New Frontiers – Getting Behaviour in Dog Training Part 2

World famous animal trainer Bob Bailey has said that an animal can be trained to do anything of which they are physically capable.  It’s a bold claim but one made by a man who has trained thousands of individual animals in hundreds of different species.  I’m inclined to believe Mr. Bailey is absolutely correct based on the incredible variety of “tricks” I have seen dogs trained to perform.  It seems that the limitations most trainers run into have more to do with their abilities to get their dogs to do the desired behaviour in order to train it properly.

 

As I said in first part of this discussion on getting behaviour, there are many clever and creative ways to teach our dogs which behaviours we want them to learn.  The techniques I’ve described thus far involve patience and some creativity in manipulating the dog’s environment to encourage the behaviour we are looking for.  But there is a new frontier in teaching our animals that actually involves the active participation of the animal; one that engages the imagination and initiative of the animal in discovering together what might be possible in training.

 

 

I use Mark and Reward training with my dogs.  It is one of a variety of reward-based training methods that encourages the dog to repeat the behaviours I want rather than discouraging alternatives using reprimands.  Lure/Reward training, Clicker training, Positive training, and many other variants all implement the same philosophy in rewarding the dog for offering desired behaviours in training and setting them up to be successful rather than testing whether they have learned yet and punishing them if they fail.

 

The product of this kind of training is the cooperative dog.  By working with our dogs in an environment where the desired behaviours are rewarded frequently and incorrect responses are largely ignored, we create dogs who are unafraid to offer what they think we are looking for in the way of behaviours.  It can be remarkable to see some of the clever variations that our dogs offer up when they are unafraid to experiment in hopes that something they do might produce a reward.  It is that cooperative spirit, the initiative and creativity, that can provide us with some of our most powerful training tools.

 

 

One of the most unexpected techniques I have learned in teaching my dogs is “adduction.”  Adduction has been described in two main applications: additive adduction and sequential adduction.  Defined simply, Adduction is the application of two or more known behaviours to create new or more complex behaviours.  An example of additive adduction, or “blending”, would be to cue a dog for two behaviours that they know at the same time.  If I cue my dog to spin in a circle using a visual cue and also to jump using a verbal cue, my dog might start to spin and, upon hearing the “jump” cue, leap up performing a “corkscrew” behaviour while spinning in mid-air.

 

The obvious question that most people ask me is why my dog doesn’t stop spinning and just jump straight up.  And this gets back to that fearless creativity that comes from reward based training.  Why not try both behaviours at the same time?  Reward based training teaches my dog that lots of new and interesting things have resulted in lots of rewards.  So she sees no harm in trying something new.  Because I keep my reward rate high when I train, if she fails on this particular attempt, the chances are she will be asked to do something that earns a reward any second so she is willing to try.

 

It is the “fearlessness” to offer behaviours created by good reward based training techniques that provides the foundation.  Many of the new and inventive ways trainers are developing to get our dogs to give us useful an interesting behaviours require a dog with this willingness to try new things without fear of being reprimanded.

 

 

One of the great training games we learned about many years ago came from animal trainer Karen Pryor.  She called it “101 things to do with a box.”  The training process is devilishly simple.  Place box on the floor and watch your dog interact with it.  Then mark and reward anything your dog does to interact with the box.  Did he touch it with his nose?  Reward!  Did he knock it with his paw?  Reward!  Did he push it across the floor?  Reward!  The process starts by rewarding the dog for just interacting in any way with the box.

 

This is where the clever part comes in – once the dog has offered a behaviour with the box a few times (say, pushing it), stop rewarding that behaviour.  The dog will likely try the behaviour a few more times to try to discover why it’s not working any more.  In that process, they will offer some variation, some different behaviour to try to earn a reward.  Mark and reward that new behaviour!  You have just taught your dog that trying something new and different can pay off.  

 

The wonderful aspect to this game that has been so useful for me with my own dogs is that it has taught them that, if I’m not giving them any cues or prompting, they are free try something on their own.  It’s that initiative, that creativity, that willingness to show me something that opens up some wonderful new training opportunities.

 

 

At a recent training conference, Dr. Jesus Rosales-Ruiz presented some interesting research coming from behavioural learning laboratories.  The studies presented concerned two different training scenarios with human subjects.  In one scenario, four different behaviours (let’s call them A, B, C, and D) were trained in separate training sessions.  The second scenario involved teaching the subject a behaviour in different stages (we will call them 1, 2, 3, and 4), each stage building on the previous stage behaviour.  The goal of the study was to watch what happened when the subjects came to a training session and were given no cue or prompt to help them know what was expected.

 

The subjects who had been trained and rewarded for each different behaviour (A, B, C, and D) responded to this unprompted test by offering the previously trained behaviour in the order they were trained.  In other words, they would offer A to see if that worked, then B, then C, and finally D.  Now here’s the interesting finding.  The subjects who were being trained in stages offered behaviours in reverse order from most recently rewarded back to the initial behaviour!  It’s important to note that in both scenarios, the subjects were engaged enough to willingly offer behaviours in hopes of getting the reward.

 

What is the value of this to training my own dogs?  Well, Dr. Rosales-Ruiz showed us a video of a woman who had trained her dog to stand on for targets with the back feet slightly raised on a higher surface.  The behaviour had been trained in stages like the “1, 2, 3, 4” example I gave above.  She then took her dog into her dining room where a chair had been pulled out from the table.  After leaving her dog to offer a few behaviours without prompting, the dog backed up to the chair and put its back feet up on the chair and held that pose.  It was repeating the most recently rewarded version of the targeting behaviour that it was learning in a different training exercise!

 

 

Reward based training has been incredibly successful in teaching both traditional behaviours like pet manners and more complex performance sports like Obedience and Agility.  But the new techniques made possible by teaching dogs to offer behaviours without fear of reprimand have not just allowed us to teach our dogs more with less time and difficulty, it has dramatically extended the range of behaviours that we are able to teach them!  These techniques are revolutionizing fields like search and rescue and service dog programs among many other disciplines.  Modern trainers are truly pushing the frontiers of what we are able to teach dogs.

 

It seems our limitations as teachers and trainers may have defined our perception of the intelligence and capabilities of our dogs.  We are discovering that dogs may be more intelligent (in their own unique way) and may be capable of far more useful and interesting behaviours than we have previously thought.  Bob Bailey’s statement that we can teach an animal anything that they can physically do now sounds more like a challenge to me than a flight of fantasy.  The questions is not whether or not my dog can do the behaviour, it is whether or not I can find a way to teach them the behaviour I want!

 

The key, I think, is to raise dogs that work with us rather than for us.  Modern reward based training techniques are giving us new and exciting ways to do that.  From simple games like “101 things to do with a box” to more sophisticated techniques of “adduction” and “resurgence”, we are learning to use the intelligence and initiative of our dogs to create better and more productive working relationships with our dogs.

 

I’ve been using this kind of training for more than 12 years and the field continues evolve new and incredibly effective ways to train.  More importantly, it’s a heck of a lot of fun for both me and my dogs!  While I’ve shared more than 160 articles about my experiences with this wonderful way to train, there are literally thousands of people out there offering classes, writing books, and making videos to help people explore and use these great new techniques.  If you haven’t yet discovered modern reward based training, you may be missing out on discovering just what you and your dogs can really do!

 

Until next time, have fun with your dogs!