Ethological challenges for the working horse and the limitations of ethological solutions in training
- McGreevy1, AN. McLean2
1Faculty of Veterinary Science (B19), University of Sydney, NSW 2006, Australia.
2Australian Equine Behaviour Centre, Clonbinane Road, Broadford, VIC 3658, Australia.
By definition, ethology is primarily the scientific study of animal behaviour, especially as it occurs in a natural environment; applied ethology being the study of animal behaviour in the human domain.
The terms equine ethology and ethological training are becoming commonplace in the equestrian domain, yet they seem to be used with a conspicuous lack of clarity and with no mention of learning theory. Most of what we do to train horses runs counter to their innate preferences. This paper summarises the ethological challenges encountered by working horses and considers the merits and limitations of ethological solutions. It also questions the utility of terms such as dominance, submission, alpha and leader.
The terms equine ethology and ethological training are becoming increasingly common in the equestrian domain and yet they seem to be used with a conspicuous lack of clarity. Among the more outrageous examples are recent attempts to establish Equine Ethology as a brand of training system. This is rather like a paracetamol manufacturer calling itself ‘Pharmacology’ or a religious sect dubbing itself ‘Theology’. The term ethologically sound training has appeared in the scientific literature but without an adequate definition or an explanation of what constitutes ethologically flawed training. Effective and humane training always takes account of the animal’s ethology, so there appears little need for the development of a distinct subclass of training.
It is time to set out the correct meaning of these terms and also to identify human-horse interactions that are especially reflective of ethological principles. This paper explores the ethological challenges encountered by working horses and considers the merits and limitations of solutions that bypass associative and non-associative learning.
Ethology is primarily the study of animal behaviour in a natural environment. It can help us understand how animals respond to environments in which they have not evolved e.g. the human domain. More accurately, it is the study of animal behaviour in the environment in which natural selection acted to shape that behaviour. Given that all aspects of behaviour are subject to natural selection, ethology is not merely the study of innate behaviours but also the study of how natural selection influenced learning strategies and capabilities. Natural selection will, for example, have influenced whether or not an animal learns well individually, or learns by observing conspecifics, or both. It will have influenced such variables as relative attention devoted to learning new food finding techniques versus scanning for predators.
Thus, equine ethology informs not only communication but equine behavioural needs and preferences, learning, value systems and motivation. It helps us to predict some of the ways horses out of their natural environment (i.e., in the domestic context) might react and cope with various challenges and how behaviourally flexible they may be. As such it underpins enlightened and effective training but is not a training system or philosophy per se.
Ethological challenges include interventions that cause horses both social frustration and environmental distress. Examples of social challenges include leaving the social group, taking the lead in the company of established leaders, being close to aggressive conspecifics, walking abreast rather than trekking in a line and ignoring displays by other horses. Enforced proximity to conspecifics can cause one horse to tread on another in ways that seldom occur in the free-ranging state. It can also be hazardous because it limits vision, such as when horses are clustered during steeplechasing. Furthermore, as jockeys well know, when galloping horses are too closely spaced, they may be prompted by their fellows to jump when they are not close enough to the obstacle to clear it safely.
Even when riding alone we may demand responses that naturally arise only in social contexts, contexts far removed from the manège. The collection and elevation required in higher levels of dressage, for instance, may be appropriate when horses greet one another but are ethologically redundant in isolation.
Examples of environmental challenges include leaving the home range, deviating from an obvious track and traversing, rather than avoiding, obstacles. Other examples of the ways in which equitation provides environmental challenges to horses that run counter to their ethology appear in the table below:
Riding brings both social and environmental challenges and is a useful example of the way we overcome horses’ innate responses and thus ignore their preferences. For example, free-ranging horses rarely maintain a fixed postural outline while changing gait. The current debate surrounding hyperflexion (Rollkür) has helped to highlight the extent to which riders can enforce a sustained, abnormal manipulation of a horse’s posture and sometimes gain a competitive advantage as a result.
