Hi Guys, Shawna here! I am often asked about the feasibility of combining positive reinforcement(R+) and negative reinforcement(R-). It is not a simple answer. Some people have referred to this practice as combined reinforcement, but this is not a scientific term or construct. I think the thing we need to consider is what was the impetus that created the behavior? In the learning process, did we elicit the initial response via seeking (appetitive) or avoidance(pressure/aversive)? It is one or the other. Many people elicit the response with a mild aversive and then feed it. This is still R-…but with a cherry on top.
Another common question people ask has to do with how to determine the best way to move toward R + training with a traditionally trained horse. Many horses already have lots of trained behaviors that do not cause them to worry. For example, many horses have learned to wear a halter and to lead with traditional methods, and many of these horses don’t need to be retaught, as long as they don’t seem to have any worry associated with haltering/leading. However, even with horses like this, I do set out to create a new, strong R+ history with those behaviors. On the other hand, some horses need to learn just about everything from scratch. It really is an individual evaluation.
Keep in mind, when transitioning to incorporating R+ into your training protocol that some horses who have adapted to the R- world, can feel quite lost if you stop using all familiar communication. They have learned how to keep themselves from getting corrections by anticipating the correct answer. Not knowing the answer can cause a lot of fear and frustration even though the corrections are no longer a part of their world. As with all training, pay attention to their threshold and stay below it. ANYTHING that causes your horse to worry should be retrained. Moving forward, I ideally won’t be utilizing R- as a tool for training. For future behaviors, I choose to use R+.
More food for thought…as neurobiologists have learned through scientific observation and testing, when organisms learn through these two different styles of reinforcements, different areas of the brain tend to be activated. With R+ the area of the brain referred to as the Seeking System lights up. With R- the area of the brain responsible for fight or flight tends to light up. The chemicals produced by the different areas of the brain are also significant- the area associated with fight or flight produces cortisol (the stress hormone) while the part of the brain associated with seeking produces dopamine and endorphins (so-called “feel good” chemicals). The release of these chemicals creates different associations and different reinforcement histories.
Both R+ and R- can be done well and can also be done poorly. So, understanding and getting either/both of them right is really important. I have worked with Olympic medalists and international competitors in show jumping, dressage, eventing, and competitive driving who are excellent and amazingly soft with R-. For the most part, the horses I worked with were well adjusted and the rider’s/driver’s exceptional timing carried right over as they added R+ to their training protocol. Working with these people showed me what really good timing looks like! But I find the average person and competitor does not actually possess the same extraordinary timing as these elite athletes. There is much more at stake when we utilize R-.
I also think it needs to be clarified that pressure doesn’t need to have an aversive association if we elicit the response to tactile stimulation with R+, meaning the mild touch is immediately removed and followed by the presentation of a target or some other tool that has been taught through free choice followed by R+. Then the tactile cue is not associated with pressure as a training tool but simply as tactile communication. But, again, understanding the impetus for the behavior is crucially important.
We can still prepare our horses for the traditional world by utilizing R+ to teach traditional cues so that the cue and subsequent response is associated with R+ and has a strong R+ history. My horses whom I began training solely with R+ over 25 years ago could be ridden and handled by people who never fed them. The horses were happy and responsive and always earned lots of compliments. They didn’t feel entitled to food for each response because they had a solid R+ foundation. Each behavior had become classically conditioned. They loved canter departs for example; it made them feel happy, so they chose to do it consistently for anyone who asked. Of course, this goes back to the endorphins and dopamine that were part of the initial learning experience and this attitude was maintained through the learning process. They felt that way about all the behaviors that they learned because of the strong and consistent R+ history.
If your horse has some sticky spots with R-, then the stakes are higher. Some horses really have a challenge with a combination of R- and R+. I equate it to having an alcoholic parent; it can feel like you are always walking on eggshells. To our horses, it can seem erratic and scary since they are never certain which persona is going to show on any given day. If your horse is overly sensitive, worried or shut down I’d consider carefully how to address him going forward and teaching new behaviors in the future.
As I mentioned earlier, merging the past and then transitioning forward isn’t quite as tricky as choosing to combine both as learning tools moving forward. The bottom line is that there is no one right or wrong answer. It really depends on your horse, his history, and your experience. I believe making informed decisions is most important. I recommend thinking about what type of relationship you want with your horse and think about what the best way will be to facilitate that relationship.