“But restraint is the only one sort of control, and absence of restraint isn’t freedom. It’s not control that’s lacking when one feels ‘free’, but the objectionable control of force.” B.F. Skinner, Walden Two
Traditional training and countercontrol
The basic goal of animal training is to control another individual’s behavior. Cooperation is the ideal outcome, but when aversive training methods are used the animal’s cooperation is compulsory, not voluntary. Worse yet, aversive control can evoke unwanted resistance and oppositional responses. As a horse behaviorist, the severe behavior problems I’m called in to see surprise me less than how readily horses give in to aversive control.
Traditional horse training utilizes various aversive techniques, such as negative reinforcement (the application and removal of pressure or pain), restraint, and even punishment. Historically, these methods have been relatively successful, leading to their continued use and stalling progress toward more humane, force-free, low-stress methods. The use of aversive control by the owner, trainer, or behaviorist is reinforced when the horse gives in, but is punished when the horse becomes anxious, aggressive, or avoidant. Rather than questioning the training technique, however, the horse is often labeled fractious, disobedient, domineering, or flighty. Especially with horses, many riders and trainers don’t think there’s a problem unless the animal behaves in a way that interferes with the person’s goals.
In his work on human nature, B.F. Skinner wrote that the use of aversive social control, including force, punishment, deprivation, restraint, and removal of positive reinforcers, is widespread in human relationships. The individual being controlled can either give in—which reinforces the use of aversive control—or oppose it, by avoiding, escaping, or attacking. These oppositional “countercontrol” responses function to regain behavioral freedom. Although Skinner described countercontrol in humans, it follows that this principle also applies to other social animals and to human-animal interactions.
The rest of this article is focused on training voluntary cooperation and using protected contact as alternatives to aversive control. These alternatives could potentially improve horse welfare, as well as increase ease of handling and safety.
Voluntary cooperation refers to having the freedom to participate in or withdraw from an activity. It’s generally accepted that having control over one’s own behavior and outcomes—including unpleasant ones—is associated with improved mental and physical well-being. When voluntary participation in animal husbandry and veterinary care is trained using positive reinforcement, problematic countercontrol responses are less likely to occur; the animal has greater control over what happens to it, and is thus less fearful, avoidant, and aggressive.
My off-the-track thoroughbred provides a useful illustrative example of how positive reinforcement of voluntary cooperation can modify a problem behavior and eliminate countercontrol responses. From the moment he first arrived at the farm, this gelding was a pleasure to work with in most respects, but one glaring exception was putting on his bridle. His previous trainer warned me, “You’re gonna have to be really quick if you want to get that bridle on!” The reason was immediately obvious; as soon as he saw his bridle, this tall athletic horse would clamp his jaws shut and point his nose to the sky far beyond my reach. The avoidance tactic was very effective—for the horse. Rather than relying on superior speed, which I do not possess, or resorting to aversive techniques, I used food-based positive reinforcement to shape “head lowering” and “taking the bit” into his own mouth. As the trainer, my job was simple: I held the bridle, marked the desired behavior, and then offered a treat.
Zoos have adopted systems to promote voluntary cooperation in training, husbandry, and veterinary care for a wide variety of animals, including elephants, primates, large carnivores, birds, fish, and reptiles. Marion Desmarchelier, DMV, dipl ACZM, veterinary medicine faculty at the University of Montreal, has developed voluntary veterinary care practices for horses based on her experience working with zoo animals. At the 2017 Animal Behavior Society conference, Dr. Desmarchelier presented a positive reinforcement protocol for training horses to voluntarily position their hoof for a radiograph. She also reported that the horses showed fewer indicators of distress during the procedure when restraint was eliminated.These photos show the first steps in training a horse to voluntarily position a hoof for foot radiography. Initially a low, wide wood block is used, and later its size is reduced. In some cases, horses are trained using protected contact behind a stall gate.
Working in a large area with the horse at liberty is preferred, however; the horse is free to leave at any time, but she will not.
Protected contact in zoos
Protected contact is a system for animal management first adopted by zoos in the 1980s. The person and animal are separated by a physical barrier, which increases both safety and animal welfare. Voluntary cooperation is a defining feature of protected contact; the animal chooses if and when to participate, and distressed animals are free to express frustration or move away into a safe zone. In protected contact systems, training involves positive reinforcement with food and praise, shaping goal behaviors, and targeting. For example, when working with elephants in protected contact, a target is often touched to a part of the animal’s body as a cue to take a certain position for a procedure or activity.
Tara Gifford (Ohio Animal Training, LLC) shared some background on the history of protected contact in zoos. It grew from an effort to increase human and animal safety during veterinary procedures by putting large mammals into restraint devices, but not all of the animals would enter voluntarily. When positive reinforcement experts were consulted, in some cases the restraint devices were more successfully used as protected contact barriers.
Free contact is the traditional system of zoo animal management. In free contact, the person and animal share the same space, and training involves both positive and negative reinforcement. Human safety is achieved through restraint, and a distressed animal’s attempt to escape or attack is prevented by increasing restraint. Many zoos use both protected and free contact, and in some cases standards of care have been established. For example, The AZA Standards for Elephant Management and Care mandates that elephants and care providers cannot share the same unrestricted space except under clearly defined circumstances.
Protected contact can be used for veterinary procedures—including foot work, skin care, blood draws, and some medical treatments—when moving animals from one location to another, and for training and husbandry. A recent study at the Atlanta Zoo compared how often an elephant’s responses during bathing sessions were positively reinforced in protected and free contact conditions. Even though all trainers claimed to primarily use positive reinforcement, it was actually used about eight times more often in the protected contact condition. Another striking finding was that, in the free contact condition, trainers used negative reinforcement (applying a tool called an ankus or bullhook) just as often as they used positive reinforcement with food. Aversive control was never used in protected contact, and the elephants were free to choose when and if to participate in bathing. Interestingly, when free to choose, elephants were slower to respond and much more likely to refuse to participate at all, particularly when the behavior was effortful or caused physical discomfort, such as lying down. Rather than being viewed as an undesirable outcome, the authors suggested that the delayed responses and refusals might be a positive welfare indicator since they showed the elephants’ control over choice.
Protected contact in horses
Some similarities between working with elephants and horses include a high risk of injury, the traditional use of free contact with reliance on restraint and negative reinforcement, and training methods grounded in dominance theory. Horses are domesticated animals that are accustomed to being handled, and protected contact is rarely used or needed. The primary goals of protected contact are to increase human safety and animal welfare, so it might be a good option if the horse is frightened or dangerous, in rehabilitation, or unaccustomed to being handled, such as young foals and wild horses.
At present, there are very few protocols for training voluntary cooperation and using protected contact with horses. Ironically, protected contact has gained popularity among positive reinforcement trainers to prevent horses from aggressively mugging them for food. Typically, the trainer works from behind a fence with the horse confined to a stall or paddock. The need for protection might be expected when aversive training methods are used, but why would it be needed when positive reinforcement is used? There are several possible explanations. As an unconditioned stimulus, high-value food can elicit unwanted behaviors and emotions that are independent of the operant training goals. Furthermore, withholding food is a form of aversive social control, which can evoke countercontrol frustration and aggression responses, to gain access to the food.
Some potential problems
Some zookeepers have resisted the adoption of protected contact systems, and horse trainers might have similar concerns. Protected contact requires the person to give up control, and to remain patient when the animal is slow to respond or refuses to participate; the trainer must set their own goals aside and respect the animal’s choice. Working from behind a barrier can also get in the way of fine-tuned adjustments of the animal’s behavior, and caretakers cannot access some parts of the animal easily.
Elephant keepers have also expressed concern that protected contact interferes with making a strong social bond with the animal. However, because protected contact has been linked to higher rates of positive reinforcement, the animal may actually have an improved perception of humans, as was found in horses trained with positive reinforcement. I also argue that reduced social contact during training might not always be a bad thing, and could result in fewer unwanted avoidant and aggressive countercontrol responses. To borrow an example from Dennis Delprato’s article on countercontrol, a horse is more likely to behave aggressively toward a person (or other horse) who is blocking its path than toward a tree branch blocking its path, because aggression is only effective in the social context.
Concern about safety and welfare have changed the standards for animal care and management. Many zoos and laboratories have replaced traditional approaches with more humane, animal-centered methods. Positive reinforcement of voluntary cooperation and the use of a protected contact system, discussed in this article, are two such methods that can easily be adapted to horse training and handling.
Robin Foster, PhD, CAAB, CHBC is a professor at the University of Puget Sound in Tacoma, WA, and an affiliate professor at the University of Washington. She has taught courses in animal learning, behavior, and communication for more than 20 years, and conducts research on behavior and welfare of working and sporting horses. She provides behavior consultations, clinics, and educational workshops in the Seattle area through Adaptive Animals, LLC.