Caitlin Coberly PhD
Secondary reinforcers and the use of a clicker as a secondary reinforcer have come up several times recently in our online discussions. I don’t know where all the hubbub Is coming from, but I thought I’d take this opportunity to discuss it.
A brief refresher
A secondary reinforcer is something that becomes associated with a primary reinforcer —something that satisfies a biological need like food, water, or sex — in the dog’s mind through association. It’s Pavlov’s bell. Basically, it’s the sound of the gumball machine just before the gumballs fall. It’s the can opener on the cat food or the crinkle on the treat pouch. It’s something that happens almost every single time a treat is about to be delivered, and it comes to mean “nice things” and “good feelings” in the dog’s mind. (For those neurologically focused, it causes a dopamine release — dopamine is an endogenous, highly addictive, “feel-good” neurotransmitter).
We can use this process in the dog’s mind to create secondary reinforcers that we can use instead of food. For example, I can ask my dog to recall, CLAP, and then feed a treat. If clapping is followed by a treat consistently, then the clapping itself starts to feel good, and I no longer have to follow it with a treat every time.
See this excellent video from Karen Pryor Academy of Ken Ramirez working with a beluga whale and using secondary reinforcers.
What we often don’t mention about secondary reinforcers is that they must be maintained. In order for the secondary reinforcer to maintain its “feel-good” value, it must occasionally be followed by a primary reinforcer.
A quick side note — some things can be taught as secondary reinforcers and never need to be paired with food again, because social interaction and play are both primary reinforcers. Thus, if you incorporate secondary reinforcers into play sessions, the animal is getting a primary reinforcer simultaneously — your attention and fun!
Marker training (or clicker training) seems similar to secondary reinforcement on the surface. We mark, or click, when the dog does the target behavior to indicate that reinforcement will follow. So the dog does the right behavior, we make a sound, and the dog gets a treat. It’s the exact same sequence!
Therefore, it is the exact same thing, right?
Not exactly. Here is where semantics comes in to play. It has to do with our intent. A marker is an indication that a reinforcer is going to follow. It causes that release of “feel-good” brain chemicals that we discussed before. All the same processes are set in motion. We intend it to mean “that behavior makes good stuff happen.” In contrast, a secondary reinforcer is an indicator that a primary reinforcer may follow. We use secondary reinforcers to mean “I am engaged and interacting, and more good stuff is likely to happen.”
Here is the critical difference. When we intentionally use a secondary reinforcer, we intend to create an emotional response that will not always be paired with a primary reinforcer. That is fine. We know that secondary reinforcers will hold up fairly robustly if they are regularly paired with a primary reinforcer. The strength of that emotional response will depend on the frequency with which the secondary is paired with the primary, as well as how much the animal wants the primary reinforcer.
In contrast, a marker is intended to act as an indicator that the animal did the desired behavior and is going to get a reward.
Are these differences critical? For the layperson, perhaps not. It’s not going to hurt your dog if you occasionally forget to mark a desired behavior, or if you pair a secondary reinforcer with a primary reinforcer when you didn’t intend to.
So what’s all the hubbub about?
Firstly, a professional trainer should be careful about their terms, especially those terms which have a formal meaning. Primary and secondary reinforcers have been carefully defined in the scientific literature. In contrast, marker training has not been so carefully defined. While it is used extensively in the animal training world, it has received little attention in the scientific community1 despite being first proposed by Skinner in 1951.2 Therefore, while we have lots of experience using markers, we don’t have a formal knowledge of the exact effects — including whether the animal understands the difference between a marker and a secondary reinforcer.
Secondly, our intent matters. Of course, we should always be careful to be precise. Practically speaking, when we are teaching a new behavior, the clearer we can be, the less frustration the learner has and the faster they acquire mastery. A marker is a useful way to bridge the gap between behavior and delivery of reinforcer. It identifies the precise moment the correct behavior occurs, and makes the process clearer to the learner. In addition, we can use a continuous schedule of reinforcement (i.e., each correct behavior gains a primary reinforcer) to provide absolute clarity.
In contrast, when we are working with established skills, using the same old reinforcer time after time can become boring.3 The animal may choose, unless deprived of basic needs, to do something else. While I allow animals to disengage with me, it indicates a failure on my part. My ideal training session is full of fun, excitement, and anticipation. By including a variety of reinforcers in my training program, I can keep the dog’s interest, as they never know exactly what they are going to obtain. By mixing up my secondary and primary reinforcers, I am, in essence, making the pattern of the game itself rewarding.
Practically, what does this mean for my training?
Use your marker (clicker or a verbal “yay!”) when learning new skills to mean “Yes, you’ve got it!” Use it to exclusively mean that a primary reinforcer is coming — be that a treat, a tug, or something else your dog really enjoys.
When you are working with already established behaviors, mix up your secondary reinforcers (a clap, a touch, a tongue click) with primary reinforcers (typically food), and keep the games going!
- Dorey, N.R, and Cox, D.J (2018) Function matters: a review of terminological differences in applied and basic clicker training research. PeerJ 6:e5621
- Skinner, B.F. (1951) How to teach animals. Scientific American 185:6
- Bromberg-Martin, E.S. et al (2011) Dopamine in motivational control: rewarding, aversive, and alerting. Neuron 68:5, 815-834.
Caitlin Coberly is perhaps best known as the founder and head trainer on Dog Training 101 Facebook group, where she oversees operations for the not-for-profit help group (over 100,000 members and 100-200 posts a day). She is a Certified Dog Behavior Consultant with IAABC, and holds a PhD in Evolutionary Biology from Duke University.
Caitlin specializes in extreme dog behavior problems, online diagnostics, and working with the human half to prevent euthanasia and surrender. She loves working with compulsive disorders (tackling the impossible), and focuses on “mental health and wellness”. Her own dogs get to enjoy 40 wonderful acres of happiness filled with bunnies and other critters. Caitlin’s experience training other species, including falcons, horses, fish, and even spiders, informs her work with dogs.
Caitlin shares her life with 7 amazing pointers, 3 horses and one big pink primate in beautiful western Oregon.