’s responsiveness. Delays to reinforcement leave room for unintentional feedback to occur. And this is where a secondary reinforcement enters. In cases where a treat or any primary reinforcer cannot be given immediately, a clicker will mark the behavior with precision.
2. Secondary reinforcers need consistent pairing with a primary
Secondary reinforcers lose power when not continuously paired with a primary reinforcer. My dog no longer reacts to the sound of the clicker. We haven’t used it in a year. For the purposes of this post I unearthed a clicker from the drawer and I clicked to see if the sound means anything to her. I flinched with aversion and Aria just turned her head indifferently.
3. Diversify your reinforcers
So, now that you know that secondary reinforcers cannot stand alone, remember that a smart dog folk use many kinds of primary reinforcers. When your dog responds, reward them with food, praise, their favorite ball, a kiss! Diversity enriches the communication.
4. Do not mistake praise with verbal secondary reinforcers
Keep your high-pitched voice for the right moment. Praise works fine as primary reinforcer for dogs, as it is inherently stimulating to them. Verbal secondary reinforcers should be short, soft but firm words. If that helps, remember that you are trying to mimic the more precise clicker. Reserve too much praise for when you want to show to your dog that she/he did a great job at the end of a training session, or for the times when your dog does something very impressive, whether in or out of formal training.
5. Secondary reinforcement can be viewed as a case of classical conditioning
The classical conditioning model describes how dogs learn by association, without any effort on our behalf. Dogs associate two or more events and react to them according to whether there are positive or negative emotions linked to one or the other. Secondary reinforcement is a concept that utilizes this functionality and puts it into the context of dog training.
Associative learning happens all the time:
- A dog learns to like cars when what he knows about them is that ”cars take me to the lake, where I can enjoy my favorite activity, swimming”.
- A stray dog fears and shows aggression towards motorbikes because ”these moth@@ers keep startling me”.
- Dogs get aroused by the sound of plastic packages, because this sound often equals a treat.
- Dogs ignore hairdryers until you turn them on and the memory of the king of all monsters, the vacuum, comes in mind.
Back to our topic and in the same fashion; Dogs learn to respond to secondary reinforcers because they are linked to positive experiences.
Anything that your dog likes via positive associations will be a useful secondary reinforcement if used wisely and in a way that makes sense to you and your dog:
- Your dog gets aroused when the loving neighbor approaches. Not all neighbors have this effect on your dog. The love she receives is a primary reinforcer that makes this particular neighbor your dog’s favorite that enriches her experience of the outside. If your neighbor’s name is Mike, your dog will react to it too.
- Your dog gets excited every time the doorbell rings. Doorbells mean guests. Or pizza. Doorbells enrich(reward) the inside experience. A similar sound would work fine as a secondary reinforcer.
- My dog’s best and undeniable girlfriend is Nali. Nali is closely tied to happy times at the park. I use the cue ‘Nali’ as a reinforcer. It works every time. (To avoid confusion: Nali, the dog, enriches Aria’s life and reinforcing her to go to the park. But, to deploy Nali for my purposes, I add the the sound ‘Nali’, in a way that makes sense, when Aria responds to me.)
Final quick thought
Secondary reinforcers have three basic functionalities. They replace primary reinforcers, they are event-markers and they link the behavior to the reward. I find them useful and ingenious. I love to play around with these concepts, as they give me more ways to communicate with my dog. If you feel like this is something to experiment with, do it. It might be what your training sessions have been lacking.
- Karen Pryor Clickertraining: pdf
- The Meaning of Click – Linda Case
- Gary Wikes: Clicker Training: What it isn’t
- Gary Wikes: Information + Motivation – The Essence of the Clicker