How can owners tells if their horses are in a learning frame of mind?
How a horse responds during training can be influenced not only by its mood, described by scientists as its affective state, and its alertness, which they refer to as its state of arousal.
However, scientists say their response will also be affected by how attached they feel to the trainer.
Dr Andrew McLean, who directs the Australian Equine Behaviour Centre (AEBC), and Professor Paul McGreevy, of the University of Sydney, Australia, gave a joint presentation on the complex combined impact these factors can have on training to delegates at the recent annual conference of the International Society for Equitation Science
McLean noted that horses had the largest amygdala – a part of the brain – of all domestic animals, which gave them what he described as “a very significant flight response”.
“They are very fearful animals.”
Understanding this innate quality and what possibly might temper such fear could be of value to those who wanted to create an attachment with their horses, he said.
One way to modify this fear may be in how we touch the horse, he said.
McLean pointed out that, historically, horse training has not involved much touching of the animal, yet horses found security with one another through touch. Recent studies have shown the positive effects of allogrooming – social grooming between members of the same species – on lowering heart rate.
McLean proposed that such primary positive reinforcement may be another tool in the training toolbox that can be used to overcome fearful insecurity in the horse.
Touch may be an important way to develop attachment between human and horse, but not in the way some people currently practice it, he said.
“Patting – rather than stroking – may not be the positive reinforcer we think it is,” said McLean, citing unpublished data from a student at his centre, who conducted research during the 2012 London Olympics.
The data showed that two-thirds of horses accelerated upon being patted, whereas horses stroked on the withers had a more amenable response, which McLean proposed may be more conducive to developing attachment.
McLean also discussed previous attachment research in monkeys and humans, then touched on both empirical and anecdotal research into the human-animal bond and stressed the need for scientists and researchers to explore further attachment styles between humans and animals.
McGreevy discussed how a horse’s alertness and mood may enhance or hinder training.
Paying attention to how the horse feels about the training environment may be a good place to start when considering training.
“What we are trying to achieve is an awareness of how these three As – Affective, Arousal and Attachment – might influence operant conditioning, and may explain why it works better sometimes than others.”
Paying attention to how the horse feels about the training environment may be a good place to start when considering training, he said.
McGreevy explained how, through classical conditioning, a horse may come to associate a fearful situation or environment with the trainer, stressing that during training “We really want to keep animals in a positive emotional state”.
Ensuring the horse creates positive associations between the trainer and the environment may magnify what McLean had referred to earlier as the “safety effect”, which could affect attachment, McGreevy said.
McGreevy delved into new research on what is known as cognitive bias, as a potential way to “climb inside an animal’s head”, and determine how each animal is feeling.
Cognitive bias – a term from psychology – describes how individuals can create a distorted perception of their world, affecting attributes of behaviour, memory, attention, expectations, risk perception and more. In essence, it captures whether the individual sees “the glass as half full or half empty”.
The cognitive bias of a wide range of species is currently being studied, with links between cognitive bias and personality showing evidence of pessimism in animals that display stereotypies – undesirable traits – or separation distress.
Being mindful of an animal’s level of optimism or pessimism could help humans enhance an animal’s welfare, McGreevy said.
Mood may act as an early warning system to identify animals at risk of behaviour problems, or may allow trainers to predict which animals are better able to cope with demanding training regimes.
Uusing three-dimensonal landscape graphs that he developed in a recent paper with one of his doctoral students, Melissa Starling, McGreevy showed the range of possibilities that may occur when a horse undergoing different levels of arousal and affective states is asked to complete various tasks.
Such visual constructs may allow horse owners to see why and how a horse’s current state of arousal and mood can influence training.
McGreevy pointed out that horses in different states are going to respond differently to the four training quadrants, stressing that trainers cannot expect all techniques to work equally well, regardless of the animal’s mood.
“For a prey animal, it’s so important we avoid aversive experiences, rather than just focusing necessarily on rewarding experiences,” he said.
“In general terms, what this means is that positive reinforcement is going to be most effective when the emotional valence of the animal is positive. Negative reinforcement is going to work if the animal is alert enough to attent to the stimulation, but not overexcited …
“Punishment, as we know, is going to fare poorly in almost all the landscapes.”
As to the possible positive implications of further research, McGreevy mused: “Is this all about training for effectiveness, or does it have a role in the horse’s welfare? I would say both. It’s definitely both; one follows the other.”
In support of the presentation, there is a free journal article, Response landscape graphs conceptualising the effects of arousal and affective state on training outcomes of operant conditioning in animal training, available here.
The manuscript contains graphs which are best viewed as a series of interactive three-dimensional conceptual graphs which require the Wolfram CDF Player to view. This free software can be downloaded from here.