There has been a mixed response to some of the topics from this year’s ASAT conference; particularly, pertaining to CAT-H and Negative Reinforcement, and the research showing that pre-signals are not necessarily conditioned aversives.

CAT-H is the acronym for Constructional Approach Training for Horses, and utilizes Operant Counterconditioning with retreat upon relaxation to allow the animal to control proximity of stressful stimuli, which, when done correctly, can be less stressful than counterconditioning through Positive Reinforcement or a non-operant approach.

Negative Reinforcement is the scientific term for traditional Pressure and Release training.

There has been a lot of resulting discussion about the place of Negative Reinforcement in training, and the concern that these discussions will lead to a green light attitude toward the use of ‘aversives.’

I for one think it’s a great thing we are having these discussions, because many times we are discouraged from even talking about these things, which is pretty ridiculous, considering that very few of us operate within a Positive Reinforcement bubble.

These recent discussions go against a lot of the theorizing that we have been reassured is an absolute science in the Positive Reinforcement community.

With any new complex information that contradicts previous oversimplifications, we experience a liminal period of cognitive dissonance, and we often find ourselves doing mental gymnastics trying to resecure our older, and easier, thought processes… for instance, the presumptions that we should always operate under threshold, that all stress is bad, that ‘pressure’ is ALWAYS aversive (or it wouldn’t be effective, right?).

The most important step in growing away from theorizing, towards observing reality and structuring our training approach around our observations, is letting go of the security of dualistic thinking.

Knowing that the answers can change with each animal, that we must build a fluidity of approach that focuses on the emotionality of the learner to determine what is and isn’t appropriate, is a lesson in vulnerability.

We often base our training decisions off of structures like the Humane Hierarchy or LIMA (Least Intrusive Minimally Aversive), but something to keep in mind is that these are meant to be used as guidelines, not absolutes, and although we may rely on them heavily when we first start learning, as we gain more experience over time, we rely more on letting the individual animal guide our decision-making.

In addition, many of us don’t realize these guidelines have been influenced by what many consider an outdated 2-branch model of the Autonomic Nervous System (ANS), which positions stress as always aversive.

This has influenced a lot of our theorizing of what constitutes as aversive and harmful training to the animal.

The modern 3-branch model of the ANS gives us a little different and less simplistic and absolutist a perspective…

In her blog post about horses and the polyvagal theory, Sarah Schlote of EQUUSOMA stated:

“There are modern horse training articles that still teach that the nervous system is either in a stress response or a relaxation response, and therefore that stress signals are a bad sign.

Polyvagal theory shows that it is far more nuanced and interesting than that.

That outdated nervous system view creates an unnecessary sense of panic or fear that people are automatically doing harm if using pressure/release…

Traditionally, the autonomic nervous system (ANS) has been taught as having two distinct branches – the sympathetic nervous system (SNS) and the parasympathetic nervous system (PNS).
This two-branch model has become ubiquitous and mainstream, and is still used by therapists, sports psychologists, coaches, riding instructors, equine behaviourists and other horse professionals.

This perspective of the nervous system talks about sympathetic arousal, commonly known as the stress response.
A common expression from this model is “you can’t be stressed and relaxed at the same time”, meaning that finding balance involves coming out of a stress response and into a relaxation response.

Some who hold this view, both in the human trauma field and the horse training field, sometimes interpret any signs of stress as being negative and evidence of distress, pain or confusion, and encourage avoiding things that increase stress, prioritizing ‘down-regulation’ as desired and ideal.
While this is not entirely untrue, it nonetheless is based on a limited understanding of stressors and disregards how stress can be useful.

Neuroscientist Dr. Stephen Porges has conducted groundbreaking work on understanding the autonomic nervous system (ANS) in mammals. Among his contributions was the proposition that the ANS has three branches as opposed to two.

Rather than one branch being on while the other was off, the three branches act more like dials, each on to varying degrees at the same time, and responding in a hierarchical manner in response to changing environmental conditions, moderated by the vagus nerve. This is actually much less complicated than it sounds.

