A long time ago (2006-2007) in a land far away (Texas), I wrote a research thesis. The idea of performing a constructional aggression treatment with aggressive dogs was floated by my professor, Dr. Jesus Rosales-Ruiz, to a meeting of the Organization of Reinforcement Contingencies with Animals (ORCA) at the University of North Texas. As a woman in my mid-40s with a husband and kids and pets of my own, it seemed reasonable, somehow, exciting, even, to launch into working with aggressive dogs. I raised my hand and CAT was conceived. Well, Jesus had already conceived of CAT, so I guess it began to incubate. Whatever. We got started very quickly. On Wednesday afternoons at 2:00 pm for the next two years, I had a Come to Jesus Meeting in his office in Chilton Hall at UNT and we watched videos of me and sometimes my dog approaching and retreating from scary aggressive dogs in very specific ways. He made suggestions, I threw in my two cents, then I went out and did what he suggested.
What is CAT?
The first word, constructional, is described in my blog post, The Constructional Approach. Click over there for a background on the concept. Then come back here. I’ll wait.
CAT is a research-based procedure for rehabilitating aggressive animals. The research was conducted with dogs, but it can be done with any species.
It involves first identifying the function of the problem behavior. Functional Behavior is behavior that results in a desirable outcome for the learner. (I will be writing a blog post about this in the coming weeks; Stay tuned.) This sounds like any old reinforcement-based training, but when we talk about functional behavior, we are typically concerned with problem behaviors that are being reinforced by something going on in the environment. So, yes, you can teach dogs to sit by giving them a treat, and the behavior functions to help them access treats. And that’s a powerful and cool training procedure. A situation in which we would be more likely to worry about function might be, the dog jumps on the countertop because that’s where his treats are kept and getting on the counter results in getting treats. In that case, the treats function to keep the dog jumping on the counter. We could easily use those treats for good and not evil by keeping them put away when not in use and by reinforcing staying on the floor with them at given times. The dog gets the treats he wants but only for doing what his person wants. Basic positive training 101, right?
A key point to understand is that with functional behavior change techniques for problem behavior is that there is a reinforcer already maintaining the behavior before you start training. The behavior works to get that kind of reinforcer. With clicker training a sit, the dog already knows how to sit, but we want to teach him to sit when we say to, so we use a treat, which we know he likes, to get him to sit when we want him to sit. The treat is not a functional reinforcer. It’s an arbitrary reinforcer that we hope will work because it’s easy to use for training. Fortunately, it usually does in those situations.
That is what CAT is. We identify the functional reinforcer for the aggression, and we deliver that reinforcer when we see better behaviors happen, and stop delivering the functional reinforcer when the problem behavior occurs. Lo and behold, the animal, who really wants that functional reinforcer, starts doing the behaviors that pay off… the safer behaviors… and stops doing the behavior that doesn’t pay off any more. This is a differential reinforcement procedure. Reinforce the behavior you want, and don’t reinforce the behavior you want to see less of. As you may have surmised, the dog may not immediately switch to friendly behavior instead of aggression. After all, the aggression is really well reinforced already, and friendly behavior in that context is not. So, you have to shape it up with wee little successive approximations.
What is the functional reinforcer for aggression? You can often look to the colloquial names given to the type of aggression to get an idea. Territorial aggression means the dog is trying to drive someone away from its space. Maternal aggression means the dog is trying to drive someone away from its babies. Leash aggression means the dog is trying to drive someone away while it’s wearing a leash, or that the dog is frustrated by the leash and wants to interact, so there are different things that may be the functional reinforcer. Prey aggression means the dog wants to kill an animal and perhaps eat it. Play aggression means the dog has learned that very rough play brings social interaction.
Although the thesis research focused on a certain kind of functional reinforcer, a contructional approach can be used with any of these kinds of aggression. But a biggie, and the most common, is the kind of aggression that functions to drive someone away. The functional reinforcer for this kind of aggression is distance from aversive stimuli.
What we do in CAT most of the time, is to set up a situation in which the aversive stimulus (usually a person or dog with a person) is positioned far enough away from the aggressive dog that the dog does not react badly but is aware they are there. When the dog does not behave aggressively, the aversive stimulus moves away. Over time the person or dog approaches closer to the dog in tiny increments, always turning and walking away when the dog does something that is not aggressive.
Send me the questions you have about CAT in the comments below, and you may see the answers in a future blog post.
- Snider, K. (2007). A constructional aggression treatment for dogs in homes and community settings. Retrieved from: https://digital.library.unt.edu/ark:/67531/metadc5120/
- Snider, K. (2018). Turning Fierce Dogs Friendly: Using Constructional Aggression Treatment to Rehabilitate Aggressive and Reactive Dogs (1st ed.). CompanionHouse Books.