What does natural horsemanship get wrong according to equine behavioral scientists?

Equine behavioral scientists have identified several specific areas where natural horsemanship theory and practice departs from what the scientific study of horse behavior, learning, and cognition supports — criticisms that apply with different intensities to different practitioners and programs within the broad natural horsemanship movement. The dominance hierarchy model that underpins many natural horsemanship frameworks — particularly the concept that humans must establish themselves as the alpha or leader in the horse's herd hierarchy to be respected and followed — does not accurately reflect equine social organization as behavioral scientists understand it. Horses do not maintain the rigid linear dominance hierarchies that the alpha concept implies; their social relationships are more complex, context-dependent, and individual than the simple hierarchy model suggests, and scientific evidence does not support the idea that horses are constantly seeking to dominate their handlers in the way that the respect-first frameworks of Clinton Anderson and others imply. The anthropomorphism present in some natural horsemanship frameworks — the attribution of complex emotional states, conscious intentions, and relationship motivations to horses that go beyond what the behavioral evidence supports — has also been criticized as producing inaccurate models of horse behavior that may lead to poor training decisions. The concept of join-up as a specific moment of social acceptance — the horse choosing the human as its leader in a decision analogous to social choices within the herd — has been questioned by researchers who suggest that the horse's approach behavior after round pen work reflects fatigue, habituation, and learned responses rather than the specific social acceptance that Monty Roberts's framework describes. These scientific criticisms do not invalidate the practical effectiveness of many natural horsemanship techniques, but they suggest that the theoretical explanations for why those techniques work need revision in the light of what equine science has revealed.