Deep Dive in the quadrants
When some people see the learning quadrants they think it’s too simplistic, too narrow, it’s too limiting to understand all that’s going on within a horse’s mind and their highly evolved social-communication skills. It can appear this way because the quadrants are so simple, but they’re not over-simplified, they’re all-encompassing. Everything in life falls within them, often many at once, often creating a wide variety of emotional impacts and behavioral changes.
In the simplified version we know of “Positive” as adding and “Negative” as subtracting, “Reinforcement” as the behavior happens more and “Punishment” as the behavior happened less. Should we look at an empty world, the horse and a single stimulus we can see which quadrant is in action.
So if the horse does a behavior and the stimulus is added – the behavior will either happen more or less in that scenario again. If the behavior happens more, it’s been Positively (because the stimulus was added) Reinforced (because the behavior happened more). If the behavior happens less we know it’s been Punished, and the stimulus was added, so Positive Punishment.
If the horse does a behavior and the stimulus is removed (negative) – the behavior will happen more or less in this scenario. If the behavior happens more then it was Negatively Reinforced. If it happens less it was Negatively Punished.
We also have the terms “Aversive” and “Appetitive”. Aversives are something the learner dislikes, something they work to avoid, something they would prefer not to be exposed to. An Appetitive is something the learner is willing to work for, something they seek out, want, desire, something they enjoy.
We know that:
If a stimulus is added and it reduces a behavior (positive punishment) the stimulus added MUST be aversive
If a stimulus is added and it increases the behavior (positive reinforcement) the stimulus added must be appetitive
If the stimulus is removed and it reduces the behavior (negative punishment) the stimulus removed must have been appetitive
If the stimulus is removed and it increases the behavior (negative reinforcement) the stimulus removed must have been aversive
We know these things by putting two and two together logically.
But in real life nothing is ever that simple. Even if we put a rat in a box and add and subtract 1 stimulus at a time, we have the simple fact that the rat is confined to a box interfering with the data. There is always multiple things in play at any given time. ALWAYS.
We have to consider the entire external environment – the pasture, the fences, the other horses, the human, the tool the human is holding, the sounds of traffic in the background, the wind blowing the rustling leaves, the smell of fresh cut grass… There is so much happening in any being’s life moment to moment. We also have their internal environment, this is their genetics, their personality, their hunger, thirst, energy level, their emotions, their social life, their learning history… All of these and many more, influence the horse’s behavior at any given time. It’s deeply complex. There is so much happening in the environment that we don’t control, but these are still influencing the learner’s behaviors through the quadrants.
Negative Reinforcement may happen when a horse goes into a shed to get away from the flies (going to the shed was reinforced by the removal of the aversive flies). A horse is frequently positively reinforced by sniffing around the ground for the best bites of grass or plants. This is happening in life all the time, many overlapping examples all the time.
That being said, in our formal, planned training we have the choice to decide what we add and subtract in the equation. We aren’t horses, horses don’t think we’re horses. Horses don’t train other horses to perform behaviors on cue. They certainly have deep and complex social lives and communication skills, but they aren’t training one another, the way we train horses. They interact with each other as I might a friend, I’m not trying to modify their behavior, I’m just hanging out with them. But if we want to actually modify their behavior, put behaviors on cue, cue them when we want them, and alter them for our convenience or ability to provide care, then we need to decide how we’re going to go about that. We need to be good listeners and observers of what they’re communicating to us, but we can choose how we communicate with them. All of the quadrants are about modifying behavior, all of them are coercive in the sense that we control them completely. There are ways to make both techniques less coercive, less controlling and more focused on choice and consent. But the entire point of training is to change their behavior.
We can communicate through aversive application and removal (negative reinforcement) or appetitive addition (positive reinforcement) to train behaviors. We could also use aversive addition (positive punishment) or appetitive removal (negative punishment) to reduce behaviors. We also have tools like training incompatible behaviors to reduce unwanted behaviors.
Our positive reinforcers can be anything appetitive, anything the horse likes. But food is the easiest and most convenient tool. We can control the amount and value of the food as desired. Using higher quantities of lower value food is more satiating and relaxing. Using smaller amounts of higher value food is more exciting and arousing.