Category: Crossing Over ¤ Author: Kathy Sdao ¤ Title: Philosophical Difference ¤ Hi Natasha, You wrote: "Secondly, my instructor isn't too keen on clicking.... He >wants to know this - if you have a dog that knows a behavior and is just >being willfully slow or is choosing to be disobedient, how do you use >clicker to compel the dog to do what you want? You can't, you need to use >compulsion. Sometimes the dog just doesn't have the option to do as it >pleases. This may go against clicker philosophy, but how do you use the >clicker in this circumstance? ie.1.) A dog can sit on command, but when >asked to it lazily plunks down it's butt. If we are playing the clicker >game they are bang on, but no clicker, no quick sit. With J&P a well timed >leash correction can snap them back into reality. Although too many kills >the desire to work. What do you do if the dog is doing this? Do you >correct or just wait to train another day? " You ask some important questions, ones that get to the root of the significant philosophical and procedural differences between clicker training and "traditional" (or "command-based") training. Your instructor wants to know what to do when a dog is "being willfully slow or is choosing to be disobedient." His explanations for a dog's failure to perform a requested behavior call for a human judgment of the dog's state of mind. They conclude that the dog is, in effect, being spiteful. How can you really be sure the dog is being willfully disobedient? Clicker trainers try hard to look at observable events rather than unobservable, unprovable mind states. Simply stated, we analyze the dog's behaviors, not his thoughts. In my experience, humans aren't very good at figuring out other *people's* motivations/"agendas," let alone those of another species. Heck, we aren't even very good at figuring out our own, at times. So, what if instead of saying "the dog is spiteful," we say "the dog didn't perform the requested behavior." A clicker trainer would respond by not clicking, and possibly by taking a time-out. These situations actually come up less often than you might imagine, though, because if the initial training has been done well, there's really no reason the dog should fail to respond to a known cue. Why would the dog not do something that is likely to lead to a good result? If the dog makes repeated mistakes on the same behavior, we tend to conclude that the dog doesn't know the behavior as well as we thought he did. In that case, we "go back to kindergarten" by dropping our criteria. We can do this in many ways: ask for a simpler approximation of the behavior, ask for it in a less distracting environment, decrease the duration of the behavior or the distance from the trainer, etc. Notice, too, that I used the word "cue" not the word "command." They really are different concepts. The cue "sit" is like a green light to the dog for that one particular behavior. The cue "sit" means "your human slot machine might give you a treat (e.g., food, a toy, a butt scratch) now if you sit within the next 5 seconds." It does not mean "sit now or I'll make you sit." When I was a marine mammal trainer at a local zoo, one of my least favorite parts of my job was hearing children yell at me throughout the day "Make the whale do something!!" My response was often "How could I possibly *make* this two-ton animal do anything?! He has to choose to do it." Now that I work with dogs, it's certainly easier to "make them" (i.e, compel them) to do requested behaviors because they weigh less ;-). But why should I have to resort to that? A training approach that combines the use of conditioned and unconditioned positive reinforcers, gradual shaping programs, consistency, creativity and patience has always worked well with every non-canine critter I've trained. I figure that it should work as well with a boxer or a poodle as it does with a walrus . Keep asking us the tough questions. The discussions help us all become better trainers. Kathy Kathy Sdao Puget Hound Daycare Plus Tacoma, WA thehound@nwrain.com (206) 572-6214