Following up on my last post on NRMs I mentioned that the only place I really use one in my training is when the lack of information will create an expectancy of reinforcement for my dog. I mentioned the teaching of weave poles as one place that a NRM is critically important. The other place is on go-outs (or send aways) in obedience. In both of these examples if you allow the dog to get to the final destination (the end of the poles or the end of your go out) the dog will be reinforced both by the act of running and the history of past reinforcements for the same job.
When I give my NRM to my dogs their ears do not pin back and their attitude does not deflate — it is not relationship altering. To my dogs a NRM means “you are very close to earning a reinforcement, try something different.”
Why do some dogs melt when their owners try to use a NRM? The reason, I believe, is the history in the training. How much has that dog being allowed to fail? Is there a history of the dog being helped by the trainer when it struggles?
A dog’s response to failure is entirely learned. Dogs that shut down are taught to shut down. It is the history of what happens after the shut down that rewards the dog for the act of giving up. For most dogs that have failed their owners will immediately make the next repetition easier so the dog get it right. I certainly don’t have a big problem with that early on, perhaps once or twice but let me tell you it is not something that I prescribe to myself.
I teach my puppies that frustration is a normal part of learning and not to be discouraged by it but rather be excited because you have eliminated another incorrect response so your reward is now that much closer! Check out the lessons 4 1/2 week old Swagger is already learning!
Now I do end up with other NRMs that sometimes I am not even conscious of for poor “manners” type behaviours. Those are always things I blurt out, usually with a chuckle when my dogs are being naughty . . . choosing to get on the couch, stealing a toy without being invited (they are usually something like “oh my” or “I don’t think so” or else similar). However I do not use any such thing to “train” a formal skill.
For myself I follow these guidelines when using a NRM:
1. Never to be used during the “value building” stage of any behaviour. What I mean by that is that I will not use a NRM during the initial shaping of any behaviour. The behaviours I described above would be in the “value testing” stage.
During any shaping session I minimize errors through strategic planning and setting the stage with clever environmental manipulation in order to set the dog up to offer what I want. The lack of reinforcement is the only feedback the dog gets from me during a shaping session.
2.If a NRM is going to be used (during value testing) the massive amount of reinforcement that has been banked during the value building stage prevents any of my dogs from considering shutting down and leaving work.
3. As I said previously, only use a NRM if without one the dog will be receiving inappropriate reinforcement for an undesired response.
These are not 3 separate entities. All three of these guidelines are in place if I use a NRM.
But again, I think the stage is set for my own dogs long before that with the lessons they learn about failing (as outlined in these two previous blog posts; 1. Planning to Fail and 2. Planning to Fail is the Path to Success ).
The lessons of working through frustration and disappointment start very young for my dogs and carry through all that we do together.
Today I am grateful for a clean house (that I didn’t have to clean . . . it doesn’t get any better than THAT!).