Because the parasympathetic NS has two branches (two brake systems), it is easy to confuse one for the other: dissociation, submission, stoicism, and shutdown may ‘look calm’ on the surface, but are really a sign of the high tone dorsal brake being on to try to control the undercurrents of overwhelm or flooding.
‘True calm’ comes from the low tone dorsal brake (rest and digest) or the ventral vagal brake (social engagement).

The over-arching goal is not to continually avoid stress, as this results in a narrow window of tolerance, but to grow the window and one’s resilience by learning that the nervous system can experience fluctuations without going into shutdown.

Someone recently told me, ‘The reason I use positive reinforcement is because I recognize that domesticated horses experience a lot of stress that is out of their control, and so I don’t want to add more stress to their existing load thro,ugh training methods that use aversives’ (as in negative reinforcement, or pressure-release).

There is a certain amount of validity to this perspective.

Similar to trauma therapy with humans, if the conditions are not optimal (as in, the conditions are not safe and supporting regulation and coherence in the nervous system), then it is not the time to do the harder work of trauma processing (which can be more activating).
However, when the conditions are ‘good enough’ (there is enough safety and stability both internally and externally), then dipping into the deeper stuff can be attempted, to help de-activate the bound charge that is still held in the body and to grow the window of tolerance.

Bear in mind, that there is a difference between therapies that encourage flooding and therapies that focus on the titrated renegotiation of little incremental amounts of survival activation to grow the window of tolerance.

Just because the conditions are right doesn’t mean that flooding is ever appropriate (if your goal is to induce compliance, then that’s another story). Titration is still crucial to not send someone into overwhelm, which is simply a re-enactment of retraumatization.

Similarly, there are horse training techniques that encourage flooding (most often methods that mis-use negative reinforcement to dominate and induce learned helplessness) and those that are more nuanced and attend to these thresholds of intensity in order to help grow the window of tolerance. Both negative and positive reinforcement can be done with this perspective in mind.

There is a big difference between negative reinforcement done to dominate and shut down (sending the horse into the ‘red zone’), and negative reinforcement that is done while attending to thresholds, nervous system states, and that uses connection and the social engagement system (ventral vagal branch of the PNS) to moderate arousal so that it is not detrimental.

We grow the window of tolerance by recognizing that what goes up will come down, and by helping the nervous system tolerate incrementally bigger thresholds without getting stuck in fight or flight activation or shutdown.

Not all negative reinforcement (-R) is above threshold. Negative reinforcement that is below threshold is not necessarily detrimental to the horse.
Negative reinforcement that goes above threshold (i.e., into a high degree of SNS without a ventral brake, or into high tone DVC shutdown) is problematic.

What the polyvagal theory and Somatic Experiencing can add to this picture is the idea that it’s maybe not as simple as ‘show the desired behaviour and you will be allowed to escape or avoid the scary thing,’ that implies tolerating and overriding.

Instead, the goal is to experience a pendulation of the nervous system in response to a tiny stimulus (a pendulation is the full arousal and settling of the ANS) to recognize ‘oh, this isn’t so bad,’ before the window of tolerance grows a little to allow a little more — without stimulus stacking and resorting to escape or avoidance to cope.
This, in turn, builds confidence and resilience.”

— Sarah Schlote

The Polyvagal Theory and Horses: An Introduction

For more resources on this, check out the blog link above, as well as the Facebook page:


For an example, we might think about how titrations can help us avoid flooding during saddling…
Most horses don’t buck with initial saddling when we manage emotional thresholds and avoid triggering instinctive reactivity. Counterconditioning works great here.

But operant behaviors like reinforced reactivity in ‘rehab’ horses can be much different than instinctive reactivity.

This is similar to the observations of dog trainers who work with dogs who have operant vs instinctive aggression.

In the photo below, we had a trigger that was very context-specific, an operant behavior built from the reinforcement of repeated successful saddle buck-offs after going over threshold with other trainers. Operant Counterconditioning can be a good fit here.

Practically, my advice would be, slow down a little, and use a breastcollar just in case.

An ounce of respondent prevention
is worth a pound of operant cure…

Fix negative associations
before you have to fix negative behavior.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s