Positive and Negative Reinforcement
When Skinner defined the 4 quadrants of operant conditioning all those years ago (early 1930s) he noted that positive reinforcement is superior to punishment in changing behavior from an ethical and efficacy stand point.
Additionally he said that punishment was not simply the opposite of positive reinforcement; shaping behaviours using positive reinforcement results in behavioral modification that is longer lasting, punishment causes temporary changes in behavior and it also creates unwanted side effects.
How Do They Work ?
Positive reinforcement is about the subject working to attain an appetitive (pleasant) outcome. When we clicker train, the appetitive outcome is that the horse gets a treat (primary reinforcer). So we add something nice when the horse gets the right answer.
Positive reinforcement increases the likelihood of the behaviour happening again.
With negative reinforcement the subject is working to find out how to remove an aversive stimulus (something they want to remove, e.g pressure). When we clicker train we don’t want the stimulus to be aversive to the horse, instead we want the pressure to be so light, and importantly not to escalate so that the horse sees this as guidance to ‘yes’ (positive reinforcement). So we remove the stimulus when the horse gets the right answer.
Negative reinforcement also increases the likelihood of the behaviour happening again.
Pressure and Release
With clicker training horses we use both positive and negative reinforcement. We can’t help it; we sit on our horses, we use halters and bridles, saddles, and other equipment, all pressure. If we can manage how the horse understands and feels about the pressure then we can ensure this is seen as guidance.
As such, it is important that we educate the horses well about any of the equipment we will use. We need to educate them about how we would like them to respond to the equipment and we can do that using positive reinforcement.
For example; teaching our horses to accept the bit in their mouth. We can break this down in to very small stages. Milestones might look something like the following;
- Can you bring your head towards the bit (no bridle attached), click and treat (C/T)
- Can you open your mouth when I hold the bit up, C/T
- Can you place your mouth over the bit, C/T
- Can you take the bit in to your mouth, C/T
…and so on. Then we can start to introduce the bridle in stages. Once we have the bit happily accepted by our horse we can start to teach them how we would like them to respond to the pressure we place on the bit.
The main keys to successful (clicker compatible) use of negative reinforcement are:
- The pressure has to be seen as information only
- The pressure must not escalate (a small amount of pressure over a period of time creates the desire for change, Alexandra Kurland)
- The pressure must be released entirely when the desired behaviour (or approximation of) is given by the horse.
- Then we click and reinforce
Pressure should be sophisticated enough to be viewed as the horse as guidance
Positive reinforcement then distinctly tells the horse they got the right answer.
Why use Positive Reinforcement ?
There are a number of reasons for using positive reinforcement; As B. F. Skinner noted when he defined the 4 quadrants of operant conditioning, using positive reinforcement is more ethical and the results are significantly longer lasting than other training methods.
One of the more notable effects of training with positive reinforcement is that it creates a horse who is highly motivated to work and learn. They become very enthusiastic to learn and as such the speed of learning is accelerated.
Horses who are trained with positive reinforcement become excellent communicators and problem solvers who want to work with us.
All fantastic reasons for training our horses with clicker training.