Positive Re-enforcement Only Techniques

[sidoney]

You can say he does this for reward. What reward? I don't give him any

That's actually quite an easy question to respond to, only I and others will do it mindful of what we are doing. The release, ie. the opportunity to run around again, is the reward.

[pgm]

Sidoney, Sidoney, Sidoney...how is being released a reward for his offering the behavior of his own accord? He is already free. How is giving up his freedom to come back into heel of his own accord a reward? What are you you saying?- he voluntarily gives up his freedom in order to be rewarded with freedom?

But even if your interpretation is correct (and to be honest I am quite happy to accept it) the dog is making a cognitive judgement, of his own accord, from his own intitiate. His behavior in other words is self-directed. And that of course, is the goal of my training, that the dog acts in a self-directed manner appropriate to the enviroment in which he lives.

Even if I accept that he is motivated by the reward that the release provides, then I am happy to give it. It is the kind of reward that I have in ready supply. It costs very little and is much more convenient than having to carry around treats all day.

Edited February 10, 2005 by pgm

[Tabata]

how is being released a reward for his offering the behavior of his own accord? He is already free.

In OC terms: he is responding to a cue (the other person or dog) which you have trained. Cues to behaviour don't have to come from you. His reward is to go back to what he was doing.

This is a game I play all the time with my dog on walks - she starts to sniff something, I call her away, her reward is to go back to the sniffing. We built it from a low level until now I can call her off truly fascinating smells and she literally bolts back to me.

[sidoney]

Sidoney, Sidoney, Sidoney...how is being released a reward for his offering the behavior of his own accord? He is already free. How is giving up his freedom to come back into heel of his own accord a reward? What are you you saying?- he voluntarily gives up his freedom in order to be rewarded with freedom?

Yup. It's a behaviour chain. He has previously learned to come to you when he sees people, by whatever method you taught him. He has the habit of it, and the anticipation of some consequences (perhaps not every time, but enough to keep it going) for the behaviour, of whatever ilk. He's looking for the reward, at the end of the chain - the release. And I would also say, for your approval - of course dogs are relationship oriented. Behaviourism gives training tools that allow one to gain predictable results, but it's not everything - dogs and other animals are not emotionless automata.

My dogs often offer previously learned behaviour, because they are looking for the consequences. And yes, some of those consequences are coming from me and from our relationship.

But even if your interpretation is correct (and to be honest I am quite happy to accept it) the dog is making a cognitive judgement, of his own accord, from his own intitiate. His behavior in other words is self-directed. And that of course, is the goal of my training, that the dog acts in a self-directed manner appropriate to the enviroment in which he lives.

Makes them much easier to live with, doesn't it? <VBG> Many dogs have to be self directed in this way.

Even if I accept that he is motivated by the reward that the release provides, then I am happy to give it. It is the kind of reward that I have in ready supply. It costs very little and is much more convenient than having to carry around treats all day.

You bet! And if you train with rewards, once the dog understands the behaviour, and when you have developed a great relationship with your dog, the reward can be positive attention from you, as well as such things as the opportunity to run free, to get into the car, etc. That is all positive reinforcement.

My puppy, now - I consciously train her with extra rewards, as we are establishing behaviours and building relationship. Later, I won't need to for the day to day stuff - the rewards will be more relationship focused.

When I am asking for an extra special effort that is difficult for the dog/s, then I use extra rewards - that would include for my dogs, doing an agility course accurately and FAST - they will do it for me anyway but to get the extra effort I need to give them extra motivation. That is the Vizslas. They are not self rewarded by the course - not enough anyway to give their all. The Kelpie - well, she's young, we'll see.

Also I train the dogs in new things with appropriate and highly desired rewards; that helps them to focus through the learning process and to keep trying even if it's hard.

[pgm]

In OC terms: he is responding to a cue (the other person or dog) which you have trained. Cues to behaviour don't have to come from you. His reward is to go back to what he was doing.

Tabata, I have no problems with this description. So maybe you and Sidoney are correct. The only thing that I would insert in this description is that training develops the dog's understanding - hence the 'cue' would have no effect unless the dog 'understood' it as a cue.

[sidoney]

training develops the dog's understanding - hence the 'cue' would have no effect unless the dog 'understood' it as a cue

You have my complete and utter agreement on that one!

Part of teaching a dog - and the most important part, in my opinion, is the teaching of the dog "how to learn". I've experienced it with young, unhandled horses and with green dogs too - there is a turning point, where everything up to then is not really doing much, then they understand that their behaviour can control the consequences, then it all happens from there.

Oh yes, I should add, the second thing they have to learn is that some environmental stimuli (cues) "open the door" to certain ways of obtaining consequences. This, for verbal cues anyway, can take a little longer.

Edited February 10, 2005 by sidoney

[snobbybobby]

These are scientific facts. Please don't put your own interpretation on them."

Really? Then could you please provide me with the scientific citation.

PGM,

B.F. Skinner wrote about behaviour of organisms in 1938 and many people who have trained animals for a variety of purposes have used the science of Operant Conditioning since then.

I suggest you get hold of a book called "How Dogs Learn" by Mary R. Burch, Ph.D. and Jon S. Bailey, Ph.D. published in 1999 by Howell Book House.

BTW the method is used by people training sea mammals in places like Sea World. These animals can't have a check chain around their necks for obvious reasons.

Does that answer your question?

I'm not about to get into discussions with people who simply want to make a laughing stock of the method.