Responses to physical discomfort under saddle generally have more to do with physiology than ethology. Here, the most obvious sources of physical discomfort are the bit, the rider’s leg/spur, the whip, the ‘carrot-stick’ and the girth. This is important because there is an assumption that the relationship a human has with a horse on the ground is identical to the relationship when he is mounted. It is by no means certain that horses connect pressure in the mouth with the rider. They have not evolved to expect that another animal can apply pressure to the inside of the buccal cavity via a piece of metal. This may account for the apparent forgiveness horses show when allowing heavy-handed riders to mount them time after time. It is therefore unnecessary and inappropriate to complicate a rider’s interventions by giving them anthropomorphic labels such as asking (e.g., asking the horse to lower his head), encouraging (e.g., using the inside leg to encourage forward movement) and supporting (e.g., outside rein to support the impulsion). The problem with the use of an anthropomorphic framework to explain rider-horse interactions is that it can justify abuse of horses that offer undesirable responses. So most horses benefit when science provides mechanistic explanations of equitation, even though some horse-lovers argue that this is undermining the bond they share with their horses.
There is an appealing notion that we can apply equine social strategies to human-horse interactions. But data and scientific rigour are lacking in this domain. In the midst of social conflict among horses, it is often the appeasement signals that determine the outcome. These may be very subtle. With their excellent vision, horses are able to detect minute cues from animals (and not just horses) around them. It seems likely that most human signals are not necessarily interpreted as homologues of equine signals (Roberts and Browning, 1998). How crude are the signals from a human to an equine observer? With no tail, fixed ears, a short, inflexible neck and only two legs we can hardly expect horses to regard us as equine. The chance that we can mimic equine signalling with any subtlety seems remote. Perhaps this is partly why humans rarely claim an ability to issue appeasement signals to horses and why agonistic advances (that may facilitate domination) prevail. Humans who fall into the trap of assuming that they can speak the language of horses with eloquence may fail to recognise the aversiveness of some of their behaviour and so put horses under inappropriate pressure. Ultimately, however, any search for equine analogues of human interactions with a horse becomes virtually irrelevant when a human gets on the horse’s back.
If, when handling horses on the ground, we are to correctly exploit their innate social organisation in training and handling them, the distinction between so-called dominance characteristics and leadership is critical. It is important to recognise that in the equestrian context, the imposition of so-called dominance manifests as the application and withdrawal of aversive stimuli and therefore cannot be considered outside the framework of learning theory. The notion of dominance evokes further thoughts of submission and respect. The term ‘submission’ is of particular importance because it occurs on every dressage test score sheet. If dressage is all about training rather than the imposition of human value systems, we should celebrate compliance rather than submission.
A common example of the sloppy use of the term ‘dominance’ arises when owners leading a horse find that it walks crookedly towards the handler and label this as evidence of the horse being dominant. The implication is that the horse doesn’t respect the handler as a leader and is actually invading the handler’s space. When we boil away the anthropomorphic interpretations, we find that if the horse leads crookedly, the best solution is for the trainer to condition it to lead in a straight line. There is no need to interpret the behaviour in anthropological or ethological terms.
As it happens, there is growing distaste for the term ‘alpha’, since this implies domination and permanency. This trend is also found in dog-training circles, and has led to a preference for the notion of leadership. The concept of humans as leaders of horses has subsequently gained currency in equestrian contexts but brings along its own set of problems. It implies that if a horse ‘respects’ a human as leader and has subsequently bonded to him or her, it will follow that human even when conspecifics are present. Leadership implies also that such bonded horses would follow humans to aversive places away from the sanctuary of conspecifics. There is no evidence of these phenomena occurring. Furthermore, those who subscribe to the notion of leadership do not explain how leadership qualities can be identified or developed.
It has been suggested that humans can enter the social hierarchy of groups of horses by mimicking their behaviour, most notably through their signals (Sighieri et al., 2003; Roberts, 1997). Based on the debatable premise that a herd is organised according to a dominance hierarchy established by means of ritualised conflict, this approach has grown in popularity but embodies some muddy thinking.
