08/20/2026
R+ CAN be coercive. Talking about it is how we minimize it.
First, R+ is not a training method.
Positive Reinforcement describes a learning process: something is added following a behaviour that increases the future likelihood of that behaviour.
That definition tells us how behaviour is being reinforced. It does not tell us what the entire training system looks like, how skilled the trainer is, whether the horse has meaningful choice, or whether the training is ethical.
And R+ is absolutely not automatically synonymous with consent, choice, agency, or good welfare.
You can use something a horse desperately wants to influence behaviour in ways that become coercive.
You can remove or restrict access to food before a training session specifically to make a horse hungrier and therefore more motivated to work for the food you have.
You can deliberately use extremely high-value food that sends a horse’s arousal and motivation through the roof, then mistake their intense willingness to perform behaviour to get that food for enthusiastic consent to everything happening in the training session.
You can make access to something the horse really wants dependent on them doing what you’re asking. Technically, they’re still performing the behaviour and receiving something they value afterward. That doesn’t automatically mean they had a meaningful choice.
You can create conflict where the horse wants what you have, but doesn’t actually want to approach or interact with you.
You can continue asking for behaviour because the horse keeps participating for the reinforcer while ignoring increasingly obvious signs of stress, frustration or conflict.
And you can technically still be using Positive Reinforcement while doing those things.
This is why saying “I train with R+” actually tells me surprisingly little about the welfare of the horse in front of you.
R+ describes a learning process. It is not a welfare designation or technique.
We still have to look at motivation, deprivation, emotional state, arousal, conflict behaviours, the horse’s ability to disengage, what alternatives are actually available to them, and what happens when they say no.
And I think this is something we need to acknowledge when people outside of R+ spaces criticize what they’re seeing.
Sometimes what they’re seeing is food being poorly handled. They’re seeing horses becoming extremely over-aroused around food, sometimes to the point that the behaviour becomes genuinely unsafe.
Sometimes they’re seeing someone describe their training as ethical, consensual or force-free because food is being used, while the horse’s body language is telling a very different story.
Those observations shouldn’t automatically be dismissed because the trainer is using Positive Reinforcement.
The presence of food does not prove good welfare.
We still have to look at the horse in front of us.
So how do I try to avoid R+ becoming coercive in my own training?
• I don’t restrict forage before training to make food more valuable.
• I choose reinforcers that are valuable enough to motivate, without creating so much arousal that the horse struggles to think or disengage.
• I pay attention to conflict, frustration and changes in body language, even when the horse is still taking food and performing behaviours.
• Possibly unpopular: if I start seeing long pauses, lack of engagement, or clear conflict between wanting the food and wanting to perform the behaviour, I stop making the food contingent on performing the behaviour. I’ll feed them anyway, reassess what I’m asking, and change the plan.
Using R+ does not exempt us from critically examining our own training.
And talking openly about the ways Positive Reinforcement can become coercive isn’t an argument against R+.
It’s part of making sure we’re using it well.
