Whoa there, friends. Don’t misquote me. Before I go any further, I want to clarify that I am talking about negative punishment.
Negative punishment is the kind where you remove something the animal wants when they do an undesired behavior. (That’s where the “negative” comes from. Something is being removed. Check out my post on the four processes of operant learning if you want to learn more.) The dog lunges for the toy in your hand; you make it inaccessible. The dog learns that lunging makes the toy go away. If you are consistent and there are no other influences involved, lunging behavior will decrease.
So I’m not talking about hurting my dogs. But I am talking about trying to squelch behavior.
Most clicker trainers find the use of negative punishment ethically acceptable in at least some situations. It is considered most useful and acceptable when paired with positive reinforcement. In the above example, you could hold the ball out of reach (or put it away in a pocket) until the dog stopped lunging. The instant the dog did an acceptable behavior, such as sitting quietly, you could whisk the ball out and toss it to the dog. These pairings of consequences can teach the desired behavior very quickly. (Still, I try not to do them as a first choice. Contrary to popular belief, and certainly counterintuitively, animals don’t have to learn what the “wrong” behavior is or get punished for it in order to perform the right one consistently.)
However, I also believe there are times when even negative punishment is clearly unfair. And I can define “fair” in behavioral terms. Fair in this case is when criteria are clear and the reinforcers (or punishers) are consistent. Unfair is when they are not.
It is a problem that negative punishment is so easy to learn to dole out. It can become habitual. And dang, sometimes it feels so good to just get the dog to stop whatever it is. Punishment is reinforcing to the punisher in that way. I don’t pretend to be immune.
So here are my three examples. The things I am refraining from punishing. See if you agree.
So a couple of years ago I was at a nice shopping mall with my friend. I had Zani with me and my friend had her dog. We were standing on the pavement chatting. Zani must have been offering behaviors and had failed in getting my attention. I surmise this because after a while I looked down and she was lying flat on her side on the cold pavement, with her eyes cutting up at me. As if to say, “Is this enough to get your attention?” I took a picture that very first time. Here it is.
I have intermittently reinforced that behavior and have it halfway on cue: “Flounder!” It has remained what it always was: an extreme form of down. Her “ultra-down.” As if Zani thinks, “If down doesn’t work, let’s try this!”
In keeping with this bid for attention, which I was OK with thus far, I started hearing this little groan when she would flop down. I knew immediately this was trouble. If it got reinforced, I was going to get groaned at in addition to being floundered at. I became super diligent about not reinforcing the Flounder if I had heard her groan first. But I was too late. I think that at the beginning she was groaning softly enough that I didn’t hear it, and that sometimes I still don’t hear it. Or perhaps at times I have reacted directly to the groan and possibly reinforced it by turning and looking at her. I must have accidentally reinforced Flounders that started with a groan because, guess what, groans are increasing!
This happens most often in the kitchen while she is on her mat while I cook. So now the scenario is that she might have been lying on her mat quietly for 10 minutes, and I feel like I really ought to reinforce such nice behavior, but, uh, did she groan first? I can’t remember. I hope she doesn’t remember either and I give her a treat.
At this point when I hear her groan, I have to hold myself back from picking her up and taking her out of the room. (This would be an attempt at a timeout; negative punishment in the form of a removal from the opportunity from reinforcement.) The behavior is that irritating to me.
Compared to a lot of things dogs do, it’s a small problem. I imagine it sounds petty of me to complain about it. Zani is adorable. But aversives get to be defined by the one who is experiencing them. When she groans now it just goes all over me. And I really really would love to make it stop. But I believe that applying the negative punishment of removing her from her mat and the kitchen when she did it would not be fair. Because from her point of view that means that sometimes she gets reinforced, and sometimes she gets punished for the same behavior.
There are ways that I could train away the groan. They could be time consuming. So at the moment I try my best not to reinforce it, try like hell to ignore it, and grit my teeth.
The Bark, Check In, and Bark Some More Loop
Clara has some pretty choice behaviors too.
This is entirely a behavior I trained. Even at the beginning I mostly realized the consequences, and in an analytical way it is preferable to almost all other choices. However, that doesn’t preclude me from getting irritated.
I have previously written about Clara’s classically conditioned and operant responses to other dogs barking and other distractions. I paired other dogs barking with treats raining from the sky, and she has a positive emotional reaction to that. It turned into a reorientation, then a recall as she started to seek me out for her treats.
You can see the version I’m discussing at about two minutes into this movie. Clara is in the back yard and I am in the house. The back door is open, as always when a dog is out. Clara barks at something in the yard, interrupts her own barking, and comes in to check in with me. I give her a treat. I love that she doesn’t stay out there endlessly barking. A few barks and a check in are fine.