Let’s consider round-pen training as an example. The undoubted merits of this type of hands-off training are purported to be that it is humane and brings little risk of learned helplessness. The chief appeal of this approach lies in the notion that it is possible to manage unhandled horses without coercion by mimicking their behaviour patterns. But round-pen training does involve coercion. For unhandled horses, being stared at, approached and touched seem to lie on the same continuum of aversive interactions as being whipped. Furthermore, rewards in round-pens often take the form of rubbing, typically on the forehead – a rather unethological response given that allogrooming of the forehead by conspecifics is far less likely to occur than wither-scratching, which has been shown to lower the heart-rate (Feh and de Mazières, 1993; McBride et al., 2004). Round-pen training presents ideological contradictions because it is supposed to create trust and partnership. But chasing a fearful horse in a round-pen is an extremely aversive interaction that, in some circumstances (with fearful horses, for example), can be both ineffective and inhumane. If used to condition horses into constant states of hyper-reactivity, it can precipitate chronic stress. Sadly, it is in danger of becoming a legitimised and institutionalised form of violence. An emerging syndrome in equine behavioural medicine seems to reflect chronic stress in horses that have been round-penned with alarming regularity.
A raft of questions is launched by the philosophy of human-as-leader: What if, despite embracing the notion wholeheartedly, handlers cannot persuade their horses to comply? If the horse fails to follow the handler into a trailer, does that simply mean the human was perceived as a poor leader? What particular aspect of the human’s leadership was lacking? How can this be studied scientifically? Are there negative welfare implications for the horse that doesn’t recognise any human as leader? If a third party leads the horse on using negative reinforcement is he/she showing subliminal leadership?
One would expect horses that have bonded in this way and genuinely regard certain humans as leaders to seek out the company of the human leaders and forsake their conspecific affiliates. However, there is insufficient evidence that horses in a paddock approach humans for reasons other than mere curiosity or because it is a conditioned response.
Horses will always find conspecifics more salient than humans as leaders. Analogues drawn between human-horse interactions and elements of the equine ethogram can be tenuous. For example, it is suggested that simply being behind a horse and driving it forward (as in long-reining) is directly analogous to the herding behaviour of stallions (Zeitler-Feicht, 2004). This assumption is very difficult to test. Perhaps we should simply accept that we are, at best, care-givers and companions, and, when we are not giving care and companionship, we are trainers. Conspecifics, including dams, can condition members of their social group but whether training is their intention is clearly debatable. Although there is some obvious overlap between care-giving, companionship and training, there seems sense in compartmentalising them. To do so helps us approach each set of activities with clear expectations.
It has been suggested that a trainer’s interactions with horses should be based on three elements fundamental to the equilibrium of the herd: flight, herd instinct and hierarchy (Sighieri et al., 2003). However, this approach overlooks the importance of foraging, coalitions, kinship and affiliation as well as the reality of the effects of conditioning on all innate responses. Ethologically sound solutions should not depend on a notion of the horse’s benevolence – that the horse is ‘wanting to be with’ or ‘wanting to please’ the trainer. This is an ideologically unsound mindset owing to its anthropocentricity.
The importance of habituation, sensitisation, operant and Pavlovian conditioning should never be under-estimated because they underpin training techniques. They are informed by learning theory, and supported by ethology. Although most training systems use a blend of all four of these processes, there are fundamental gaps in the understanding and acceptance of their place in equestrian coaching (Warren-Smith and McGreevy, 2006). Studying equine ethology demands consideration of how natural selection shaped horse behaviour and horse learning capacity. Training philosophies that embrace learning theory can be ethological in the sense that they might take into account the types of stimuli horses are most likely to respond to and the types of reinforcer that are most rewarding (from knowledge of ethology).
