HOW HORSES LEARN: PART I

©Lesley Skipper 2003

How do horses learn? All mammals (including, of course, humans) learn in pretty much the same ways. The strengths of some species may lie in different areas of learning than others, but essentially the processes are the same.  Different modes of learning have been identified, all of which are important to everyone concerned with the care and training of horses. Although it is not possible to do the subject complete justice in the space of one or even two articles, I want to try to clarify some of those points which frequently lead to misunderstanding.[1]

Imprinting

This is nature's way of ensuring that a duck knows he is a duck, a goat that she is a goat, and a horse that he is a horse.  The image of the mother is 'imprinted' on the baby's brain, so that it knows who she is and will not try to follow either another of the same species, or an animal of a different species, which in some cases could be fatal.

Silent Learning

The foal, having learned who his mother is and how to move around without falling over, must then learn all kinds of things about his environment. We absorb all kinds of information about our surroundings and the people we associate with; so do horses, who are particularly good at this kind of learning.

Insight Learning

Humans are great solvers of the kind of problems that require insight, i.e. the sudden realisation of the answer to the problem. Horses in the wild would have little need to tackle this kind of problem, so in general they are not so good at this, although some of them show flashes of insight when learning new manoeuvres.

Habituation

Horses, like other animals, have to learn what elements in their environment are worthy of their fear, and what may be safely ignored – in other words, they grow accustomed, or habituated to things which would otherwise cause constant alarm.

One Trial Learning

One ability horses do possess in a high degree is that of ‘one-trial learning’: that is, they can learn after only one experience.  This can work for us or against us.  The horse can learn bad things just as easily as he can learn desirable things, and one careless error may affect a horse, in some cases for the rest of his life, in others, certainly for a very long time.

Social Learning

Clearly horses learn from their environment, but do they learn from each other?  There is a great deal of controversy about whether horses do learn by direct observation; relatively few studies have been made of this type of learning, and it is highly possible that these were not designed well enough in the first place. I suspect that motivation has a great deal to do with how well horses do in such experiments; it may be that they simply don’t see the point! One day experiments that are more 'user friendly' for the observer horses may well show more positive results, but at this stage we lack experimental proof one way or the other. On the other hand there is ample anecdotal evidence that horses do learn from each other, as well as from humans. Whether horses really do copy bad habits from each other, as is generally believed, is difficult to assess, since such habits can arise from a variety of causes. In general, though, it would be little short of miraculous if a social species like the horse did not have the ability to learn by observing others.  How else would they learn so much about what it is to be a horse?

Conditioning

Many people will have heard of the Russian physiologist I.P. Pavlov and his famous experiments with dogs. Pavlov trained the dogs to associate being fed with the ringing of a bell, and eventually the sound of the bell alone was sufficient to make the dogs salivate in anticipation.

This is known as 'classical conditioning', and it can occur naturally. For example, a horse who has happily gone past a gateway for years may have a frightening experience there – perhaps a dog rushes out barking, or a plastic bag in the hedge flaps and gives the horse a fright.  The horse subsequently refuses to approach that gateway, even though whatever frightened him may no longer be there; he has been 'conditioned' to fear it.  Humans too can be affected like this; they can develop phobias about certain places, sights, sounds or smells because they are associated with some unpleasant experience.  Understanding this helps to explain a great deal of equine behaviour.

The other kind of conditioning is known as 'operant conditioning', commonly used in laboratory experiments on learning.  In these, the 'operant' (a rat, a pigeon, a monkey or whatever) works either to earn a food reward or to avoid a punishment.  This type of learning works on the basis of stimulus-response.  Once a 'cue' or stimulus has elicited the desired response, a new stimulus or cue can be 'paired' with the old one by presenting first the new stimulus or cue, then the old one. Eventually the 'operant' will come to associate the 'new' stimulus or cue with the correct response. The stimulus (or cue) could be almost anything; in the case of a rat in the experimental box, a buzzer or a light, in the case of a ridden horse, the feel of the rider's leg.  The idea is that all the required responses have to be taught by means of conditioning. For many people, this forms the basis of horse training, and to some extent this is true. However, it is very far from being the whole story – as we shall see.

Disconnected brains?

Horses may appear not to connect something they have seen, say, on their left, with the same object when they pass it again on their right. Many people believe that the two halves of the equine brain are imperfectly connected, and that the two sides of a horse’s brain do not communicate effectively. So, we are told, we have to teach the horse everything twice, once for each side. In fact, there is no physiological reason why the two halves should not communicate just as well as they do in human brains. Indeed experiments carried out by Dr Evelyn Hanggi in the USA have shown that horses can certainly transfer information efficiently from one side of the brain to the other. The apparent disconnection between their left and right may have more to do with the way they see things than with anything else. Because their range of binocular vision is limited in comparison with ours, it may well be that they do not immediately make the connection between an object seen from one side, and later from the other (try looking at a reversed image of a familiar scene, and you too may experience an initial lack of recognition). Or it could be that they did not pay sufficient attention to it in the first place!

          In addition horses, like humans, are usually right- or left- ‘handed’: that is, they are suppler on one side than the other. So it is one-sidedness, not an imperfectly joined brain, that prevents them from performing equally well on both reins.

Mirror, mirror on the wall

Which is the brightest horse of all?

In general, I dislike questions like this, because they imply that there is some single quality called ‘intelligence’ which we can measure and use to compare one horse (or group of horses) with another. However, nobody has yet succeeded in defining ‘intelligence’ adequately. It is not a thing, but a concept encompassing many different aspects of dealing with the world. So if we can’t even decide what intelligence is, how can we measure it? What we can do is try to measure how quickly horses of one breed or type learn in comparison with horses of another breed or type. We have to be careful, though, because there are so many variables which can affect learning ability, such as ill health, stress, poor nutrition (especially in early life), management, training, individual temperament, etc.

           In one study, Arabians were far and away the quickest learners. This will come as no surprise to owners of Arabians, but although it can be a pleasure to train horses so quick in the uptake, it has its downside: Arabians are just as quick to learn the wrong thing as they are to learn the right thing. This means that extra care needs to be taken that you do not give the wrong aids, or signals, or inadvertently reward the horse for doing the wrong thing (more will be said about this in the next article).

           Horses of Iberian breeding are also well-known for their trainability and quickness to learn. This, together with their unflappable temperament, is one of the reasons why they are so much in demand as circus performers and stunt horses.

At the other end of the scale, some studies have shown that the least quick to learn were the Thoroughbreds. This is not necessarily because Thoroughbreds are less bright than other horses; it could instead be due to their generally higher level of reactivity, which may make them more easily distracted.

           I must emphasise that the comments above are generalisations. Horses are individuals, and no two horses, no matter what their breeding, will ever react in exactly the same manner. The point is that each horse learns in his own way, and attempts to compel all horses to learn in the same fashion, at the same pace, are doomed to failure.

In the next article I shall be looking at ways in which we make use of conditioned learning in training our horses, and the difference between positive and negative reinforcement. I leave you with a question:

What is the difference between a cue and an aid? Think about this, and you will find that the answer lies in the words themselves.

For more on this subject, go to How Horses Learn: Part II

Back to Xenophon home page

Last revised: January 18, 2004.