Operant Conditioning in Dog Training
So after reblogging the post earlier discussing negative punishment vs. negative reinforcement, I got 4 or 5 messages asking me to elaborate on the different quadrants and if x tool falls in which quadrant.Â
These are the four quadrants:
Positive doesnât mean good or bad in this context. Positive in the context of operant conditioning means âto addâ. You are adding something. That is all it means. Negative in this context means the exact opposite: you are removing something. Again, these terms cannot be understood in terms of every-day semantics.Â
Punishment means you are attempting to stop or discourage a behavior. Reinforcement means you are trying to create or encourage a behavior.Â
So here is the breakdown:Â
Positive Punishment means you are adding something to stop or discourage a behavior.Â
Negative Punishment means you are taking something away to stop or discourage a behavior.Â
Positive Reinforcement means you are adding something to encourage or create a behavior.Â
Negative Reinforcement means you are taking away something to encourage or create a behavior.Â
So when people ask questions like âso is this tool positive punishment or negative reinforcement?â it depends on the context of your use.Â
The quadrants in and of itself are mere arbitrary nomenclature that you are supposed to cover in dog training 101 and then move on from there. The tools you use cannot be neatly labeled and put into a corresponding quadrant box. Iâm continually surprised at the number of people who donât think that a head collar can be an aversive tool, or that a reward isnât always food.Â
Even a flat collar can be used as a tool for positive punishment. The tool itself is neutral, just like the quadrants. It is how you apply them that determines whether it is positive punishment or something else.Â
So what do these quadrants look like in practice?Â
Positive Punishment: Remember, you are adding something to stop a behavior. If my dog is pulling on her leash and I apply a strong force and jerk her backward, I am using positive punishment. If my dog barks at birds and I zap her with an e-collar, or spray her with a garden hose, I am using positive punishment. If I have a prong collar around my dogâs neck and it applies pressure to her when she pulls, I am using positive punishment.Â
Negative Reinforcement: Remember, you are taking away something to encourage or create a behavior. If I have prong collar around my dogâs neck and it is applying pressure to her, because she is pulling, and she stops pulling and the pressure is removed, I am using negative reinforcement. Yes, that means that in one single training moment, the prong collar is being used by two quadrants. If I want my dog to sit, and I zap her with an e-collar until she sits, I am using negative reinforcement.Â
Positive Reinforcement: Now here is the example, where most R+ trainers lose their minds. I have personally witnessed one person fail their dog trainer exam over this very question, because she could not wrap her around it. If I tell my dog to sit, and she does not move fast enough, and I zap her with the e-collar once to spurn her on, I am using positive reinforcement.Why? I am adding something (the shock) to create or encourage a behavior (the sit). Here is something I want to stress and I cannot stress it enough: positive reinforcement does not equal force free. You should have seen my face when my teacher told me this. My world view was shattered. The most common and popular example of positive reinforcement is to reward a dog for doing a desired behavior by giving them a treat, but it is not the only example. Use of an aversive can be positive punishment or positive reinforcement or negative reinforcement all at the same time, depending on how you are using it. Aversives are a whole different terminological league.Â
Negative Punishment: We are removing something, to stop or discourage a behavior, so an example of this is if I am playing with my dog and she is being too rough, so I leave the room or stop playing with her until she plays nice. Every time she plays too rough, I stop, until she eventually stops being so rough. Another example is if I am teaching her to not pull. She wants to keep walking because she wants to sniff that very cool looking butterfly, but I stop walking because she is pulling and I want her to knock it off. In this moment, she wants nothing more than to keep walking, and I am punishing her by not allowing the walk to continue. Once she stops pulling, we keep walking, which is her reward (positive reinforcement).Â
The quadrants were never meant to describe a training style, but to explain the four basic ways we can influence canine behavior during training.R+ training has become an accepted term to describe non-aversive trainers, and many popular and famous dog trainers use it in that way, but itâs technically inaccurate. Positive does not mean âgoodâ in the colloquial sense. It simply means âto addâ. And even a âpositiveâ tool can be an aversive, depending on the dog.Â
On a final note, no training is ever âpurely positiveâ in a (admittedly extremely) technical sense. We all use negative punishment at some point or another, and that is okay. In terms of dog training terminology, negative is a completely arbitrary word used to describe a course of action.Â
It is important to understand the quadrants in their entirety, but donât spend so much time worrying about whether youâre being a positive enough trainer that you neglect perfectly valid training methods to help your dog, or even condemn a method entirely simply because it has the word ânegativeâ or âpunishmentâ in it.Â