One upon a time there was an adolescent dog in an open admission shelter who had one day left.
And a woman who knew nothing about dog training, already had a smaller dog, and made an impulsive decision to go get the shelter dog. Some problems ensued.
That’s how a lot of people discover dog training, I think. And I suspect that I’m not the only one for whom that dog who started it all has a special place in the heart.
As a positive reinforcement trainer, I know that just about whatever a dog is physically capable of, you can train and put on cue. If you can figure out what is reinforcing to the dog, you can get reliable behavior. And you can even create new reinforcers by pairing them with ones the dog already has.
On the other hand.
In 1961 Keller and Marian Breland published a paper called The Misbehavior of Organisms. It was in part a response to B. F. Skinner’s work The Behavior of Organisms. The Brelands outlined examples of training difficulties having to do with countering the natural, instinctive tendencies of animals.
I think of this when I look over my motley crew of dogs, noticing not only their individual quirks but how their personalities and interests might be related to their breeds (mixed in most cases). There are certain breeds of dogs that have been selectively bred for generations to thrive on work with humans. Herding dogs and retrievers come to mind. My dog Clara exhibits the tendencies I’m talking about. A strong valuing of almost any kind of activity with a human and a strong focus and attention span to work.
My dog Summer is different. She has an abundance of independent varmint dog genes. Her are some shots of her in her element.
I hadn’t put any of this together when I decided to go into competitive dog sports with her. Summer is a non-traditional obedience breed mix, putting it lightly. I’ve related part of my painful learning curve in the post, “Ant Sized Treats.” I talk about how I finally learned how to really motivate her with food. But I didn’t mention in that post the work I have put in on trying to use tug as a reinforcer.
Summer does like to tug. She will tug very heartily with me for a little while. And I have had a tiny bit of success using tug as a reinforcer. But only when there is no food in the picture. If we are already playing tug, I can ask for behaviors and reward them with tug. But if it is a “training session,” food trumps tug. (And playing in the hose trumps some food, even.) Problems like this with food and tug are perfectly solvable. Here is a nice video my friend Marge made teaching Maple, a boxer puppy, how to enjoy multiple reinforcers in a session. So please let me be clear that this is my choice as a trainer not to pursue food and tug issues with Summer until they are solved. I’m sure it could be done. But I have four dogs and plenty of essentials to train. With Summer it would be uphill. Zani and Clara will both tug in the presence of food and vice versa, and I’ve done much less work with them on it.
What Summer loves to do is to carry off and dismember toys. Perform squeakerectomies, fuzz-ectomies and anything-that-sticks-out-ectomies. That is by far her favored part of the predatory sequence. Allowing this type of self-reinforcement (with the human out of the picture) is somewhat Frowned Upon by many of the sports dogs trainers. We are supposed to make sure that the dog plays with toys with us, not on her own, or certainly not extensively by herself.
But Summer is a beloved pet and I am fine with her tearing up toys on her own. So: how then should I handle the challenge in our upcoming AKC Rally Advanced trial where a dog must heel past distractions on the floor including food and toys? We have practiced a lot with food and have a protocol for that. So I wanted also to figure out a way to reward Summer with a toy in such a way that would 1) maintain the Zen behavior she has and build on it, 2) be fair to her, and 3) be truly reinforcing. Letting the Treat Fit the Feat, as I’ve written about previously.
For this dog, grabbing up a toy and whooping it up expecting her to tug with me would not be reinforcing, especially if I took the toy away before she could rip it up.
With help from the Training Levels list and my teacher I came up with the following plan: In Rally context (recognizable to her), Summer never gets the treat off the floor or the toy off the floor, even after the run is over. She has to exercise Zen self control and for her, I can better accomplish it with giving her a treat that is separate from the one on the floor. She gets a great reward at the end of the run and it fits the challenge: big bite of some great food, or yes, a toy to shred, or both. Both come out of my pocket or outside of the Rally setup premises (mimicking what we will do in a trial). Summer already knows that routine: long behavior chain, then run to the crating area for something great.
I bought a bunch of very cheap stuffed toys so shredding could be part of the routine each time we practiced.
