Stubborn Dog or Poisoned Cue?
Does your dog have a delayed or unreliable response to Come, Down, or other commands? Do you see certain stress signals, like lip licking, looking away, sniffing the ground, moving slowly, or yawning? These responses indicate a poisoned cue.
First, we need to go over the difference between a Cue and a Command.
A Cue invites the dog (or any other species) to perform a behavior, which is followed by a reinforcer— a treat, praise, toy, play, or whatever else is valuable to the dog in that context. If the behavior doesn’t happen, there is no reinforcer and no aversive consequence (“correction”). What to do next is a topic for another day.
A Command, originating from the military method of training, carries an implied threat: Do it or else. Come, for instance, may have been trained by pulling the dog towards the handler by the leash. Or in the case of Down, by pulling towards the floor. Even if the dog was rewarded with a treat after getting into the right position, the Command itself is now associated with something unpleasant.
The typical response of an animal trained with positive reinforcement is happy, even joyful. Oh boy! A dog who associates a cue with only positive consequences looks very different from one who has experienced yanking, dragging, shock, or threats (like showing the water bottle or shake can, or taking a confrontational posture). Uh-oh or Oh no!
One of the sessions I attended at Clicker Expo this month addressed how easily cues can become poisoned and what to do about it. The speaker was Dr. Jesus Rosales-Ruiz, a professor of behavior analytics at the University of North Texas, who is also on the faculty of the Karen Pryor Academy, the organizer of the annual Expo and the premier resource for learning about clicker training and behavior science.
Equipment Can Be Poisoned Too
Although I was familiar with the term, I hadn’t realized how far it can go. Take equipment, for instance: a prong collar or uncomfortable harness predicts that pain or discomfort will be experienced sooner or later on the walk. That can set the stage for reactivity, by the way.
A collar or leash that’s used to deliver “harmless pops” (as some trainers call them) can be viewed by the dog as a warning not to pull or a prediction that he will be forced to sit, even when he feels unsafe, for example in the presence of an unfamiliar, scary dog or person.
Even a dog’s name can be poisoned if it’s often said in anger or followed by punishment—that is, anything you do that decreases an undesirable behavior. That’s why it’s usually a good idea to change the name of a dog with previous owners.
Dr Rosales-Ruiz demonstrated in videos of 100+ training sessions conducted by his students how a small poodle learned to respond to the cue Ven (using clicker and treats only) versus the command Punir (using leash pulling at first, together with clicker and treats). The dog’s body language and response to Punir was significantly worse, even long after the pulling had stopped.
Ambiguity Causes Stress, Hesitation
The conclusion from this and other similar studies is that when the Cue or Command becomes ambiguous—that is, the dog is unsure whether it will be followed by a reinforcer or an aversive—the desired behavior breaks down and becomes unreliable.
When people don’t recognize those stress signals (above) and assume their dog is just being stubborn, then the relationship can start to break down too.
Here is more proof that if you want your dog to love learning as much as you want his responses to be quick and reliable, you'll use positive reinforcement and avoid so-called corrections—and avoid trainers who use both and may call their method “balanced.”
If you believe some (or all) of your cues have become poisoned, the good news is that you can simply change the cue or the equipment and start over with positive reinforcement only.
Copyright Lisa Benshoff 2018