Henrynchlo

[pgm]

henrynchlo,

I'm sorry I can't help it, the theory of OC is a laughing stock. Not my fault, go speak to Skinner...

[Guest Tess32]

henrynchlo,

I'm sorry I can't help it, the theory of OC is a laughing stock. Not my fault, go speak to Skinner...

PGM....I thought it was the positive trainers that could be insulting?

Nat

[sidoney]

Oh dear ...

I don't think we need to insult each other or each others' ideas.

Can we please keep it to an interesting discussion and not make it personal?

[snobbybobby]

PGM....I thought it was the positive trainers that could be insulting?

Nat,

PGM is not insulting, just has a small, closed mind and I feel for any animals who are caught up as the result of this discussion.

Where would the world be today if everybody had that sort of outlook?

We would still be wondering if we would fall off when we rowed too far!

Henrynchlo (thinking "what's the use, the animals are the losers"!)

[pgm]

Henrynchlo,

in order for my mind to open, you need to put forward an idea. So far you have put none forward...

So as far as I am concerned you could for all I know be a complete idiot.

Then again, my mind is still open...still waiting for you to put forward an idea

Do you have any?

[sidoney]

If we all recall that the board was almost closed down because of personal behaviour on it ... let's get back to the topic, huh?

[boxagirl]

I believe that positive methods CAN work on ALL dogs, if used correctley.

A GREAT book on positive dog treaining is "The Complete Idiot's Guide To Positive Dog Training" by Pam Dennison. It is probably the best and most informative book I have ever read. It is simple and explains the theory and science behind all the behaviors. It covers basically every aspect of dog training, and actually explains why positive dog training is better. It completely changes the way you think when training a dog (and actually makes you stop and think before just yelling or yanking the leash!!) I strongly suggest that anyone who is interested in dog training reads this book.

[snobbybobby]

A GREAT book on positive dog treaining is "The Complete Idiot's Guide To Positive Dog Training" by Pam Dennison. It is probably the best and most informative book I have ever read. It is simple and explains the theory and science behind all the behaviors. It covers basically every aspect of dog training, and actually explains why positive dog training is better. It completely changes the way you think when training a dog (and actually makes you stop and think before just yelling or yanking the leash!!) I strongly suggest that anyone who is interested in dog training reads this book.

Thank You for supporting my contention

I have the book that you suggest and find the same theory being put forward.

I not only use positive reinforcement with my two dogs but, I use the same techniques with shelter dogs, in particular puppies!

I find it amazing the response that is obtained when an action is REWARDED rather than PUNISHED. A dog that sits then is rewarded with a treat will sit and sit again until the treat is forthcoming! That, my friend, is positive reinforcement.

Henrynchlo

[Kavik]

I don't think anyone is arguing whether positive reinforcement works (as all training methods use positive reinforcement, including traditional techniques),

Rather whether positive ONLY techniques work as well as those that are balanced, on every dog and in every situation.

Edited February 10, 2005 by Lia

[sidoney]

Good point Lia.

I am vastly mostly positive reinforcement. However I do use snippets of other bits when the situation calls for it. Realistically, unless you stick your dog into a lab cage and keep the human and interactive/social aspect out of it as well, how are you going to keep everything to positive reinforcement? Even the act of putting a leash on, if the leash has any pressure on the dog's neck, imposes something other than positive reinforcement. And as pgm has raised, there is more to dogs than acting simply as stimuli processing automata. They have emotion and sociability and lots of other things going on in there - exactly what I cannot know for sure as I can't see inside!

I choose my balance to be strongly on the side of positive reinforcement, and I keep my negatives as mild as possible.

Edited February 10, 2005 by sidoney

[Bonniescot]

Sidoney........ :D

You explain things well!!

[pgm]

Sidoney: "They have emotion and sociability and lots of other things going on in there - exactly what I cannot know for sure as I can't see inside!"

PGM: actually sidoney you can know what is going on inside a dog's head to same extent that you can know what going on inside a humans head. The process of understanding what another human being is thinking is the same process for a dog. Simply put, you read the external cues to infer what's going on inside. Human language is such an external cue.

Of course, most of us can do this much better with humans than we can with dogs, but then most of us have infinitely more experience with humans than with dogs. We also belong to the same species with other humans, so that makes it easier as well.

But whilst I do not claim this ability to read what's going on inside a dogs head, I do believe that the more experienced a trainer is, and the more experience s/he has with a wide variety of different dogs, the better they become at reading a dog's mind, and the more accurate their judgements become.

[Guest Tess32]

PGM: actually sidoney you can know what is going on inside a dog's head to same extent that you can know what going on inside a humans head.

- This may be true for experienced dog trainers, but for 90% of the population, it isn't. We can use external cues to interpret human behaviour because we have ourselves and our internal state as a state of reference.

A dog that hangs its head and runs off when you get home after it ripped up your couch 5 hours ago LOOKS guilty and if we use external cues, guilt would be the emotion I'd call that kind of behaviour. But would most experienced dog trainers conclude guilt, or just claim it is a dog that is expecting punishment and thus, acting submissively?

I do believe that the more experienced a trainer is, and the more experience s/he has with a wide variety of different dogs, the better they become at reading a dog's mind, and the more accurate their judgements become.

Actually I think very often it comes down to the trainer's bias. A purely positive, OC fan (and that aint Orange County!) will interpret the dog very differently than a trainer who adheres to the dominance model no matter how many dogs they see.

Nat