How Horses Learn

If we want to train our horses as finely and Sportsfreundlich as possible, we should not only deal with how the anatomy and biomechanics of the horse work. We should also concern ourselves with how horses learn. Only then can we make our training as horse-friendly and motivating as possible.

Autorin: Karolina Kardel, 360 Grad Pferd

An important point that we should always be aware of when dealing with our horses: Learning does not start when we sit in the saddle. Learning starts when we enter the paddock!

The communication scientist and psychologist Paul Watzlawick (1921-2007) once said:
"You can't not communicate. "

And just as you cannot not communicate, you cannot not learn. Communication, after all, always goes hand in hand with behaviour and, accordingly, learning.

Often we leave this aspect out when we are with our horses. For many of us, training begins when we enter the oval track and ends when we get off. However, we often forget that our horses - just like us, by the way - are always learning. So it can easily happen that behaviours develop that we call bad habits and that we scold our horses for. But if we take a closer look, we have (unconsciously) taught our horse this undesirable behaviour with our behaviour and our reactions

When we deal with the learning behaviour of horses, we have the opportunity to question and optimise our own behaviour again and again. In this way we can better understand how (undesirable) behaviour arises, how we can prevent it and also how we can ensure more motivation in our joint training.

Three forms of learning

Basically, three forms of learning can be distinguished:
1. Learning through imitation
2. Habituation
3. Learning through consequences

Learning through imitation

In learning by imitation, a horse learns by copying a behaviour from other horses. A foal, for example, learns from its mother. Horses within a herd learn from each other, for example how to use an automatic drinking trough.
Horses not only learn behaviour through observation and imitation. Mood can also be transferred, especially fear and stress. If a horse in a herd raises its head in alarm, everyone is in "beware" mode. This is an instinct that is essential for horses to survive.

We should always be aware of the aspect of mood transmission in training. If we are afraid, this is often transferred to our horse. If we ourselves have fun and enjoy training and give motivating praise, our horse's motivation will also increase. Conversely, the horse's stress level increases when we are also tense and annoyed. (We have already reported extensively on stress in horses here.)

Habituation

Learning by habituation means that a horse is repeatedly exposed to a stimulus that has no consequences. The horse learns that it does not have to react to this stimulus and can remain calm.

Habituation is a lifelong learning process that is very important for horses. For example, a young horse that has no experience with road traffic may react to a passing car with fear, stress and uncertainty. However, the more often this horse is passed by cars and experiences that the stimulus basically has nothing to do with it, the less it reacts to it. This learning process is especially important for the flight animal horse, which uses its energy consciously and sparingly.

Another example of a classic habituation is spraying: The horse learns the hissing sound and learns that spray (diffuse stimuli) then hits the body. It learns that nothing else will happen to it and that it is not worth spending energy unnecessarily. Note: Some horses have problems processing the diffuse stimuli, they expend a lot of energy and react with strong defensive behaviour. In this case, the horse should rather be supported in processing the diffuse tactile stimuli.

Learning through consequences

Learning through consequences is the most important part of horse training. In learning by consequences, different learning theories are distinguished, which we would like to explain to you in more detail in the following.

Classical conditioning

In classical conditioning, two stimuli are linked together. For example, when a horse hears its food bowl rattling, it knows that there is something to eat. A basically neutral sound is associated with a (positive) expectation. Conversely, the mere sight of the saddle can also cause stress if the horse associates it with pain, for example.

The best-known example of classical conditioning is Pavlov's dog. In 1905, the Russian researcher Ivan Pavlov found out in an experiment that dogs begin to salivate at the sound of a bell tone alone, if this bell tone was heard repeatedly shortly before food was given. The dogs learned that the acoustic stimulus was followed by food (source: Wikipedia).

Operant conditioning

Operant conditioning is learning by success. The name "operant" comes from the fact that in this form of living the horse actively communicates and interacts with its environment and learns through consequences.

Positive and negative reinforcement

Reward (= something pleasant) and punishment (= something unpleasant) can either be added (+/positive) or removed (-/negative). This creates four learning quadrants:

Something positive (e.g. food) is added -> The behaviour is shown more often. : Something unpleasant (e.g. driving help) is removed -> The behaviour is shown more often. : Something unpleasant (e.g. whip) is added -> The behaviour is shown less often. Something pleasant (e.g. food) is removed -> The behaviour is shown less often. When we talk about positive or negative reinforcement, this has nothing to do with good or bad in the sense of our human values. They are merely mathematical quantities.

Classic training systems that can be classified here are clicker training and conventional horse training. Clicker training uses positive reinforcement and negative punishment (food is added or taken away).

In conventional horse training (ranging from Natural Horsemanship to Academic Horsemanship), the pressure-and-release method is used. This means building up and releasing pressure. When we lunge our horses, for example, we drive them with body, voice and whip into a faster gait and as soon as we get the appropriate reaction, we release the pressure. When riding, we also drive with thigh and crop and stop when our horse reacts accordingly.

So pressure and negative reinforcement are not bad. It is simply a different form of training than reward and positive reinforcement. And no matter which training system you use to work your horse: It is always important to have a well thought-out training structure, exact timing and correct execution of one's own aids. Otherwise we have a stressed horse that cannot learn. We want to bring you closer to the topic with some practical examples.

