Thinking through this problem a bit. I agree completely with a couple of points that various people in this thread are discussing.
Thread consensus, as I read it: It is not what the horse is reacting to. It is how the horse is reacting. Regardless of the root cause is fearfulness, or wilfulness, by the horse.
Some of us might want to understand the trigger. But there seems to be agreement that we work on the horseās response to the trigger, to any trigger.
But, imo ā how to work on the response is different for fearfulness vs wilfulness.
Fear is uncomfortable, so in the horse mind, calm is a better deal, helping the training along.
But wilful behavior is self-rewarding, like the counter-surfing dog mentioned above. Changing a horseās wilful self-rewarding behavior will likely need mean a) blocking/preventing the self-rewarding behavior, and b) a more tangible reward, to get the best possible outcome for behavior change. Or at least behavior modification to reduce the necessary tools over the horseās lifetime, depending on the success level.
I do think the nature of the reward for wanted behavior may need to be altered between fearful and wilful horse behaviors.
For the fearful horse, fear feels bad, calm feels good. Calm is a self-reward for avoiding fear.
Wilful behavior is the opposite of the fear-reaction, because self-rewarding wilful behavior feels good.
The self-rewarding wilful behavior needs to be blocked, taking away the self-reward ā every single time, with the necessary tools depending on the trainer and situation.
And the reward for a new behavior has to feel very good to give the horse the incentive not to fight for the old self-reward (which some will do). But instead to more willingly change its behavior pattern.
The horse perceives that the self-rewarding wilful behavior is now closed off, but instead of fighting for it, go with the new rewarding behavior instead. Thatās the strategy, imo.
I think OP is pinpointing this, having determined that her horse is wilful, not fearful. A crucial starting point and I applaud her attention to detailed observation to determine this. An observation skill that will always be needed with this horse.
Unfortunately there isnāt much room for failure with a wilful behavior. Failure will again create a self-reward, and therefore a remembered incentive to try the self-rewarding behavior yet again at some unknown future point. The more failure, the more the wilful behavior will return in the future. This is proven in animal learning studies.
That said, the nature of the control tactics and tools can be key, because the horse has to learn not only that it is not possible to bolt, and it is uncomfortable to try. As a result of the horseās own action in bolting, not because of anything the horse perceives the handler to be doing.
Bringing the head around to prevent the bolt is likely to be enough to prevent the self-reward, and make it uncomfortable, without much pain. No reward. Not comfortable. Horse brain can process that simplicity.
With both wilful and fearful horses, a calm horse brain is necessary to allow a horse to mentally process a change of reaction. To delay the old reaction for even a second or two to allow the process. So, a smart, calm, observant, assured trainer is essential. And yet another reason punishment doesnāt work because it agitates the horse and interferes with learning. As well as creating a whole additional set of reactions to the punishment, as various posters illustrated above.
Horses powered by adrenaline ā due to fearfulness or wilfulness ā tend to be less controllable and predictable. But as we see in racehorses and cross-country horses, horses can learn consistent cues to guide their adrenalized behavior, and not just default to random scatty prey-avoidance behavior.