I don’t think this post is going to win a popularity contest, but here goes anyway. I can’t get it off my mind.
Trainers regularly work hard to teach people alternatives to endlessly saying “No!” to their dogs. Even those of us who know the pitfalls of the habit lapse into it from time to time.
But I seem to disagree with many others about what exactly those pitfalls are.
Here’s why I think yelling “No!” is a bad idea: most people who are doing it haven’t taught it as a cue for a behavior trained with positive reinforcement. It ends up as an aversive method and carries all the usual potential for fallout. It relies first on a startle response. If the dog habituates, then people escalate the aversives.
But that’s not the objection I usually hear.
The Common Objection to “No”
I read it again the other day, in a discussion advising someone who was dealing with an undesirable behavior by her dog. She had been telling her dog “No!” when he performed the behavior. Several people chimed in, pointing out two related things: “no” is not a behavior, and saying “No!” didn’t tell the dog what he should do.
Both true statements. But they point to a failure in training, not some magical property (or lack of property) of the word.
The statement that “no” doesn’t tell the dog what to do is also true for every single verbal cue we use—we have to teach the association. For instance, merely saying the phrase “turn around” doesn’t give the dog any information about what we want them to do, either. A cue and a behavior are two different things. We train the latter and associate it with the former.
R+ trainers commonly say two things that are contradictory.
- On one hand, we tell newbies any word can be a cue. This is true. “Lightbulb” can cue sit. “Resonate” can cue the dog to look at me. Trainers just have to remember them and be able to teach the dogs. Cues don’t even have to be words. A cue can be a hand on a doorknob, the sound of a car approaching, a time of day, or the odor of vinegar. This takes a while for most of us to comprehend, because the language aspect is typically much more salient to us humans than anything else. And we tend to backslide. We persistently mix up the meaning of the word with its function as a discriminative stimulus. I discuss this in my blog post, “Good Sit!”
- But then we also tell people that “no” is not a behavior. That’s also true, but not really relevant. When we say “sit,” “down,” or “lightbulb,” those aren’t behaviors either when they come out of our mouths. They are cues. “No” is not a behavior, but it doesn’t have to be. It just needs to indicate reinforcement is available for a behavior. We don’t say that a hand on a doorknob or the smell of vinegar can’t be cues because they aren’t dog behaviors.
Singling out “no” as uniquely meaningless isn’t logical.
The Real Problem with No
I believe the root problem with “no” is that people don’t train it; the word doesn’t point to a behavior that will be followed with positive reinforcement. And if saying it doesn’t successfully interrupt the dog, people usually escalate. So “No!” comes to predict aversive conditions: nagging, yelling, stomping, clapping, or even physical aversives like hitting.
Dog trainers rightly advise their clients to start over and use another word if they are going to teach a “leave-it” or an interrupter, because most of us rarely say the word “no” to dogs nicely.
But we can. I have a friend who practiced for ages to use “no” as her leave-it cue for her service dog so she could say it in a pleasant and neutral tone of voice.
When I Yelled “No!”
Believe it or not, I yelled “No!” on the same day I started this article, right after I was pondering this whole thing.
I make a baked dessert out of oatmeal, egg whites, almond butter, dried cranberries, and dark chocolate. A lot of dark chocolate. I warmed a piece of it that night on a plate and put it on the counter. You know what’s coming. I turned around and Lewis was countersurfing. He had his nose up, sniffing the dessert, about to take a bite.
Even though I have taught Lewis a leave-it cue, I panicked, yelled “NO!” and clapped my hands. I did exactly what I’ve been describing. I yelled, hoping to startle him, and when that didn’t work instantly, I clapped, with the same goal.
What did Lewis do?
He didn’t cringe or cower or run away. He slid slowly down from the counter and calmly came to me, expecting a treat. I gave him a handful, then I removed the dessert from his reach.
I haven’t trained the word “no” as a cue, but I’ve trained several other words that function to interrupt, and he is accustomed in particular to being called away from the counter. So to him, it didn’t matter what I said, nor, apparently, how I said it. Lewis associated a behavior (reorienting to me) with my saying “No!” because of the foundation of training we already had.
I taught him “Pas” (leave it), “Excuse me,” (put all four paws on the ground), and “Lewis” in a high, singsong tone (come here). None of those words or phrases “was a behavior” when he first heard them either, but now they signify good stuff if he performs the behavior I’ve associated with them. And by generalization, so did the “no.”
I used to train “Hey!” I carefully conditioned it to predict great things for dogs who come to me, since that was what usually came out of my mouth when I panicked about something that affected a dog. I even practiced it in an irritated tone, so the good reinforcer hopefully counterconditioned my cranky tone. You can see a demo here. I should do this with Lewis as well.
There is a lesson to be learned here. The positive reinforcement-taught cue for Lewis to get down from the counter is: “The lady says something while I have my feet up on the counter.” Yes, any word can be a cue, but often it’s not the word at all. We humans are the ones stuck focusing on the words.
And of course, I’m not suggesting that yelling “No!” to our dogs is a good thing. I’ve delineated the problem with it already. It worked out for me in that instant without fallout, but only because it resembled real training I had done. We might not have been so lucky. It would have been safer if I’d come out with one of my trained cues. I need to practice more, or maybe I should condition “No!” as well as “Hey!”.
Not Only a Semantic Argument
I thought hard before publishing this. It may give people the false impression that I am supporting yelling “No!”. I’m not! Or it may seem pointlessly picky. Maybe.
But my motivation is practical. Focusing on the word “no” and what it means or doesn’t mean feeds into the idea that cues drive behavior. If we center our argument on the word “no” not being a behavior, we are very close to implying that words like “sit” and “down” are behaviors. And this can strengthen our unconscious tendency to believe that dogs automatically understand language the way we do.
That’s the downside of saying, “No is not a behavior.” It adds to the confusion about words that are both cues and verbal descriptions of behaviors. Sometimes cues may describe behaviors, but it’s not necessary that they do.
I understand that the statements people make about “no” that bother me are shortcuts. Trainers don’t usually give a lecture on discriminative stimuli when first introducing people to R+ methods. And it’s true that people yelling “No!” are not usually thinking of what they want the dog to do; they are thinking of what they want the dog to stop doing. So it’s great to introduce the concept of training with positive reinforcement and get people thinking about building incompatible behaviors instead of repeatedly reacting in the moment.
I’m not a pro trainer; I don’t work with humans training their dogs every day. If telling people that “no doesn’t tell the dog what to do” helps most of them break the habit, then great.
But I bet there are others like me who eventually want to understand this stuff about cues a little better, and the claims about “no” can slow that down. I know, because it’s taken me 10 years to unravel even a little of it for myself.
Copyright 2022 Eileen Anderson