Mixing +R with +P
First off, I admit to not reading this whole thread of 15 pages of comments (now 28) regarding operant conditioning (clicker training) with horses. However, I was alerted to the recent discussion of mixing +P (added punishment) to +R (added reinforcement). Not to mince words, but that is crazy making!
When I first started clicker training my horse (for six years now), I was lucky to find a well-trained, very experienced horse clicker trainer not too far away. This person has worked closely with Alexandra Kurland for over 12 years.
I started my horse behind protected contact and sometimes go back to that for a refresher. In fact, whenever I want to enter his stall, I wait for him to back (I’ve shaped the back up) and treat at arm’s length over the stall gate.
The mugging problem: Many horses new to the new paradigm of CT, often mug and even scrape the palm of your hand with their teeth. The mugging needs to be addressed through working in protected contact and making sure that your treat delivery mechanics are clean, crisp, and clear – always separate to three events of clicking, reaching for the food, and delivering the food. Many of us, even experienced clicker trainers, get sloppy.
Early in our CT relationship, my horse got more than muggy, he bit. He bit badly. I have the scars to prove it. It was very frustrating! And painful! Of course, I lashed out and slapped him for it. It did not feel good to me emotionally. And here are the points to consider when doing that.
First, punishment, by definition, reduces or eliminates unwanted behavior. The slap or the punch in the nose might work for that particular training session, but if the behavior continues during the next training sessions, then, by definition, the slap or the punch was not actually punishment. “What!?”, you say. “It wasn’t punishment? But I hit him! Wasn’t that punishment?” No, not by definition. Since the behavior continues, it wasn’t punishment.
So what to do? Hit harder? Longer? With something heavier? With a whip? Will that work? It might. But then what’s the fallout? What will happen in your horse’s mind when you start punishing him like that? Will he match your aggression with some of his? If he does match your aggression with his aggression, what you are you going to do next? Get even more violent? “Art ends where violence begins.” Sorry to say it, but the violence began when you hit back.
If you’re not allowed to hit back, then what do you do? What did I do when my horse bit me and I lashed out and hit him? The slap didn’t work. It might have suppressed the biting for that session, but, in the long term, it didn’t help at all. What helped, was my taking a long, hard, soul-searching look at how I trained. Were my food delivery mechanics up to snuff? How about my Rate of Reinforcement (RoR)? Was it high enough to keep him engaged? Was I training in a way to avoid frustration – both for me and for him? I guarantee you, that if you’re frustrated, so is your horse (or dog, bird, whatever).
What’s causing the frustration? Again, look at your mechanics of click timing, RoR, and food delivery mechanics. According to Bob Bailey (the last living direct connection to B.F. Skinner and the Brelands), 80% of trainer problems are caused by 1) Timing, 2), Criteria, and 3) Rate.
(If you don’t know who B.F. Skinner, Keller Breland, and Marian Breland Bailey and Bob Bailey are, then you need to do some research. These are the people who discovered and refined the use of operant conditioning.)
A good training plan should include identifying:
- Timing: What to reinforce. Is your click well timed? Or are you off by just a bit, either too early or too late?
- Criteria: What to reinforce and for how long to reinforce at one criterion before moving on. Are you clicking for only one criterion per session or are you all over the map? Did you raise your criteria too high too fast?
- Rate of Reinforcement: Making it worthwhile for the animal. This is how fast you’re reinforcing behavior. A good rule of thumb is to try to deliver 10 treats in 90 seconds or less.
To these I would add: mechanics and planning.
- Mechanics – I’ve already talked about those.
- Planning – have you sat down and really spent some time thinking about what behavior you want and how you plan to get it?
If you are not getting the behavior you DO want or you are getting behavior you DON’T want, then it is up to you, as the “smarter” half in this equation, to take a look at your training skills.
Now back to specifically addressing mixing +R and +P. First off, you’re not really mixing them – you’re instantaneously switching from one to the other. And, in the process, totally confusing your horse. Are you to be trusted as his new, reliable, consistent best friend? Or are you some crazy being that could go off at any time? Which way would you like to be treated? Imagine having someone offer you a cookie and then punching you in the nose or slapping your face? Then doing it again and again. At what point would you decide that the cookie was not worth the effort or that you couldn’t trust that it was really going to be given to you? When would you shut down and walk away?
Peggy Hogan has many tips, especially safety and food delivery, available for download both in PDF and as a video on her website, www.thebestwhisperisaclick.com
, and she has a Facebook page where many thought-provoking discussions and education take place – Clicker Training Horses group page.
Alexandra Kurland (www.theclickercenter.com
) has several books and DVDs for sale that cover all the basic topics, especially safety and food delivery. If you buy her “Riding with the Clicker” book, you can join her Yahoo! Discussion group, The Click That Teaches.
Shawna Corrin Karrasch also has info available on her site, Shawna Karrasch and On Target Training | Positive Reinforcement Clicker Training | Horse Training
If you’re going to delve into CT for horses, please read up on it and research it so that you know what it’s all about and how to use it correctly. Learn the science. Do the research.
I recently finished working in two five-day chicken workshops with Bob Bailey (see above). Bob Bailey: "We believe trainers should invest time knowing their technology. We believe trainers should invest time developing skill at the training craft. We believe in what they are doing and how they are doing it!"
"We believe animal training should be:
“The better a trainer understands and practices science and technology, the better can be the trainer's art and craft.
“Animal training is a mechanical skill. Practice the mechanics without the animal first. Then introduce the animal and practice the mechanics some more.
“Animal training demands learning and applying information.
"OC (Operant Conditioning) is not about:
“Developing relationships or heirarches (please, no alpha rolling the chickens).
“Thoughts (Vulcan mind melds are not permitted.)
“Emotions (Emoting with a chicken will not get you behavior.)
“Tricks or gimmicks (We get behavior the hard way, we earn it!)” If the animal gets frustrated, you've probably raised the criteria too high or lowered the rate of reinforcement too low.
One needs to completely understand each of the four quadrants AND know exactly how to use each one effectively, efficiently, AND HUMANELY! Bopping a horse on the nose is not humane! +P is not humane!
Bob Bailey, whom I just spent 10 days with, learning as much as I could, who spent 40 years training over 140 species of animals, did use all four quadrants, including +P, but he did THAT maybe a handful of times and always reluctantly.
If you have to resort to +P AT ALL, there is something wrong!
If you get frustrated, think about what the animal is feeling. If you're frustrated, I guarantee that the animal is too. Take a break. Think more. Be more creative. Write down a plan. A training plan includes identifying: Timing (when to reinforce), Criteria (what to reinforce and for how long to reinforce at one criteria before moving on), and Rate of Reinforcement (making it worthwhile for the animal).
There is a whole body of scientific research and data and experience in operant conditioning. It's simple in concept, but is complex in practice. Simplify whenever possible. Split as much as possible and then split some more.