Articles

THE PLIGHT OF THE WORKING ELEPHANT

Elephant training throughout Asia is ancient and has its roots many millennia ago in Assam. Like the training of any ridden animal such as horses, donkeys and camels, elephant training techniques are entangled in folklore and legend, handed down from generation to generation. Interestingly, in all ridden species, the prevailing view of the recipe for successful human-animal interactions arises from mankind’s historical preference for hierarchical relationships since hunter-gatherer times – a fixed linear hierarchy where rank is maintained by ritualised aggression.   In other words mankind has embraced systems of dominance and submission hierarchies where one individual is made to respect the one above. This has been manifest in the military, in schools, business and even in family relationships.  Nowadays however, the notion of dominance is diminishing (except in the military of course) in favour of more dynamic and interactive relationships. However, in traditional horse and elephant training across Eurasia, the belief that success in training comes from dominance through ritualised aggression is still mainstream.  Central to this belief structure is the mindset that the animal already knows what it should do, but until you make it respect you, it will not be of service or ‘willing to please’. 

On the other hand, what humanity now knows about animal learning is relatively very recent, less than a century old. This body of knowledge is called ‘learning theory’ and began with the tradition of the behaviourists culminating the amazing work of B.F. Skinner whose theoretical contributions form the basis of training animals in zoos. This knowledge also explains the effects of the interaction between riders and their charges including elephants and their mahouts.  Basically animals are motivated to obtain certain goals, and on achieving them, their actions are rewarded and the stimulus response connection is reinforced by repetition. For instance, animals are motivated to remove even the smallest amount of annoying vibrating pressures on their bodies, and if any action on their part causes the pressures to go away, the animal repeats the reaction that worked. So in early learning the young elephant learns that tactile vibrations behind its ears disappear when it goes forward, it learns to go forward each time these stimuli occur, and thus the animal feels secure in the implicit knowledge that it has the control to diminish its irritations. Security is the antithesis of aggression. So the mahout should stop the pressures when the animal gives the correct response. Food is another motivation. If the animal learns that a certain reaction causes the delivery of food, it will repeat that action. These ideas are the basis of negative and positive reinforcement.

Furthermore if the animal learns that a certain event repeatedly precedes a known motivator (food or comfort), the event comes to predict that reward. This is how animals learn the various signals human trainers use to control them.  In elephant training, the problem is that some reactions  are complex (such as picking up a coin with the trunk), however this training feat can be achieved by setting up a training protocol where the task is broken down into related steps, where each step leads obviously to the next and is rewarded and consolidated by repetitions.  For example the elephant may be rewarded with a morsel of its favourite food for sniffing near the coin with its trunk, and when that is repeated by the elephant each time, the task is modified to the next step so that now the elephant is rewarded only for touching the coin and so forth. By this time the elephant is very keen on interacting with the coin and soon is rewarded only for picking it up. Then it is rewarded each time the coin is brought closer to the human. This successive approximating toward a goal is called shaping and is a vital tool in training.  Almost anything is possible in training if the steps are small and well thought out.

Using pressure-release and food morsels (which is gradually phased out) to motivate, reward and shape desirable responses and then installing cues or voice commands by association is the basis of the training system I have developed for the working elephant, and finetuned with the help of Laurie Pond, my knowledgeable and practical elephant training colleague with a lifetime of hands-on expertise.  With many repetitions the elephant learns to do what is asked of it by sheer force of habit. Animals are happy and secure as a result of an accumulation of solid habits as these give the important benefits of security, controllability and predictability. In the same way as humans develop habits (such as riding a bicycle or driving a car) the animal learns its training as second nature and it soon becomes so deeply embedded in its life that it knows nor desires any other way: such is the power of habits. We are no different, we have the power to change many things in our lives but we generally stick to our entrenched life patterns in work, rest and play.  This ‘learning theory’ approach to elephant training involves no violence, and so none of the problems associated with violence occur compared with traditional approaches. It just takes patience. But do not be fooled into thinking this method takes longer, because once begun, results come surprisingly quickly because the elephant is motivated – he is on your side. You might say he is now a willing partner.

The early training of elephants is traditionally based on breaking its spirit and making it submissive. It involves much pain and suffering which is inflicted until there is no resistance or reaction, over a period of many days. This is not only counterproductive to training an animal to respond to pressure signals, but also sets up some very serious problems that threaten the safety of mahouts and other personnel associated with  the elephants.  Traditional methods invariably involve punishment – pain delivered after an act, or after an incorrect response. Punishment for non-compliance is not only difficult for the elephant to fathom; it increases the probability of latent aggression. When animals are subjected to pain or fear, their brains are wired to seek to control that fear or pain by reacting, running away or fighting it. Occasional fear or pain followed by escape or attack has little negative effect as the animal learns ways of controlling the situation by fight or flight. However, when animals (or people) are subjected to repeated inescapable pain because they are tethered, imprisoned or trapped, (punishment with the ankus whilst ridden falls into this category) they learn that they do not have control over their environment. The result typically can lead to effects akin to chronic PTSD (Post Traumatic Stress Disorder) phenomena. 

Some of the cardinal features of chronic PTSD in humans are: hyperarousal symptoms, aggression, marked avoidance behaviours, and the re-experience of the symptoms of the trauma. In soldiers experiencing severe inescapable and prolonged trauma in war, post trauma reactions are frequently evident, and according to research, significantly higher in prisoners of war. PTSD may show up as sudden hyper-aggression and violence triggered perhaps by a mild fearful experience or a relatively benign situational trigger, such as the noise of a balloon bursting at a children’s birthday party or a conflicting emotional situation. Such events are not restricted to humans. Elephants show the same patterns. With an alarming similarity to human cases where severe PTSD is implicit in human tragedies, the elephant sometimes stands guard over the dead mahout. In Kerala, India, with a population of 900 working elephants, 2 people die through elephant aggression per month. There, elephants are frequently used for ceremonial purposes, so fireworks can typically trigger spontaneous aggression in vulnerable and insecure elephants. Elsewhere in India and Asia, there may be no single trigger, but just a steady increase in the elephant’s threatening behaviours leading to aggression at some point later.  Unfortunately and unnecessarily, traditional elephant training practices throughout Asia centred on creating submission through inescapable pain during ‘breaking-in’ and punishment for non-compliance throughout life are responsible for the sudden hyper-aggression shown by wild caught and captive elephants.

Elephant training must be extricated from the veiled brutality of folklore and handed to behavioural science to reorganise. There is no question here of ‘if’, it is only a question of when, and which country will take the first steps. Nepal is perhaps somewhat ahead at this point in time, however India is beginning to show great interest through the Wildlife Trust of India and that country has an advantage as a result of its more developed education system. Each country that trains elephants must develop a national consensus to elephant training.

There is also the mahout. The mahout should have a positive view of the importance of his vocation. He must feel its importance through pay and conditions worthy of such a culturally important task. Without exception, mahouts need education in elephant learning processes, and training principles and tasks that arise from such principles. So mahout schools need to be set up in the countries that train working elephants. If this essential education were the focus of governmental initiatives, great steps forward could be made in sparing the elephants the agony of unnecessary stress in early life and in saving the lives of mahouts. There are around 15,000 working elephants throughout Asia but there is a way that you can assist. The Human Elephant Learning Programs Foundation has begun working with elephants and mahouts in Nepal and India but in order to broaden the dissemination of ethical elephant training systems, HELP requires urgent funding (see www.help.org).  The early life of every working elephant is a preventable tragedy that is in our hands to change.

Andrew McLean, PhD | Chairman of Directors,

Human Elephant Learning Programs (The HELP Foundation Ltd)