Not many who read this blog would argue against the fact that a positive reinforcement based training plan is almost always the best path to take when attempting to train an animal (or a person). And yet, something I am going to call “fork in the road” training is immensely popular amongst sport dog training, specifically agility training. Presenting our dogs with a fork in the road; essentially two choices, and providing them with reinforcement only if they choose the correct path is considered an acceptable for sport dog trainers. After all, the only consequence are reinforcement or lack thereof. There are no collar pops, no harsh words, and no electric shocks. The dog is simply presented with a choice and paid for making the “right” one. For those of us who came from the dark ages of competitive obedience or other sports that regularly employ severe corrections, this can feel pretty great. But I’d like to suggest that there might be a better way than fork in the road procedures.
“Fork in the Road” Training
Very common ways of training with a fork in the road (FITR) style include most default impulse control protocols, a good number of discrimination training protocols, and the vast majority of what I find dog trainers referring to as “proofing.” Give the dog two choices, reinforce the desired choice. In the case of the video below, baby Felix could choose to go for the treat, or not. If not, he got paid. If he chose to go ahead and grab the food, the food was withdrawn:
Here, I am working on Stig’s A frame contact. I’d like him to hold his position until released no matter what. I present him with the choice to wait for the release, or to self-release to the toy. If he self-releases I withdraw the toy, as well as the opportunity to work for a brief period. (Enjoy the cameo by Tundra, who left us in June of last year–I know I did).
In the first example, I am providing forks only. In the second example, my dog has a stronger background knowledge about what will get reinforced in this context, but I am still presenting him with forks. While neither dog is being mistreated, neither dog seems ready to quit or overly frustrated, I could do better. Rather than presenting my dog with a fork in the road on the way to reinforcement, I could light the path for him that will certainly take him there. Both of these videos are a couple of years old, and the trainer I am today, while far from perfect, would do a better job of illuminating the way to reinforcement for my dogs.
“Light the Path” Training
The most skilled teachers in the world are very good at deliberately lighting the path to reinforcement for their learners. Increasing criteria in such a way that the learner is not even aware of the increase is what I would call “light the path” training (LTP). When we do this we rarely see errors (of course they happen, but they are infrequent), and our learners benefit from a very high rate of reinforcement. The repetitions that involve punishment are few, and while in some instances our teaching process takes a little longer (the fewer approximations we can hit to get to our final behavior, the faster the learner will get there) it doesn’t feel as long because no one is frustrated.
Here, I am introducing my sister’s dog Kenobi to the concept of discriminating between cues. The concept of cueing has already been introduced through systematic LTP work. Each of the cues I am giving him have a strong history of positive reinforcement (as strong as a 14 week old pup can have!) and this is the first time he was asked to discriminate between all of them. What I like about this session (besides my cute dog-nephew) is how engaged and joyous Kenobi appears, and how quickly I provide him with a new path to reinforcement each time an error occurs.
In teaching Felix to stand for his agility measurements I knew I needed to light the path for him. The wicket (measuring device) is wobbly and usually metal, which speaks of one of Felix’s biggest fears; metal baby gates. I knew that the inherent stress surrounding this sort of thing for him meant that simply taking a FITR approach would only increase that stress, and might lead him to simply opt-out of something that was vital for his agility career. I took at careful LTP approach, and am really pleased with the results. Below is his first session at this, and one of his final ones before actually getting measured.
Felix is now one AKC measurement and two USDAA measurements in, and he has done beautifully. He even bounced back from the wicket falling off the measuring table and crashing to the concrete floor; something I worked hard to avoid but ran into anyway. I could not be happier with his ability to be measured in a stress-free and cooperative way. LTP training is what got us there. Could FITR training have achieved the same results? Possibly, and maybe even faster. But the stress that would have caused us both would not be worth the ends for me.
Next time you set out to make a training plan, think about whether you will be presenting your dog with forks in the road to reinforcement, or with a clearly lit pathway there. Both routes work; and both have side effects. The side effects of FITR procedures are often frustration or stress, while the side effect of LTP procedures is joy.