René Descartes died in 1650. His argument that animals were automata — sophisticated biological machines without consciousness, without felt experience, without inner life — was the dominant philosophical framework for human relationships with animals for the next two centuries and, institutionally, for considerably longer.
The Paradox of the Classical Masters
The classical masters worked within this framework. Pluvinel, de la Guérinière, Baucher in his First Manner — all trained horses as one might train a complicated machine. And yet, empirically, they arrived at methods that we now recognise as welfare-positive: the looped rein that released pressure as reward, the descente de main that tested self-carriage rather than enforcing it, training based on repetition and release rather than sustained force.
They stumbled on what works without understanding why it works. The horse's nervous system, its capacity for learned behaviour, its response to release of pressure as reinforcement — none of this was available to them as a theoretical framework. They discovered by feel what the science would later confirm.
Darwin in 1872 and the Tradition That Ignored Him
Darwin published The Expression of the Emotions in Man and Animals in 1872. The argument that animal emotional experience was continuous with human emotional experience — that fear, pain, frustration, and contentment were not uniquely human phenomena but evolutionary continuations of capacities shared across species — was stated clearly and with substantial evidence.
The German military training manuals that codified what became the FEI Training Scale were published in 1886 and 1937. Long after Darwin. Without integrating animal consciousness into their framework. The horse in those manuals is a training problem, not a sentient participant.
The Five Domains and What They Find
The five-domain welfare model, developed by David Mellor and colleagues, specifically includes mental state — whether the animal is experiencing positive or negative affect — as a welfare criterion. Competition horses assessed under this model are clearly experiencing negative affect in large numbers. The model exists. The data exists. The sport continues as before.
Modern dressage's institutional structures were built on Cartesian assumptions that the science abandoned 150 years ago. The Cartesian horse is not a contemporary innovation. It is a medieval philosophy that competitive sport never updated. Updating it is not an optional enhancement to what the sport does. It is the prerequisite for the sport being what it claims to be.