Let’s get operating
There’s been a lot of talk in recent years about operant (also known as instrumental) conditioning and it’s relevance to horse training. More specifically, about negative and positive reinforcement. Lots of discussion on how much of traditional horse training, European and American, and natural horsemanship is based predominantly on the use of negative reinforcement, and how those that use the ‘clicker’ focus on positive reinforcement. Lots of kneejerk reaction to the terms ‘negative’ and ‘positive’. Lots of confusion, because the end effect on behaviour is the same- reinforcement, or strengthening. Lots of articles written explaining what these terms mean, both in theory and in practice. Lots of division between trainers and justification for why they prefer one or the other as their focus in training.
A mechanistic approach
But for me, there is a big danger in focusing too much on operant conditioning. Yes, we most certainly need to understand it in order to train. BUT, focus too much on it, and training can become mechanical, treating the animal (or person) like a machine that is being programmed to follow instructions.
Why does that matter, you may ask? For many reasons (not least safety) we want our animals to be able to follow instructions. So do I…. But I don’t want to treat them like a machine. Like humans, there is growing scientific evidence that animals are sentient beings, that experience emotions just like ours.
We may not ever know what it is like ‘to be’ a horse, or a dog, or a fish, but the physiological evidence supports the idea that they can experience fear, joy, form attachments and grieve over losses. They share many of the same cognitive processes as we do. In fact, much of our understanding of human behaviour is based on animal studies. We are all animals!
A little detour through history
For a number of years now, I have been teaching my students and clients the importance of understanding classical (also known as Pavlovian, associative or respondent) conditioning. You may have heard the term ‘Pavlov’s dogs’ bandied about.
Pavlov was a Russian physiologist whose studies involved the collection of saliva from dogs. Like all good scientists, he was an observant man. He was aware that saliva production is a reflex action, not under conscious control. The reflex is triggered by the presentation of food. So in order to collect the saliva, the dogs were given food. Pavlov began to notice that the dogs started salivating BEFORE the food was produced, when the lab assistants appeared.
How could this be? It couldn’t just be because the dogs ‘knew’ that food was coming- salivation is a reflex, not under conscious control, that should only occur in response to one stimulus- food. Yet a new stimulus, lab assistants, was provoking the reflex response.
Pavlov had discovered that it is possible, through ‘association’ to pair the stimuli, so that a new, previously neutral, stimuli could lead to the response.
But what does that matter to us?!
Salivation isn’t the only reflex we have. There are all kinds, relating to self preservation and survival, including observable external reflexes such as the eye blink (to protect the eye from injury), choking, the knee jerk, and more interestingly for us, internal reflexes such as endocrine responses, such as would happen in response to rewarding or aversive experiences.
So this means that emotions are subject to the same possibilities as salivation. Emotional responses can become associated with stimuli that wouldn’t normally elicit them.
Time for some examples
Let’s imagine two horses, Misty and Thunder, and their owners, Bob and Jean. Bob and Jean are new to horses, have never seen horse training and have no preconceived ideas. They are each given a whistle, a bucket of carrots, and a lunge whip.
Bob takes Misty off into the paddock. He has this idea that he’d like to make Misty move, so he blows the whistle, but nothing happens. He blows it again, and decides this time to try cracking the lunge whip straight after. After just a few repetitions, Misty is cantering away from him as soon as he blows the whistle, and he doesn’t need to crack the lungewhip.
Meanwhile, Jean has taken Thunder off to a different paddock. She blows the whistle, with the idea that she’d like Thunder to move. Nothing happens. So she blows the whistle and picks up the bucket of carrots. After just a few repetitions, Thunder is cantering towards her as soon as she blows the whistle, before she can lift the carrots.
After a day or two, Bob finds he is having problems catching Misty, yet Thunder gallops to the gate as soon as he sees Jean coming.
In this example, the whistle is the neutral stimulus that initially means nothing to the horses. By ‘associating’ the whistle with either a stimulus that provokes the flight response (fear/ avoidance reflex) OR with a stimulus that provokes salivation and the positive affect (emotions) that accompany eating, the whistle has taken on the properties of those stimuli. So the whistle has the power to make one horse feel good while the other feels fearful.
