These are my notes from the ORCA Great Minds conference at UNT in March 2009. The Great Minds conference brought half a dozen top trainers together to talk about animal training, clicker training, operant conditioning and the future of training. This is the third post of my thoughts and notes and the last page of notes about Steve Martin. To read more about the conference in general, please see this post. To read the previous pages about Steve’s talk, please see this post (for part 1) or please see this post (for part 2).
In my first post on Steve Martin’s lecture, I discussed his ideas for what makes training an art, and what makes the difference between a good trainer and a great trainer. I’ll discuss some of my thoughts about his last two suggestions in this post. Just to refresh here’s the basics of his talk:
According to Steve, the artistic trainer is able to do the following things:
- Accept Responsibility for the animal’s behavior
- Empower the animal
- Motivate the animal
- Participate in 2 way communication
- Avoid aversives when at all possible
Participate in 2 way communication
Steve’s discussion of two way communication seemed to cover a whole slew of things, but the basic gist I got from it was that the animal is always reading your body language and you should always be reading the animal’s body language for the cues and signals the animal is giving you. Not sure if this is completely what he was driving at, but it did help me begin to think about some interesting stuff!
He started out talking about how a trainer always must be aware of the animal’s body language and the subtle signals and messages the animal is sending us. For instance, we must be careful not to invade the animal’s personal space and to wait for the animal to invite us to approach. I find that some people really get this concept, but many don’t. Many animals have learned to tolerate people who invade their space, but it doesn’t mean they enjoy it. I recently read Patricia McConnell’s The Other End of the Leash (which is an excellent book, by the way!). In it, she has a whole series of pictures of people hugging their dogs. Most of the dogs look utterly miserable. Of course, there are some dogs that genuinely do enjoy getting hugs. However, having a large creature stoop down and surround you in it’s arms is not natural for the dog and most dogs, even if they learn to tolerate it, do not enjoy it.
One of the most fascinating things I heard all day was Steve’s ideas on bridges and marker signals. We traditionally think of the clicker as the marker signal. However, the animal picks up on all sorts of other signals we give, such as voice and body movement, which can end up serving as markers without us even realizing it.
Many animals use visual bridges much more than we realize and Steve challenged us to consider how big of an influence the clicker is actually having on the animal’s behavior. He showed several clips of trainers working with animals in the zoo setting, mostly involving stationary behaviors, such as targeting, holding a pose, or touching something. While watching, he had us observe carefully for when the animal broke pose, signaling the end of the behavior. We would expect, if the clicker was truly serving as a marker signal, the animal would break pose when he heard the click. Instead, in almost every case, the animal held the pose after the click and moved as soon as the trainer reached for the reward. How interesting!
If the sound of the clicker is not a good predictor of when a reward will be delivered, then the animal depends on your body language and signals rather than the click. This just goes to illustrate that the magic is not in the clicker, the magic is in the trainer’s ability to train. In several zoos, Steve has taken away the clickers and the trainers have been able to achieve the same results and success as they had before.
Does this mean we should all throw away our clickers? I don’t think so at all. It means we should make sure our animals are associating the sound of the clicker with a reward. The clicker is a great tool for precise timing and communication when we use it correctly.
Also, I think the clicker serves an important role in teaching the trainer good timing skills and when and what behaviors to reinforce. However, this does help drive home the point that there is a huge difference between clicker training and training with a clicker! Just because someone is holding a clicker, does not mean they understand the principles of training and positive reinforcement. On the flip side, though, just because someone isn’t holding a clicker, does not mean they aren’t training with the principles of clicker training.
Steve also talked about giving the animal a short opportunity to perform a behavior. If the animal doesn’t perform the behavior, take the cue or object away. (The example he used for this was teaching a bird to step up onto your hand). If we hold the hand out for the bird for a long time, the bird learns he can take however long he pleases to step up and he will still get the reward. However, if we take the hand away after a second or two, the bird quickly learns that if he wants a reward, he needs to step up quickly when presented with the hand.
I found the topic of 2 way communication incredibly interesting, as I know some people who shy away from clicker training because they say it is too mechanical or impersonal. However, good training is about being able to read the animal and create a bond and relationship with the animal. A good trainer knows this and uses this, whether or not they clicker train.
Avoid aversives when at all possible
This one is a big one. Aversives compromise the relationship and have negative consequences and side effects. Instead of relying on an aversive to get rid of behavior, a far better approach would be to create a plan where positive reinforcement can be used to create good behaviors. I won’t talk too much about aversives here, because one of the other speakers talked at length about aversives.
Punishment is anything that follows a behavior and decreases the rate of that behavior. (If something is being used as a punisher but doesn’t actually decrease behavior, then it’s ineffective or abusive.) However, even though punishment will decrease behavior, it comes with other consequences. Some of the main consequences of using punishment include an increase in escape and avoidance behavior, and increase in aggression, apathy, and a generalized fear of the environment. One example he discussed was bird dogs that are trained using shock collars. The dogs often develop superstitious fears, such as being afraid of stumps after being shocked by the collar when passing by a stump.
Even if we can find ways to use punishers without too many detrimental side effects, we need to be cautious with punishers because they only teach the animal what not to do. We might decrease the “bad” behavior, but the animal still has no idea what he is suppose to be doing. Also, animals trained with aversives will only operate at the level to avoid the aversive.
Steve showed an interesting video clip of a trainer teaching birds to step up on a scale. At first, the trainer was working with one bird at a time and shooing away the other birds, which was working with limited success and causing both the trainer and birds to be frustrated. She was only telling the other birds what not to do. Instead, she eliminated punishment by teaching each bird to go and stand on a target station. Then, she could work with one bird on the scale, and the other birds still had a job and something that they were suppose to be doing. In the second clip, everyone was much happier.
Final Thoughts on Steve Martin
Although I enjoyed all the lectures, Steve Martin’s was probably one of my favorites. He was witty and enthusiastic, and shared plenty of stories, pictures and video clips to help illustrate his points. I felt that he really had worked out exactly what he wanted to tell us, and then broke it down into small chunks that would be easy to begin to understand (even if they might take several lifetimes to complete understand!) Also, I felt he gave us really solid ideas that I can use when I think about how I train and that he presented them in a way that the could be understood by those newer to clicker training, while still being applicable and intriguing for those who had been training for decades.
Above all, Steve told us to be flexible and make sure we were training at the animal’s pace. For example he said many trainers will take 3 months to train a particular, complex behavior because the book says it’s suppose to take 3 months to train that behavior. If the trainer was paying attention to the animal, she might see that the animal was completely, utterly bored and could have learned the behavior easily in 3 weeks.
Be sure to check back tomorrow and later in the week, as I’ll be blogging about the rest of the conference speakers. If you don’t regularly read my blog, but find this interesting, I encourage you to subscribe via the RSS feed or by e-mail subscriptions (both located at the top right. Thanks for reading!