Except, what typically happens next? Lather, rinse, repeat, that’s what. The door is still open. Whatever was out there for her to bark at is probably still out there. So what is she going to do after she has checked in and gotten her treat? Hang around doing nothing? Nope. Run out there and bark again and come back again. And again. I remember that this is exactly what I taught her to do. I have richly reinforced the behavior. I didn’t convey to her, “And you can only do this once! Afterwards you have to be quiet and stop being a dog.” Doesn’t work that way.
But that doesn’t stop me from being irritated. This usually happens when I am trying to make my lunch, and also letting the dogs be outside for a while to break up their day. What I feel like doing on one of her trips in is sticking her in a crate. But that would be unfair and unproductive. How can it be right that sometimes I would give her a treat for coming in (away from something exciting, I might add), but that sometimes I would give her a sour look and stick her in a crate?
For this situation, my solution is to allow two or three iterations, then the last time I go and close the door so she can’t keep going in and out. (You can hear me mention this in the movie.) There are times when she doesn’t start barking again so I don’t want to jump the gun the first time she comes in. I am careful not to associate anything negative to coming in and checking with me. That’s not the problem! The problem is her going out again to repeat the process. And of course I am pleasant about it when I finally shut the door. I don’t make it a timeout from reinforcement to be in the house with me.
Poor Zani gets mentioned twice this time. Her other annoying behavior is also related to matting in the kitchen.
Because of the logistics of four not entirely compatible dogs, Zani is most often in the kitchen when I am. She has a mat to get on that is out of my immediate working area. She gets reinforced for lying quietly on her mat while I work. Before I continue I want to remind myself and my readers that the problem is mostly in my head. I have a dog who will go get on a mat on cue, and stay there for long periods of time for pretty sparse reinforcement. That is a great thing! I am truly sweating the small stuff, but that’s how it is sometimes.
But my dream of how her behavior should be is that my walking into my area (I used to even have a piece of tape on the floor to mark it off for myself) cues her to get on her mat and stay there until released. In return she’ll get some food treats, perhaps part of what I am cooking if that is appropriate. Also part of my dream is that I completely avoid dropping crumbs, so there is never anything enticing on the floor in my area. Dream on.
But because of some complications, I decided I couldn’t rely on the cue of my walking into a certain area of the kitchen. I made a conscious decision long ago that I would verbally cue Zani when I wanted her on her mat, and that the rest of the time she was allowed anywhere in the kitchen. This works fine.
Except that sometimes I forget. And sometimes she gets on her mat first and I think I have cued her but I haven’t. And also, she eats her meals out of a food toy in the very area that I want to be offlimits the rest of the time, so of course it is an enticing place.
Add to all this the fact the Zani is the most intense scavenger of all my dogs, and I have a little dog who comes into “my part” of the kitchen and sniffs around fairly frequently. The worst thing is that I am convinced in my own mind that she shouldn’t be in there at all, even though I made the conscious decision that she should only stay out when I cue her to get on her mat. So I can be doing something completely different, say, sitting at the kitchen table. When that is the case, and I haven’t cued mat, anywhere in the kitchen is permissible for her. But it still really really bugs me when I see her go sniffing around in the cooking area.
Again, I am tempted to perform a timeout. Remove her from the kitchen the instant she heads into that area. Just get her out of there for a while and show her that it’s not OK. For many dogs that would be negative punishment, as one is briefly removing the opportunity for reinforcement. But for sensitive Zani, either being bodily picked up or led out of the room by her collar in this situation would qualify as positive punishment. Adding an aversive to the environment to decrease a behavior. Not just neutrally making reinforcement unavailable. Those things would be very uncomfortable for her. But I can tell you right now they would be insufficient to decrease the behavior. Her urge to scavenge is way too strong. So that’s another reason not to resort to this aversive technique. Even if it were “fair,” it wouldn’t work.
I have shared these scenarios not as some kind of confession of being an awful person. I think I am probably a pretty typical person who lives with a lot of dogs. (Is that an oxymoron?) I get irritated sometimes. I get tired of doggie behaviors. One of the first books that taught me about the mismatch between human and dog behavior was The Culture Clash by Jean Donaldson. She has been saying it for years. They bug us. We bug them.
The problem is that since I know these neat negative punishment methods for dealing with behaviors, I am tempted to use them as a shortcut to getting what I want.
I shared the stories as an example of the kind of introspection that seems necessary, for me at least, to be a good person for my dogs. This is being tough with myself. And also I wanted to describe once again the seductiveness of punishment. It’s always lurking in the corner, ready to pop out and be put to use.
I remind myself that it is not fair to apply any kind of punishment, positive or negative, to a behavior that is also being reinforced, sometimes directly by me! And I remind myself how good my dogs really are.
Thanks for reading! Stay tuned for:
Stay tuned for: