Mechanical approaches

Most naïve horses respond to humans as they would to any predator. They behave in ways that help them avoid physical or psychological pressure, by moving away bodily or posturally. These basic evasive responses can be modified successfully to produce a highly responsive equine performer or unsuccessfully to produce a problem horse.

When trainers encounter horses that rapidly enter conflict and as a result do not comply, they often experiment with increased pressure to overcome resistance. Mechanical restraints and stimulants may be used to magnify the pressure that riders can apply. Because of the perceived need to force horses into an ‘outline’ and make them work ‘on the bit’ (see Glossary), the tendency is to use ever-stronger bits as the first approach to solving any resistance. This accounts for the multiplicity of bits on the market. As with any device, it is true to say that if there are many of them available, the chances are that none of them provides all the answers. Furthermore, enlightened trainers recognize that where aversive stimuli have failed to elicit the desired response and have begun to cause behavioral conflict, the application of more force is contraindicated. They rarely reach for such gadgets.

Frustrated trainers often try bits that apply pressure to different parts of the mouth or to the same area of the mouth but with greater severity (Fig. 15.2). Even though they can and sometimes do sever the tongue,⁵ saw-chain bits (so-called mule bits and correction bits) are readily available to riders who are having problems getting horses to respond to milder bits. If changing or increasing mouth pressure is unsuccessful, an alternative or additional means of making the horse adopt the desired shape is to apply pressure to other parts of the head, so training devices such as curb chains, gags, hackamores, bosals, draw reins, balancing reins and chambons may be employed. Unfortunately the tendency is to develop a reliance on these extra pulleys, rather than use them solely for retraining as the name might imply. Martingales and tie-downs are similar in that they apply pressure to prevent evasive raising of the head and are rarely dispensed with.

Figure 15.2 Etching of 18th-century bits. At times it seems that more effort has gone into designing bits than addressing poor equitation.

If the horse fails to show sufficient forward movement or impulsion, trainers direct their attention to the flanks where they can increase the pressure and more effectively drive the horse forward or simply away from the rider’s leg using whips and spurs. Although, for some, these stimulants are distasteful, they are not necessarily contraindicated since they can be used with subtlety and employed transiently to empower the rider’s leg signals. In contrast, tying horses’ heads down (with side reins and draw reins) and clamping their jaws shut with nosebands⁶ are counter-productive in the long term. Indeed there is growing concern that these approaches can automatically compromise welfare by virtue of their inflexibility and inherent absence of pressure-release.⁷

Behavior modification

Behavior therapy can help overcome undesirable equine responses that may have an innate component but are largely learned. The temptation may be to treat such traits in isolation. However, a holistic approach has tremendous merit since better results tend to occur if the human–horse relationship is nurtured when the rider is both on the ground and in the saddle. Time spent training horses in-hand seems to pay dividends even in the ridden horse,⁸ presumably through a process of generalization. Achieving stimulus control in-hand seems to facilitate stimulus control under-saddle.⁹ Inter-specific communication may help horses overcome fear and therefore reduce their tendency to use counter-predator responses. It is important that the translation of trained responses from cues in-hand to cues under-saddle is better understood by practitioners. It may simply be that an unconfused horse in-hand is a better prospect for training than a confused one. In any case, correctly identifying the effects of reduced confusion and fear and distinguishing stimulus control from mere compliance may allow us to describe and even measure the bonds that form between horses and their trainers.

By the same token, punishment is regarded by many owners as unenlightened, since it presupposes that horses know the difference between good and bad behavior. So, it is increasingly inappropriate for veterinarians to advocate punishment as a response to undesirable behavior in the horse. It undermines the human–horse bond and carries the risk of injury, which may place any practitioner who advises it in a legally liable position. Examples of acceptable techniques used for behavior modification include habituation and counter-conditioning (see Chs 4 and 13). Habituation is the key to overcoming fear-related responses such as traffic shyness but it can be expedited with counter-conditioning, an approach that relies heavily on the shaping of alternative responses through operant conditioning (see also Fig. 15.3). At its most elegant, equine operant conditioning takes the form of positive reinforcement and subtle negative reinforcement. Consistent use of negative reinforcement is the key to retraining the basics, but is notably more likely to work in horses that have not habituated to pressure. Techniques for restoring sensitivity to pressure cues demand exquisite timing of pressure-release. Meanwhile, clicker training employs positive secondary reinforcement that is particularly useful for refinement of responses, once negative reinforcement has established the basics.

Figure 15.3 Horse being clipped in the presence of food – an example of counter-conditioning.

(Reproduced with permission of Australian Equine Behaviour Centre.)

Horses with innate or learned evasion responses may have the rewarding outcomes of these responses removed by a process of extinction. Furthermore they can be trained to associate the fear-eliciting stimuli with positive outcomes in the same way that laboratory primates can be trained to voluntarily present their forelimbs for intravenous injections. (A case study that required a pony to learn to cope with hypodermic injections appears at the end of this chapter. It followed the injection desensitization training technique developed by ethologists at Pennsylvania State University.¹⁰) A far less humane approach to fearful horses is to flood them by exposing them to an overwhelming amount of the fear-eliciting stimulus in the hope that their responses are exhausted. For example, so-called ‘tarping’, a method of breaking by flooding, involves evoking learned helplessness by casting the horse, covering it with a tarpaulin and beating it until it no longer resists.

Practitioners are strongly encouraged to become extremely conversant with learning theory (see Ch. 4) before counseling owners on the application of these techniques. All of the undesirable responses in the current chapter can be remedied by behavior modification. But a prerequisite is to correctly identify the motivation for a given response. Readers are urged to consider the ethological content of this entire volume when determining causal factors. Beyond that the success of behavioral therapy depends on the consistency of administration and compliance of owners, who must be counseled that the latency for desirable outcomes depends largely on the number of times the horse has trialed and learned from the unwelcome response. With a thorough knowledge of learning theory, practitioners can customize behavior modification programs, modify advice in the light of initial responses and counsel owners when proximate results surprise the uninitiated. So, for example, when as a result of an extinction protocol, undesirable responses become more frequent, the accomplished student of learning theory can explain that this is a manifestation of the frustration effect (see p. 103 in Ch. 4) and that it is transient.

Inappropriate obstacle avoidance

It is appropriate for horses to avoid hazards that may jeopardize their safety, but such innate self-preservation responses sometimes inconvenience riders (Fig. 15.5). Examples include avoidance of ground hazards, manifested as running out and refusal of fences, ditches and water jumps, while avoiding lateral confinement may prompt a horse to refuse to enter starting stalls. Unfortunately, the original causes of these responses may be obscured and the subsequent application of aversive stimuli may simply confirm to the horse the need to avoid such obstacles because they become associated with pain.

Figure 15.5 A horse avoiding obstacle despite (or perhaps even as a result of) the efforts of a novice rider.

(Photograph courtesy of Mellisa Offord.)

In other words, riders and handlers may focus on a given hazard and, through the use of pressure cues, may inadvertently make it increasingly aversive. It is therefore preferable to first work on deficits in the quality of operant responses such as leading to a light cue and straightness in less challenging contexts. Again, the sequential steps in a linear scale of training (see Ch. 13) should end in proof only after the preceding steps in the scale have been achieved repeatedly. Shying behavior is most rapidly learned and incorporated into the horse’s behavioral repertoire when it permits complete escape. As the horse escalates its shying behavior, the reinforcement from escape is compounded by the incremental loss of control by the rider.² Additionally, riders who predict the responses may inadvertently apply pressure that may increase the horses’ motivation to find freedom.² Re-training the ‘go’ response in such horses (see Ch. 13) is the preferred remedial course of action.

The importance of rhythm and direction in basic training is explained when one considers horses that learn to refuse jumps. Habitual losses of rhythm and tempo or directional line often lead to stalling while approaching fences.² When horses learn to stall in the jumping effort, outright refusal is generally their next step.

Hyper-reactivity responses

While innate obstacle avoidance responses may paradoxically help to ensure the safety of riders (e.g. by prompting the horse to steer around a hazard unseen by its rider), other innate self-preservation action patterns of horses, especially those that keep them safe from predators, are sometimes triggered too readily or are performed with speed that riders are unable to predict. These are chiefly flight responses motivated by a need to get away from the stimulus, rejoin a group of conspecifics or return to the home range. Depending on the human observer, horses that readily offer theses responses are variously described as sharp, keen, fizzy or flighty.¹² Of course, these labels blame the horse rather than identifying deficits in its training.

There are several common examples of hyper-reactivity in the ridden horse. As discussed above, shying is an avoidance response that sends horses leaping laterally or backwards away from auditory, olfactory or visual stimuli. Many horses in unfamiliar surroundings offer this response innately but some adopt it as their default response to novelty. Traffic shyness is a particularly important example. The most successful approach to this dangerous problem involves a process of habituation, but the challenge for trainers is to present a comprehensive range of diverse stimuli (a model for habituating former racehorses appears in the case study in Ch. 4).

Jogging, on the other hand, is largely learned and manifests as a horse that is unresponsive to signals to walk (Fig. 15.6). This undesirable gait is often accompanied by persistently repeated extension of the neck to pull the reins through the rider’s hands in horses that have learned to ‘pull’. These horses demonstrate a clear need for reinstallation of the ‘stop’ cue. Similar unresponsiveness takes an extreme form in horses that bolt, i.e. that gallop with no response to rein pressure. These animals may learn to ignore bit pressure whenever they wish to remove themselves from a perceived threat or return to their home range or social group.

Figure 15.6 A jogging horse. Jogging is a choppy trot that wastes energy and often indicates the need for better training of clear transitions.

Agonistic responses to conflict

When faced with discomfort or a threat, most horses move away. If horses cannot escape from such a stimulus, they enter behavioral conflict and increase their kinetic effort in a bid to relieve the pressure (Table 15.2). Some horses develop seemingly irrational phobias by becoming sensitized to associated stimuli and anticipating the escalation of bit or leg pressure that riders use to make them ‘behave’. The therapeutic approach to conflict is discussed in detail in Chapter 13.

Table 15.2 Selected examples and descriptions of agonistic responses to conflict

Evidence of pain and irritation

When a horse fails to respond as requested by the rider, the innate human tendency is to repeat the cue with greater force in a bid to increase the animal’s motivation to relieve the pressure. However, it is important that somatic causes of a horse’s disinclination to respond are eliminated before more aversive stimuli are applied. There is a strong argument for the attending veterinarian to use analgesics as a diagnostic probe when horses fail to respond appropriately, especially if this departure is sudden. Pain is recognized as an important contributor to unwelcome behavioral responses.¹³ Occasionally horses may learn to relieve pressures caused by riders by removing the riders themselves, e.g. by bucking (Fig. 15.7), rolling or even rubbing the riders against fixed objects.

Figure 15.7 A horse bucking under-saddle.

(Photograph courtesy of Andrew McLean.)

Gentle snorting is usually a good sign in horses under-saddle because it tends to be accompanied by signs of relaxation such as spontaneous head lowering. In contrast, grunting and groaning are expiratory noises associated with tenesmus and abdominal guarding. As such they reflect low-grade discomfort in the very general region of the horse’s forequarters. Far more profound an indication of thoracic and lumbar pain is seen in ‘cold-back’ syndrome during girthing and mounting. This often includes responses to being saddled, mounted and especially to being girthed that vary from aggression to recumbency.

Horses commonly learn to reduce the discomfort of the bit by manipulating it to lie in relatively insensitive parts of the mouth. Common names for these evasions include raking, boring, getting the tongue over the bit and seizing the bit between the teeth (Fig. 15.8). Mouth pain is usually associated with heavy-handed riding or inappropriate equipment, e.g. the action of some jointed bits is thought to cause pinching between the commissures of the lip and the second premolar. Some maintain that in the event of a horse fighting the bit, comfort in this part of the mouth can be maximized by creating a ‘bit seat’ or ‘cheek seat’, i.e. removing an appreciable portion of the second premolar.¹⁴ While one study reported improved athletic performance in most horses after the creation of bit seats,¹⁴ an abiding question is whether a simple change of bits (e.g. to an unjointed design) would be just as effective. Alarmingly, some overzealous equine dentists suggest that partial tooth removal should be considered for all horses that are required to wear double birdles. This is an outrageous suggestion since it effectively means that all horses competing in elite dressage should be surgically altered in this way. The use of double bridles (and the jaw-clamping nosebands that are frequently used with them)⁶ should be questioned before horses’ dental arcades are routinely modified to accommodate them.

Figure 15.8 A draft team in which several horses can be seen evading the bit.

(Photograph courtesy of Les Holmes.)

Evidence of poor physical ability

While some horses do not comply with riders’ signals because they associate the responses desired by riders with musculoskeletal discomfort, others are insufficiently athletic because of poor conformation or lack of fitness (Table 15.3). Clearly, as with those resulting from pain, these cases are unresponsive to behavior therapy. It is also worth checking that riders are giving signals in consistent anatomical sites, e.g. lack of consistency can explain why a short-legged rider may fail to prompt a satisfactory ‘go’ response from a horse that has previously been ridden only by long-legged riders.

Table 15.3 Selected examples of inadequate performance under-saddle that reflect poor physical ability

Evidence of learned helplessness

Horses may learn that they are unable to help themselves when responses they use to relieve pain or discomfort or threats (or their precursors) are unsuccessful. By a process of habituation, such horses often become unresponsive to the stimuli, but it is unclear whether they all do so with a concurrent reduction in physiological distress.

Horses that exhibit general reluctance to work and resistance to signals from the rider are described as ‘stale’ or ‘sour’. Specifically they may be referred to as being hard-mouthed (Fig. 15.9) or lazy and sluggish, whereas, more correctly, they should be considered to have habituated to rein pressure or leg pressure, respectively. It is difficult but not impossible to re-educate these horses (see Ch. 13). Classically, animals that have learned helplessness are generally unresponsive to multiple stimuli in their environment, not just training-related stimuli.

Figure 15.9 A hard-mouthed horse.

(With kind permission of Elvira Currie.)

Human causes of unwelcome behavioral responses

Some unwelcome responses disappear if management deficiencies are corrected or if the rider becomes more skilled or is replaced by a more talented equestrian. Horses learn to evade discomfort in both appropriate and inappropriate ways. Sometimes, for example in rodeos, horses are required to produce exaggerated agonistic responses as part of their work. Indeed Schonholtz¹⁵ points out that some horses are purpose-bred for a heightened anti-predator response (e.g. one line of buckers accounted for 30 horses used at the US National Rodeo Finals in 1996). While this is claimed by some to be evidence for the increasingly humane treatment of rodeo animals, it may simply indicate that hypersensitive horses can be bred. This means that selected animals are less likely to habituate to being ridden with a bucking strap in place because they continue to find the procedure aversive. So purpose breeding seems an unconvincing justification for the use of selected rodeo animals on welfare grounds.

While schooling in the form of operant conditioning can make a horse respond desirably, it can also evoke resistance, conflict and learned helplessness when administered crudely, inconsistently or too rapidly.¹⁶ Avoiding and resolving problems in the ridden horse have both already been explained in Chapter 13.

Poor riding technique

The importance of rider position is always stressed in riding manuals, with good reason (see Fig. 7.24). Without this prerequisite, riders attempt to maintain or resume a centered position and usually have a significant effect on the horse’s movement and balance in the short term (Fig. 15.10). In the long term, a rider’s lack of balance allows the horse to practice unrequested responses. The more opportunities the horse has to trial responses that are not under stimulus control, the greater it perceives its freedom from the control of the rider. This allows an escalating series of random behaviors to be trialed and possibly reinforced. These tendencies, in turn, can affect a rider’s confidence and thus prompt the emergence of a so-called defensive seat, characterized by a forward tilt of the upper body with excessive rein contact and some gripping with the lower legs. For these reasons, novice riders should never be paired with uneducated horses.

Figure 15.10 An inexperienced rider showing lack of balance, with the hands uneven and the legs placed too far forward. Typically these characteristics predispose the rider to being ‘left behind’ the horse whenever it moves forward with unexpected speed, an outcome that prompts many novices to compensate by using the reins to balance.

(Photograph courtesy of Sandro Nocentini.)

Hyper-reactivity

While reactivity is selected-for in some high-performance breeds, especially those used for racing, it is unwelcome in the breeding stock of others such as those used for draft purposes. Motivation to respond to pressure may be reduced in the more stoic animals that are commonly labeled sluggish. Perversely the so-called warmbloods may be so tolerant of bit pressures (Fig. 15.11) that they can perform in dressage competitions while subjected to bit pressures that hotbloods cannot ignore.

Figure 15.11 Unfortunately, strong rein contacts are normal in contemporary dressage, perhaps because judges struggle to identify lightness.

(Reproduced with permission of Julie Taylor.)

Hyper-reactivity may arise from inappropriate matching of horses with the work required of them. For example, some horses cannot be used in traffic because they are intolerant of large objects moving in their peripheral vision. Additionally, most traffic-shy horses learn to anticipate aversive stimuli from riders (including rein and leg pressures) when traffic is encountered. This means that instead of habituating to traffic as most horses do, they recognize the rider’s alarm and become especially responsive to traffic-related stimuli, in other words, they become sensitized (see Ch. 4). As discussed with shying, even in the absence of anticipation, inadvertently applying pressure may increase the horses’ motivation to find freedom (Fig. 15.12).^2,22

Figure 15.12 A rearing horse under-saddle.

(Photograph courtesy of Sandra Hannan.)

Beyond mismatching horses for work, two general causes of hyper-reactivity are management and schooling. We know that the frequency of unwanted behavior during training is increased by stabling.²³ By feeding horses inappropriately and housing them without regard for their need for conspecific company and spontaneous exercise and play, humans increase the likelihood of explosive displays of locomotory behaviors. So, the influence of intensive management should always be considered when a horse presents with training problems, e.g. because shying appears more likely in stabled horses, it is appropriate to consider the effects of post-inhibitory rebound²⁴ and the role of over-feeding high-energy foods as contributors to hyper-reactivity in the wake of confinement. As far as the role of schooling is concerned, it is worth remembering that some riders train their horses to be acutely sensitive to leg pressure. This is appropriate only if the horses are never ridden by novices who may trigger responses inadvertently, e.g. by using their legs to grip and balance.

Horse-related causes of unwelcome behavioral responses

If horses continue to perform poorly, despite demonstrable improvements in management or technique, they may have innate or acquired physical anomalies that make them unsuitable for ridden work. This is not meant to suggest that the horses are at fault, since most of these problems can ultimately be attributed to human intervention or omission.