Instinctive responses predicted by ethology can facilitate horse handling without the need for learning. For example, allelomimetic behaviour, mimicry, stimulus enhancement and social facilitation (McGreevy et al., 2005) are all mechanisms for changing behaviour without conditioning. However, these are adaptive mechanisms that evolved for group cohesion and they can and do act upon behaviours that may have been subject to conditioning themselves.
Ethologically based training solutions can capitalise on leadership by conspecifics and even the effects of psychopharmaceuticals (including pheromones). Dressing up training systems as being forms of ethology denies the importance of learning theory and implies that we must speak the language of horse. This may be beguiling but it is ultimately an illusion.
The illusions of horse-owners are generally harmless as long as they do not create unrealistic expectations. Learning theory can and should be used to explain all training techniques no matter how elaborately they are camouflaged. For example, advance-and-retreat (Blackshaw et al., 1983) is as much based on negative reinforcement as the physical pressure-release systems used in the ridden horse (McLean and McGreevy, 2005). Equitation science uses learning theory to demystify and simplify training and, although still in its infancy, it is already beginning to show how horse wastage can be reduced and welfare enhanced.
More rigour is needed in using the terms ethology and ethological in the equestrian domain. An understanding of equine ethology enhances effective horse handling but cannot explain the majority of the responses made by working horses. There is danger in confusing dominance and leadership with conditioning. Acknowledging the role of learning theory in human-horse interactions is the preferred means of avoiding over-interpretation of equine responses to humans.
Blackshaw, J.K., Kirk, D., Creiger, S.E. (1983) A different approach to horse handling, based on the Jeffrey method. International Journal for the Study of Animal Problems. 4 (2) 117-123.
Feh, C. and de Mazières, J. (1993) Grooming at a preferred site reduces heart rate in horses. Anim. Behav. 46, 1191-1194.
McBride, S.D., Hemmings, A. and Robinson, K. (2004) A preliminary study on the effects of massage to reduce stress in the horse. J. Equine Vet. Sci. 24, 76-81.
McGreevy, P., McLean, A. (2005) Behavioural problems with the ridden horse. In: The Domestic Horse. The Origins, Development, and Management of its Behaviour. Eds: Mills, D.S. and McDonnell, S.M., Cambridge University Press, Cambridge. 196-211.
McGreevy P.D., McLean A.N., Warren-Smith A.K., Waran N. and Goodwin D. (2005) Defining the terms and processes associated with equitation. Proceedings of the 1st International Equitation Science Symposium (2005), Eds: P. McGreevy, A. McLean, N. Waran, D. Goodwin, A. Warren-Smith. Broadford, Victoria, Post-Graduate Foundation in Veterinary Science, Sydney, 10-43.
Roberts, J.M., Browning, B.A. (1998) Proximity and threats in highland ponies. Social Networks 20(3):227–238.
Roberts, M. (1997) The man who listens to horses. London: Arrow Books.
Sighieri, C., Tedeschi, D., De Andreis, C., Petri L. and Baragli, P. (2003) Behaviour patterns of horses can be used to establish a dominant-subordinate relationship between man and horse. Proceedings of the 2nd International Workshop on the Assessment of Animal Welfare at Farm and Group Level, University of Bristol, UK, 4-6 September 2002. Animal Welfare 12 (4): 705-708.
Waring, G.H., Dark, G.S. (1978) Expressive movements of horses. Paper presented at Animal Behavior Society Annual Meeting, University of Washington, Seattle.
Warren-Smith, A.K., McGreevy, P.D., (2006) An audit of the application of the principles of equitation science by qualified equestrian instructors in Australia. Proceedings of the 2nd International Equitation Science Symposium (2006), Milan. Eds. M. Minero, E. Canali, A. Warren-Smith, A. McLean, D. Goodwin, Mari Zetterqvist, N. Waran, P. McGreevy.
Zeitler-Feicht, M. (2004) Horse Behaviour Explained. Origins, Treatment and Prevention of Problems. Manson Publishing, London.