While I’m no expert in dog training, I’m going to take a stab and explaining poisoned cues. I’m also including an article from dog expert, Karen Pryer on the subject. It’s a little academic, but for those who want to learn more about poisoned cues, it’s a good place to start.
What is a poisoned cue?
First of all, let’s talk about the difference between a “command” and a “cue.”
Both commands and cues are signals for the dog to do something specific. The word “command” is associated with traditional dog training methods, where compliance does not usually result in reward. The dog is simply expected to do what it’s told and noncompliance brings on a correction.
For example, in order to teach my daughter’s first dog, Jami, to sit,
- I’d say “sit.”
- If Jami sat, I offered verbal praise,
- If she didn’t sit, I’d correct, by pushing into her chest and down on her backside until she sat.
(I used a similar strategy with Ripple, but offered treats and praise even if she required follow through.)
During these early days, there was no motivation for Jami to sit for me. Despite this, she learned to follow the command, probably to avoid my correction.
In dog trainer and author, Karen Pryer’s article on poisoned cues,( http://www.clickertraining.com/node/164) she refers to a cue as something that always “opens the door” to reward. She and other clicker trainers refer to a trainer’s signals as a “cue,” distinguishing this from the traditional word “command.” One big
difference between the two terms has to do with correction. Clicker trainers ensure that the dog is offering the behavior before they name it.
- watch her, so that we notice whenever she sits.
- whenever she sits, we click
- whenever we click, we offer her a treat.
Rewarding a behavior that you want increases the likelihood that the dog will keep doing it.
Once the dog seems to understand the cause and effect (the act of sitting = click/treat,)
then the cue, “sit” is added. If the cue is given and the dog does not sit, the only consequence is that no reward is offered until the dog complies. Correction or follow through is never imposed.
A poisoned cue is a command, cue, or signal that the dog no longer sees as an open door to rewards, but instead views as something that leads to a worsening condition.
Ripple does not consistently “sit” when told to do so. In fact, at times her response to this command is to back up, so as to avoid being touched in the rear. Ripple has come to associate the command “sit” with punishment.
And to make matters worse. It’s my fault.
What can a handler do?
Go back to the basics
When Ripple hears the command, “sit,” she likely thinks, “I don’t want to be touched,” which has made her less compliant. One way of repairing this is by building good associations with the act of sitting. So, we’ve worked with Ripple using similar strategies to the ones we are using to teach Cassidy.
- Whenever Ripple would sit we’d click/treat.
- It hasn’t taken her long to figure out that ‘sitting,’ is an opportunity to earn good things.
- The next step is to pair the act of sitting with a name. (Because the cue, “sit,” has been poisoned, the best practice is to choose a different name, like, “rest,” or, “easy.” Our problem is that Cassidy is now responding to, “sit.” We are hoping that repairing the damage done to Ripple, can be accomplished while still using the same cue.)
Consider what unintended condition a cue may open the door to.
I don’t mean to imply that traditional dog training methods always result in poisoned cues. As I mentioned, I used similar methods with our dog Jami, and she grew to become a very compliant, well mannered dog.
Ripple came to us with insecurities. For her some things that we wanted her to do, like having her go through the doggie door or allow us to touch her backside, made her unsure or even frightened. We made the mistake of using important cues, like “sit,” and her name, when she was confronted with these uncomfortable conditions, instead of first taking steps to desensitize her to them. (We are currently trying to desensitize Ripple to being pet. She has the habit of wincing and backing away when someone reaches out to pet her. I’m not sure why, we’ve had her since she was three-months old, and have never raised a hand to her. Ripple and I now have petting sessions as part of our training, where this experience is repeated many times and paired with clicks and treats.) As a result, Ripple’s response to a given command is to stand back and consider just what she’d be getting herself into.
Now that we are aware of this, we are more observant of Cassidy’s insecurities, and are careful not to use important cues in context with situations where she seems unsure. We are also trying to only say her name in conjunction with praise and treats. So, if she is chasing the cats we are trying to say, “No,” rather than using her name and saying, “No.” If we slip, (which does happen) when she runs over to us (because we’ve said her name) we give her praise and a treat (in both cases we’ve resolved the issue, because she is no longer chasing the cats.)
We are also clicking or giving praise each times she calmly approaches or walks by the cats.
A last thought
If we work towards avoiding negative association with the cues we are teaching Cassidy, and make every cue given an open door to reward, there’s a good chance that she will happily comply with almost everything we ask.