Learning theory is the term given to the processes by which animals and humans learn. There are a variety of different theories but the one that is best applied to training animals is behavioural theory, a conceptual framework that lays out how behaviours are acquired, or learned, through conditioning.
Conditioning falls into two categories, classical and operant. In classical conditioning the horse learns a simple association between two events (think Pavlov’s dogs or a horse getting excited by the sound of the feed room door opening) and the behaviour becomes a reflex response to a certain stimulus. In operant conditioning the horse chooses behaviours that are reinforced either positively or negatively. The horse learns to make associations with repeated words or actions through negative reinforcement, in which something undesired or uncomfortable is removed such as leg or bit pressure, or positive reinforcement where something good is anticipated such as a treat which can be used in methods such as clicker training.
The process of ‘shaping’ is used to build on these responses and refine them in order to make a horse respond to the subtlest cue.
Despite over a hundred years of the study of animal behaviour (ethology) and the psychology surrounding the science of learning, the equine industry still has a way to go in promoting best practice in horse training. Unlike the canine industry, where knowledge is generally passed down with success from professionals to the general population, there are countless horses who are suffering from confused training methods and many riders frustrated as a result of misunderstanding.
Although there are probably many examples of poor dog training, the rudiments of how learning theory is applied to dogs is familiar to most owners who will understand the thinking behind teaching a dog to sit or stay. It would be a much better world for the domesticated horse if the same situation were true in the equine industry. If every rider understood the principles of teaching a horse to respond to their cues without confusion, fear or conflict, both parties would be so much better off.
In 2007, a group of equine professionals came together with the goal of facilitating research into the training of horses to enhance welfare and improve the horse / rider relationship. The International Society for Equitation Science (ISES) was founded in 2007 and now unites a multi disciplinary membership of academics, students and interested practitioners worldwide. It promotes eight fundamental principles of learning theory in equitation, with the ambition that every rider and horse owner base their training interactions around them.
Training Principles of Learning Theory The following eight principles were originally defined in the peer reviewed paper by McGreevy and McLean in 2007 (The roles of learning theory and ethology in equitation,’ published in the Journal of Veterinary Behavior: Clinical Applications and Research, Volume 2, 108-118). The application of the following eight principles is not restricted to any single method of horse training. There are many possible systems of optimal horse training that adhere to all of these principles.
1 Understand and use learning theory appropriately Learning theory explains positive and negative reinforcement and how they work in establishing habitual responses to light, clear signals. The terms ‘positive’ and ‘negative’ when applied to reinforcement are not value judgements (as in ‘good’ or ‘bad’) but arithmetical descriptions of whether the behaviour is reinforced by having something added or something taken away, such as pressure. For example, when the horse responds to a turn signal and the rein pressure is immediately released, negative reinforcement has been applied.
It is critical in the training context that the horse’s responses are correctly reinforced and that the animal is not subjected to continuous or relentless pressure. Prompt and correct reinforcement makes it more likely that the horse will respond in the same way in future. Learning theory explains how classical conditioning and habituation can be correctly used in horse training.
2 To avoid confusion, train signals that are easy to discriminate There are many responses required in horse training systems but only a limited number of areas on the horse’s body to which unique signals can be delivered. From the horse’s viewpoint overlapping signal sites can be very confusing, so it is essential that signals are applied consistently in areas that are as isolated and separate from one another as possible.
3 Train and shape responses one at a time It is a prerequisite for effective learning that responses are trained one at a time. To do this each response must be broken down into its smallest possible components and then put together in a process called ‘shaping.’
4 Train only one response per signal To avoid confusing the horse, it is essential that each signal elicits just one response. However, there is no problem with a particular response being elicited by more than one signal. Sometimes a response may be complex and consist of several trained elements. These should be shaped (or built up) progressively. For example, the ‘go forward’ response is expected to include an immediate reaction to a light signal, a consistent rhythm as the animal moves in a straight line and with a particular head carriage. Each of these components should be added progressively within the whole learned response to a ‘go forward’ signal.
5 For a habit to form effectively, a learned response must be an exact copy of the ones before For clarity, a complete sequence of responses must be offered by the horse within a consistent structure (e.g. transitions should be made within a defined number of footfalls). Habit formation applies to transitions in which the number of footfalls must be the same for each transition and this must be learned.
6 Train persistence of responses (self-carriage) It is a fundamental characteristic of ethical training systems that once each response is elicited, the animal should maintain the behaviour. The horse should not be subjected to continuing signals from leg or rein pressure.
7 Avoid and dissociate flight responses When animals experience fear, all characteristics of the environment at the time (including any humans present) may become associated with the fear. It is well known that fear responses do not fade as other responses do and that fearful animals tend not to trial new learned responses. It is essential to avoid causing fear during training.
8 Benchmark relaxation to ensure the absence of conflict Relaxation during training must be a top priority. When conflict behaviours are observed in the horse, the training method must be carefully examined and modified so that these behaviours are minimised and ultimately avoided. To recognise the importance of calmness in enabling effective learning and ethical training, any restraining equipment (such as nosebands) should be loose enough to allow conflict behaviours to be recognised and dealt with as they emerge.
1. Immediate release - soften the pressure of the signal the instant the horse responds appropriately. 2. Differentiation - signals must be clearly distinguishable from each other. 3. One step at a time (shaping) - train each response component of complex movements separately. 4. Repeat with consistency - the horse will automatically respond in the desired way if the behaviour is precisely targeted. 5. One response per signal - reinforce only one response for each separate signal. 6. Avoid fear - during all interactions, make sure that characteristics of the environment, including the humans, do not become associated with fear. 7. Train persistence - reward the horse for maintaining a behaviour by not applying pressure until the next signal is given. 8. Check for relaxation - strive for relaxation when training each response. Techniques and equipment must not be used to mask distress or undesirable behaviour.
Find out more about the International Society for Equitation Science visit www.equitationscience.com