My plan was to finish the run, make a big deal about the food and toy, then leave her to her shredding. But the first day I implemented my new plan I found out something amazing. It turns out I AM part of Summer’s play with the toy. I gave her the toy, she was surprised, but immediately made it clear that she wanted me to sit with her while she pulled it apart. She simply didn’t want it unless I was there, too. I was touched almost to tears. This is my independent dog. This is after years of on and off work on my part to make playing with me fun, but then watching her continue to prefer to take the toy into a corner and shred it herself. So she had her toy, I hung out with her and bragged on her for 5 or 10 minutes, and she was SO happy. The second day we had our routine down better and She. Started. Bringing. Me. The. Toy. Again, you’d have to know the history. But it turns out that all the other stuff I had done to get a toy fetch, rewarded with tugging, was too much pressure. When I gave her some time, she went back and forth between shredding it herself and bringing it to me to tug and handle together. I left my jaw on the floor somewhere that day. The whole experience also brought home to me that for her (in a household of four dogs), time alone with me is very special.
So we ended up having, rather than an instantaneous reinforcer (even a whole jar of baby food takes less than a minute for her to eat), but a reinforcement period. There are some tricks to that–some things the trainer chooses to offer are bound to be of lower value than others so there may be moments of disappointment–and perhaps I’ll write about that in another post. But I feel like spending several minutes with my attention entirely on my dog and what she would like to do was a Treat that Fit the Feat.
On the third day, I brought out the video recorder. What you will see in the video are excerpts from an 8 minute session that Summer and I had in the back yard. I set out the plates of distractions, we did rally practice for about four minutes, then we finished, I released her, and we ran to another part of the yard. I gave her two huge bites of pumpkin cake, then gave her a disposable toy. I hung out with her. She gutted the toy. We tugged a little. She never once turned around to check out the former distractions, but just hung out with me in a relaxed way. She was free to leave and choose a different reinforcing activity at any time. After a time we went together (not cued by me) and I picked up the plates (during this part you can see that she is still very interested in them), then went back and hung out some more and she solicited some petting.
This will probably look pretty low key to a lot of you. Unless she is aroused about something, Summer is a pretty low energy dog. But check out the photos at the top again and compare them to her demeanor in the video. In the video Summer pays lovely attention to our work. After she is released, she doesn’t leave. We don’t see her patrolling the perimeter, digging, hunting turtles, or going up to the top of the porch to check out the neighbors, or any other favorite activity. She reacts not at all when a dog barks or the neighbors use their chain saw. She doesn’t take the toy off to a corner. She doesn’t prowl back to the plates. This is the most amazing thing. I mean, our Zen cue does not have anything like that duration. She is choosing to hang out with me over the chance to sniff and pilfer some stuff off the ground, which is always hugely enticing to her. (I would have picked up the plates if she had tried, as part of our rule structure. But the point is she didn’t even seem to think about them.)
So my hypervigilant dog chose, out of all the available reinforcers, to hang out with me in a relaxed way in a distraction filled area.
My miracle dog.
I realized after some discussion in the comments that I had not talked at all about the fact that in my video and in my practice lately, there is a delay between the behavior and the reinforcement period. As most of you probably know, in most cases a delay between behavior and reinforcement makes for ineffective training. The relationship between the two can break down, or never form in the first place. On the other hand, there is ample research about delayed reinforcement that shows that animals can learn to connect delayed reinforcers with the behavior. Sue Ailsby in the Training Levels and some other trainers have techniques to teach the dog about this connection.
With Summer I have followed Sue’s technique and am pretty sure that she does make the connection. I spent several weeks a year or two back going to Rally practice wherein I didn’t give any treats during the run, but afterwards we ran back to our crate area and she got a whole jar of baby food. Her performance and enthusiasm improved markedly during that period, evidence that she connected the great food treat with the rally sequence. We do the same thing for agility runs, and have a routine that even includes putting her leash back on before running for her goodies.
I don’t give training advice on this blog but I want to give a simple caution that if you have not taught your dog about delayed reinforcement, a period of food treats, play, and attention such as I show in the video would likely not be connected to a previous behavior chain. And frankly, I don’t think that by the end of my hanging out with Summer she was thinking, “This is all because I did my Rally moves so nicely!” But I think at the beginning she probably did experience the doggie equivalent of that. And if we are going to have fun and hang out together, it certainly doesn’t hurt to pair it with an activity that I want to have good associations for her.