Conventional horse training: negative reinforcement and positive punishment

Classic horse training, which I am sure most of us know and use, relies on the learning concept of negative reinforcement.

You want your horse to make a forehand turn. To do this, you touch him on the side with the whip. This (light) touch is an unpleasant stimulus that the horse avoids. As soon as your horse reacts correctly and steps sideways with the hindquarters, the unpleasant stimulus subsides. Your horse learns that the unpleasant stimulus ends when he steps sideways and will show this behaviour more often in the future. At some point, even a whip pointing towards the hindquarters will be enough and your horse will make a forehand turn.

So negative reinforcement in horse training is not a bad thing at all, as long as you give your signals as finely as possible and set up the training so that your horse has a chance to understand what you want him to do.

The great danger in training with negative reinforcement is that a horse is forced to react. This covers up the fact that the training set-up is not optimal and that the horse cannot perform the task as desired (physically or mentally). Because what happens when a horse does not react to the crop impulse as desired? Often the pressure is then successively increased until the horse reacts accordingly. If the horse does not react as desired and stops or moves forward or backward, in the worst case the unpleasant stimulus (touching) becomes a positive punishment: the touching becomes a beating combined with a loud scolding: "At some point the stupid horse has to move its feet. "

The horse learns: If it swerves forwards or backwards instead of stepping sideways with the hindquarters, it gets into trouble. So it shows the undesired behaviour (standing still, swerving forward) less.

Positive reinforcement: more than just clicker training

In most cases, positive reinforcement is immediately associated with clicker training. But positive reinforcement means nothing else than: A desired behaviour is followed by a pleasant consequence. The horse notices that a certain behaviour is worthwhile and will probably show it more often in the future. You signal to your horse that it has behaved correctly with a feedback that is pleasant for the horse. Pleasant feedback does not necessarily have to be a treat. Pleasant feedback can also be an extensive crawl session. It is important that the horse really enjoys the petting session and finds it pleasant.

The reason for combining positive reinforcement with clicker training is as follows: behavioural researchers have found out that the time span between behaviour and reward must not exceed three seconds so that a horse can still establish the link between praise and behaviour. If this time span is exceeded, there is a danger that the horse will link the reward with something else. If, for example, your horse paws between the behaviour worthy of reward and the digging out of the treat from the jacket pocket, it will probably link the pawing with the reward and paw more often in the future. It has learned that it is worthwhile to scratch because there is a tasty reward. More at 360 Grad Pferd.

This is where the clicker comes in: it marks the correct behaviour with its acoustic signal and the horse knows that it gets the reward for behaviour X and not for behaviour Y - provided it has been conditioned accordingly beforehand.

In contrast to training with negative reinforcement, where your horse tends to react, training with positive reinforcement puts the horse in a more active role. In a forehand turn with positive reinforcement, the horse would not avoid the whip with its hindquarters, but would, for example, actively move towards a target with its hindquarters. (Explanation: A target is a target object that has to be touched).

Of course, when training with positive reinforcement, horses do not only show desired behaviour. If, for example, your horse paws with its hoof, this behaviour is not punished by scolding it or pulling on the rope, but is simply ignored by, for example, turning your back on your horse and ignoring it. This way, the horse cannot earn a reward and learns that it is not worth spending energy on pawing.

One exercise, two training approaches: the Spanish greeting

A good example to show you the differences between the two training systems is the Spanish Salute or the Spanish Walk. In the Spanish walk, your horse alternately lifts its front legs far and high and goes forward. In the Spanish Salute, the horse also lifts its front legs high, but remains standing. The Spanish walk is usually developed from the Spanish salute.

We describe the exercise structure once trained conventionally and once trained with positive reinforcement. We are not trying to give a rating of good or bad (see above, positive and negative are only mathematical values), we just want to show you the different approaches.

Option 1: Conventional training
If you want to work on the Spanish step or the Spanish greeting conventionally, it usually works like this: You touch your horse's front leg with the crop until he lifts his leg. Then you stop.

With sensitive horses, this approach usually works quite well and they lift their leg in response to a light touch. With less sensitive horses, however, it can happen that a light touch quickly turns into a strong knock and a "now lift your leg".

The long-term goal is that at some point you will no longer have to touch with the crop, but that a finger point is sufficient for your horse to lift his leg.

Option 2: training with positive reinforcement

When training with positive reinforcement, a target is used in addition to the clicker and food praise. The horse should touch the target with its front tarsal joint. The target can be a fly swatter, a pool noodle or something similar.

Here, the training structure can look like this:
Touch your target to the horse's tarsus and click. Repeat this a few times. Then hold the target a little away from the front leg and as soon as your horse moves his leg towards the target, click again. This distance is increased bit by bit. Every time your horse touches the target with its front leg, you give a click and praise.

Conclusion

It doesn't matter whether you train your horse with positive reinforcement or conventionally with negative reinforcement: It is important that you always make the training Sportsfreundlich and stress-free as possible. Because horses are only able to learn in a relaxed atmosphere. If you make your training as small-step as possible, you give your horse the opportunity to understand what it is supposed to do. This way you avoid frustration and stress on all sides.