In this blog I’ll talk about the NRM/interrupter that I choose to use the majority of the time my dogs make an error within a chain. This is not technically a NRM because it does MORE than mark the moment of the error; it totally interrupts the dog so that the dog is not engaging with whatever caused their error in the first place. At the same time, I go to a good deal of trouble to remain cheerful so that the dog know that I still love them and want them to stay in the game with me. In effect, I want to maintain the dog’s effort and ensure no hard feelings. Used in this manner, my “cheerful interrupter” (CI) is reasonably safe, even for dogs in the learning phases.
Using a CI, my goal is to get the dog set up to try again as quickly as possible, to provide a positive (rather than a neutral) tone, and to give assistance if needed so that the dog may finish successfully and receive a classic reinforcer.
Here’s a working example of a classic NRM:
Your dog is sent to retrieve the dumbbell over the jump. On the way, your dog heads for the dumbbell and suddenly catches a smell on the ground which causes them to drop their head and sniff for a second. At that moment you execute a verbal NRM like “wrong” with a neutral tone.
Assuming the dog understood, then when the dog hears the NRM, he should stop mid-sniff and simply return for another try. Even if the dog grabs the dumbbell, it’s irrelevant because the dog cannot “recover” the exercise after a NRM has been given. You begin again.
If the dog did not understand that sniffing was the problem, or if you’re timing was off, or if, after hearing the NRM the dog decides to continue sniffing, then you’re a bit out of luck. Generally you’ll have to enter the territory of punishment if you want to continue.
In contrast, here’s a working example of a cheerful interrupter:
Your dog is sent to retrieve the dumbbell over the jump. On the way, your dog heads for the dumbbell and suddenly catches a smell on the ground which causes them to drop their head and sniff for a second. When your see this, you immediately move towards the dog, speaking and moving in a manner that both interrupts their behavior and re-engages them with you.
When I do this, I’m perfectly cheerful! I usually say something like “Watchcha doing!” in a relatively happy voice and often I’ll remind them of the delicious cookie which they will not be getting. The combination of movement, cheerful chatter, and possibly showing the dog the reinforcer almost always interrupts whatever the dog might have been doing (in this case sniffing when they should have been retrieving), and in a short period of time it becomes clear to the dog – no reinforcement for that effort. I’ll pick up the retrieve object and repeat the request as fast as possible.
I like this approach better than a simple NRM because I think it more effectively maintains the relationship, specifically because it IS emotional. I’m cheerful because I value the engagement but there will be no reinforcement for that attempt. A CI interrupts whatever the dog might have been doing because they tend to notice my movement and they definitely hear my chatter. If I’m quick I can get the dumbbell so the dog doesn’t try to “recover” the exercise.
I also like this approach because I can use it in the learning phases. The only difference I’d make with a “trainee” is that in addition to interrupting the dog’s error, I’d consider making the next attempt easier in some fashion.
The reality is that our dogs care about traditional NRM’s for two reasons. One is the loss of reinforcement, and the other is the perceived disapproval of the “giver”, which can be emotionally hard on some dogs.
This is also the same approach that most of us use with human children but maybe we’ve never thought about it. Consider this example:
You’re working on multiplication tables with your child. The first three are perfect, so you move quickly through your flashcards, cheerfully saying “Yes!” for each correct response. You’re on a roll; your child is engaged by the flow and you’re having fun together. Then you get to the fourth problem and your child is wrong.
You don’t suck the life out of your terrific session by saying “no” unemotionally. Instead, you probably say something like, “close; try again!”. There is no disapproval because you know that it is your job to keep your young learner engaged. Maybe you lean forward a bit to show that you’re not upset or maybe you give a bit of help by repeating the question slowly to give your learner a chance to slow down and give it more thought.
And…what do you do if the child answers incorrectly two times in a row? Hopefully you pull the problem out of the chain. If the problem is 3 x 6, maybe you quickly review the “3” tables together – and after you get to 3 x 5, maybe you allow your student to complete the next one, 3 x 6, on their own.
And then….you go back to that fast paced session, quite possibly putting 3 x 6 in there many more times than you would have had you not become aware of the issue.
Of course, exactly how you handle a given session would depend on a lot of factors, not the least of which is the temperament of you and your child and what type of working relationship you have developed, but you get the basic idea. You work to keep them engaged.
So now that I’ve explained my CI, when might I use a traditional NRM?
When I’m working with a dog in the polishing phase I’ll consider using a NRM if I believe that the dog can perform correctly if they simply try harder. In that case, I’m asking the dog to work strictly for the possibility of reinforcement. With higher drive dogs that love to work and who possess a strong foundation, I find that this issue almost never comes up because the dog is always trying, but with less driven dogs who want the reinforcement but don’t want to put out effort to get it, then I’ll consider the use of a traditional NRM.
If the dog opts out of training rather than putting out more effort (which is a possibility), then that’s ok. We’ll end the session for the day and I’ll take some time to seriously reflect on the appropriateness of whatever I asked. In the next session, I may pick up exactly where I left off and push through. Or I may decide that I asked too much and we’ll change direction. As in so much of training, it simply depends.
I ONLY use a traditional NRM if I want to communicate to the dog that they are now 100% responsible and I will not help them, probably a dog that is basically trial ready for the exercise. Excessive use WILL erode your working relationship with your dog. Under more typical circumstances, I use a Cheerful Interruptor.
Here’s a video example demonstrating a Cheerful Interrupter with Brito.
Here we are working on a send to a platform in a novel environment, so we have distraction issues to work on as well as the actual exercise of getting up on a platform and sitting.
Notice that I talk relatively little when he is actually working. I talk much more when we are between exercises and being rewarded, but almost never when he is heading for the platform. Notice that I use the cookies to re-engage Brito when I lose him but I don’t actually give him the cookies – he has to earn them.
36 sec – I lose him when I set up for another send; I show him the cookie to bring him back. I give him that cookie for “setting up” because he is young and our recall is super important. I would not do that with an older dog.
43 sec on the very next send I lose him again. The lesson for me is that if he has just gotten a cookie for setting up, he may be even less likely to perform the target behavior. So this time I send him without a cookie for setting up and he is then successful.
1:01 – this time he goes on his own. We are just starting to add the idea of “stimulus control” so I call him back with my CI and we try again. To avoid another failure I send him directly without another set up. If he were an older or more experienced dog then I would not have simplified my expectation.
When I send him at 1:29 you’ll see his head drop on the way to his platform. That could easily become a part of the behavior chain if I allowed it to continue , so now I”m “on notice” to watch for it. If I see it again, I’ll need to make some kind of changes within our training.
In the next blog, I’ll consider the behavior chain of heeling, NRM’s and CI’s.