The whistle in Jean and Thunder’s case, has been associated with an APPETITIVE stimulus. In Bob and Misty’s case it has been associated with an AVERSIVE stimulus. And by default, the sight of Bob makes Misty feel fearful and avoidant, while the opposite is true for Jean and Thunder.
Operant Conditioning, Appetitives and Aversives
In Operant conditioning, appetitives (let’s call them the ‘Goodies’) and aversives (let’s call them the ‘Baddies’) are used to reinforce or punish behaviour. Reinforcement is the increase in frequency or intensity of a response. Punishment is the decrease in frequency or intensity of a response. The four quadrants of operant conditioning are normally taught with the emphasis on whether or not it is positive or negative, reinforcement or punishment.
My argument is that we are better to teach with the emphasis on whether or not it involves AVERSIVES (baddies) or APPETITIVES (goodies), because, both from a scientific perspective and from my years of observations of horses in training situations, this is what REALLY matters to the horse.
Both positive and negative reinforcement result in an increase in the frequency or intensity of a response. From the perspective of behaviour, the end result is the same.
HOWEVER, positive reinforcement involves the horse experiencing a ‘goody’ (an appetitive stimulus).
While negative reinforcement involves the horse experiencing a ‘baddy’ (an aversive stimulus).
From an emotional perspective these are two very different experiences (as anyone who has ever played the training game at one of my clinics or talks will know!). Since the evidence suggests that horses have similar emotional responses to aversives and appetitives to us humans, then we would do well to think carefully about this.
Let’s bring classical conditioning back into the equation
So we now know that horses have an emotional response to both appetitives and aversives. We also know that, through classical conditioning, these responses can become associated with all kinds of innocent stimuli!
For example, in a training paradigm that is based predominantly on the use of negative reinforcement IN OTHER WORDS AVERSIVES, it isn’t just the tools (e.g. the lunge whip, the reins, the schooling whip, the spurs, the legs) that are aversive. By association, saddle and bridle or even the presence of the trainer themselves has the potential to become a conditioned stimuli, i.e. one that, by association, provokes the same response.
So the trainer may, in essence, make the horse feel uncomfortable just by their presence.
Many trainers are lucky to avoid this, because they also are the source of appetitive experiences for their horses, such as feed, water, scratches, treats, turnout with friends.
Here’s the crux of the matter: your relationship with your horse is, in good part, determined by the balance of appetitives and aversives that you use with them on a daily basis. Like a set of scales, tip the balance too far one way, and the nature of the relationship can change from a happy one to an unhappy one. This is all because of classical conditioning.
It may seem like a rather simplistic way to view it, and I appreciate that there are other processes involved. But we cannot afford to dismiss this.
Why does it matter if my horse associates me with aversives more than appetitives?
It matters because, by their nature, aversives create avoidance. If you are too heavily associated with them, then your horse will want to avoid you. Horses avoid mainly by flight, but, if flight isn’t an option, fight can appear. Eventually, you may run the risk of learned helplessness and a ‘shut down’ horse. This is not the groundings of a safe, secure relationship.
In training, we talk about ‘reinforcement history’. Reinforcement hostory refers to the weighting of goodies and baddies, I.e. aversives or appetitives associated with a stimulus or situation. Well, another word for reinforcement history could be ‘relationship’.
Whether you own the horse or not, whether you intend to keep them for life or sell them tomorrow, you have a responsibility, especially those of you working with youngsters.
You are creating that horse’s relationship with the human race as a whole. You can make it or break it. I always say my job, initially, when working with youngsters, is to help them fall in love with me. But my goal, is to help them fall in love with the human race.
A trusting, happy, gentle horse generally receives kinder treatment than one that is avoidant or defensive.
New evidence also suggests that the levels of fear, grief and distress experienced in childhood can have permanent damaging effects on the brain….. now that’s a whole other blog post!!