In the past two years I have moved away from teaching what I will refer to as “default impulse control.” This is typically taught in the form of food games (think “it’s your choice”) that teach our dogs NOT to eat available food via a negative punishment procedure. When the dog is fluent with food the “rules” are extended to all other reinforcers (like toys). I taught this for a long time (ten years at least) to my own dogs and those of my clients. If I were still teaching pet dog classes, I may not have evolved past this stuff when I did, but I am so glad my current dogs and clientele have lit the path to a smarter way to get similar behaviors. My primary clientele is currently the sport dog community; mainly dogs with issues that are considered arousal-based. I have taught my Worked Up program across the globe in seminars and online courses, and I can say that the most prevalent error dog trainers are making in my sport of agility, the error that is leading to these behaviors labeled “over-arousal,” is that we are not being clear enough about our reinforcers. The overwhelming popularity of “impulse control” training done with negative punishment procedures is a big part of this problem. I suspected this for a long time; every single email that lands in my inbox inquiring about my Worked Up program states that the dog has great impulse control training, but still fails to hold himself together on course. Could it be that supposed “impulse control” work is more than just Dumbo’s magic feather, but actually the culprit in some cases? I’m not sure I know the answer to that, but I do know this much: lack of clarity surrounding reinforcers produces big behaviors that handlers don’t like: behaviors like spinning, barking, biting, humping, zooming, sniffing, and so many more. When we teach our dogs NOT to take reinforcers rather than actively teaching them WHEN to take reinforcers, I believe we create these problems. As always the end goal, the trainer’s skill, and the dog’s comfort levels must all be considered in any training program. The good news is that teaching our dog clean stimulus control around reinforcers—rather than default impulse control—serves all three in most cases.
What’s the Difference?
In most impulse control training the dog is taught to leave food alone through a negative punishment route; any time the dog moves toward available food it is removed from the dog; so an appetitive stimulus (food in this case) is removed when a certain behavior is observed (motion toward the food/trying to eat the food) successfully inhibiting (punishing) motion toward the food/attempts to acquire the food. Here is an example of a pretty typical early session of this work with a friend’s board and train dog:
The way I am doing things now (and the way Felix, my youngest dog has been raised) is void of this procedure. Instead I have taught him cues to take reinforcement (we call these marker signals). He has a varied vernacular when it comes to his reinforcers; I use multiple markers for both food and toys.* This has helped him to be a level-headed partner for me when he may not have been, and I am now talking about these concepts in all of my seminars and online classes. I prefer this route for multiple reasons; not the least of which being that teaching our dogs what to do, rather than what not to do, is a professional and personal choice I have made in all facets of my training. I also like the clarity of the behaviors I get; in my seminars across the globe I have witnessed dogs with default impulse control training exhibit confusion surrounding reinforcers in ways that really slow down training. I have also seen too many young border collies that were already not keen on food completely turn off of food in all learning environments once this training was begun. Of course these things are not always going to be the case, so my primary reason remains that I choose positive reinforcement as my route from point A to B whenever possible, and teaching clean markers is that route in this case.
Teaching Reinforcer Stimulus Control
The best news here is that this game is easy! Teaching a dog to do something they are already quite good at (like eating or grabbing a toy) on cue is always a quick process. Like when teaching any other behavior we need only to see the behavior first happening reliably and then slap a cue on the behavior. Then, we prevent the dog from taking the reinforcement only in the absence of the cue, and we do so in such a way that the dog is successful from the start.
Here, I am teaching the cue “dish” to Chace, a dog I met in New Zealand. Chace doesn’t have a concept of this work yet, and has very little food-based default impulse control work (the least of the dogs available to me that day). I first pair the cue with the behavior, and then begin interrupting the dish behavior with other cues, cues I reinforce with “dish.”
Here, I am teaching Ghost the cue “good” which means hold still while I bring you the food. She hasn’t had default impulse control work since she left her breeder five years ago. I start by feeding her in a rapid fire manner, and over the course of three sessions she is waiting to be fed on my cue “good.”
The Final Product
My end goal for this work is twofold; a dog that makes training easy for me by not taking advantage of reinforcers out of turn, and a dog for whom training is easy because he fully understands when he gets to have his reinforcers. Watch Felix working a behavior for which I want to reinforce him in position with food, and then release him ahead to a toy on the ground. Because he understands that the toy on the ground has not been cued he doesn’t take it, and because he understands “good” means wait to be fed he holds position as I do so.
One Size Doesn’t Fit All
As I mentioned last week, all types of training will be individual choices. I am the last person to tell a person whose training choices are working well for them that they need to change. My only goal here is to encourage people who are having struggles to consider that perhaps some of the choices that are considered “best practices” may not be. Happy training!
*While I used different markers pertaining to food vs toys as long as I have been training with both food and toys, the concept of using an elaborate marker system involving location and type of reinforcers was first introduced to me by Shade Whitesel.