How can we integrate equine learning and psychology in the training of sport horses to enhance their welfare?
It has been demonstrated recently that horses can show depressive-like behaviours and learned helplessness as a consequence of improper management or training. Furthermore, it is demonstrated that mental health and well-being in athletes are essential to good performance. Many horses are thereby “lost” due to the onset of ‘behavioural problems’ that make them unfit to perform the task for which they are bred. Putting these facts together, the urge to consciously manage and train our horses, not only for their optimal welfare but also for optimal performance, is established. This blog post summarizes the actual scientific and practical knowledge of equine learning as guidelines to enhance the physical and psychological welfare of horse and trainer. We will start with explaining the different animal learning theories, i.e. the non-associative (habituation and sensitisation) and the associative (classical and operant conditioning) learning with practical examples, and finish with giving you some final conclusions and practical tips. Feel free to drop your questions or come up with more stories about examples of the animal learning theories in practice!
The non-associative learning principle refers to a relatively permanent change in a behavioural response to a single stimulus due to repeated exposure to that stimulus. Non-associative learning is widely used in training to desensitise horses (equals habituation) to decrease anxious or other undesirable behaviours on the one hand, or to do just the opposite, meaning to sensitise the horse to gain more alertness and bigger responses. It appears that horses either adopt habituation (reduced behavioural response) or sensitization (increased behavioural response) to repeated exposure of a novel stimulus, depending on how threatening their perception of the stimuli is. It is worth forming a ‘shaping plan*’ to gain the required final behaviour without evoking dangerous situations because of anxious or aggressive responses.
*A shaping plan refers to training of the horse performed by putting together the elements of the behaviour, starting from the horse’s natural responses and ending, through step-by-step actions, to a complex pattern of an essentially unnatural behaviour.
With repeated exposure, horses habituate to non-threatening stimuli. In order to effectively train a horse with habituation, stimuli must be offered equal and constant throughout the training plus stimuli must be applied sequentially following the step-by-step procedure to achieve the required final behaviour. If the stimulus changes shape, intensity, or presentation modality, the horse might react to it as it is a new stimulus and the process of habituation will thereby be invalidated.
With repeated exposure, horses sensitise to (in their perception) threatening stimuli. The threat could represent any kind of discomfort and is very dependent on the temperament of the horse. The application of sensitisation is a delicate case and can often be misused to increase responses or alertness. However, sensitisation can be very beneficial as well if used with right intentions and a long term plan. An example of these practices is to sensitise a (sound!) horse to riding aids, when it is trained before with ineffective riding aids which can lead on itself to subtle abuse or the self-learned helplessness phenomenon.
Associative learning is a process that allows the animal to establish the connection between two events. It can be subdivided into two categories, i.e. the classical conditioning and the operant conditioning. The first one of these two subcategories is well-known as the Pavlov-theory. With classical conditioning, animals learn which environmental cues predict future events so that they can behave accordingly. In such cases, the animal has no control over events; and the response is not under the control of the animal. The second one of the two associative learning categories represents the learning processes in which an individual’s voluntary behaviour is modified by its antecedents and consequences. In operant conditioning, a certain cue is given to make the animal operate in a specific way and it therefore allows the animal to associate two events over which it has control. In other words, this learning process is used to teach the animal specific behaviour and responses.
Classical conditioning: the Pavlov-experiment
The classical conditioning learning principle was first described by Pavlov based on an (ethical) experiment in dogs. Pavlov noted that when a neutral stimulus (the sound of a bell) was consistently associated with an unconditioned stimulus (the view and smell of food), the dogs soon responded to the originally neutral stimulus (the sound of the bell) as if it was the unconditioned stimulus (the view and smell of food). Well, if we transfer this experiment to the well-known phenomenon of our horses asking for their food as soon as the feeder comes in the stables in the morning, this principle is easy to get. Originally, a person coming in the stables is not per se linked with getting food. But if the feeder comes in first thing in the morning and starts feeding them immediately every morning, the horses will learn quickly enough that the feeder means food is coming.
Operant conditional: reinforcements and punishments
Operant conditional learning can be accomplished in four different ways, the one more beneficial and animal-friendly as the other.
A positive reinforcement is the addition of something pleasant, such as a treat, a scratch or verbally rewarding. A negative reinforcement is the subtraction of something unpleasant, such as a correctly used leg aid or pressure to move the horse sideways. The opposite definition counts for the punishment terms. Positive punishment is the addition of something unpleasant or painful, such as aggressively use of the whip, bit or spurs. Negative punishment is the subtraction of something pleasant, such as not being given a treat as reward for their behaviour.
It is established that positive reinforcement is more likely to make the horse display the target behaviour and show increased motivation. The negative reinforcement of incorrect behaviours should be avoided as this cultivates the incorrect behaviour, but can be used as reinforcement of correct behaviours. An example of correct use of the negative reinforcement is taking away the leg aid when the horse goes forwards. Punishment, both positive and negative, should in all times be avoided as it can lead to dangerous reactions and reduce spontaneous activity. These reactions are associated with decreased attention and learning, a negative perception of the environment and depressive-like disorders. Thereby, punishment causes damage to the physical and mental welfare of the horse, on short term and on long term. Moreover, it won’t produce the wanted behaviour and state of the horse anyhow, as the above mentioned consequences of punishment (depressive-like disorders, …) in sport/ working horses will degrade their performance.
Tips and tricks
Besides the different learning theories, it is important to take the following parameters into account as well: the duration, the intensity, the amount and the intention of used stimuli (e.g. pressure) and the time interval between the stimuli and the required behaviour and the amount of stimuli. To make it even more complicated, although evident, research has shown that all these parameters should be carefully adjusted to every horse at every moment.
For example, increasing aversive stimuli are likely to induce habituation, which leads to a reduction in response and ineffective management/ training strategies. In practice, this is seen in riders giving repetitively ineffective leg or rein aids. This phenomenon can potentially lead to subtle abuse of the animal or the well-known learned helplessness in horses.
Another example is the consequences of the amount of stimuli at the same time. Horses will choose their response themselves when too many stimuli at the same time are applied. Additionally, the time interval between the given aid and the required behaviour determines how the horse interprets the aid.
We thereby advise you to always be conscious about how you interact with your horse and which responses your horse gives you back. And besides research, you might have felt the magic yourself already when you feel like your horse and you are so tuned into each other that it feels like you can read each other minds and that all interactions go fluently without aggression or frustration. It is so worth it to spend some extra time to integrate these principles, cause your horse will appreciate you (even) more and more… These evidence-based guidelines are beneficial for both welfare and performance, both physically and mentally, for riders and trainers in all disciplines and at all levels. We thereby hope that you can take something with your from this post and try to bring these principles into your daily equine practices!