I am with SCH3 and Ringsport on this one.
Sooner or later a dog is basicly completed with training. If the dog has had a proper mix of compulsion and reward the dog will be reliable.

Any decent working dog will complete training and at that point will work without immediate consequences. I can take my Giant and work him in a group obedience for an hour off lead for little or no praise and no corrections. Now if I don't work him in obedience for a long time it will take a few corrections and some praise to tighten the behaviors back up, but after that it will go back to just plain old off lead obedience.

 Quote:

Originally posted by Ellen:
.... when it knows it will not be rewarded with food, toys, etc. and/or when it knows it can not be physically corrected(compulsed) or punished if it does not perform?

That is the part of the question that confuses me. How does the dog in your example "know" of anything, be it reward or aversive, if it has never been introduced? Do you mean a dog that will work just for the sake of working, with no rewards or aversives? Or do you mean a dog that will perform even when the rewards or aversives that have been used in his training up to that point become absent?

In the former, sure I think it's possible.....depending upon what the work is! If you're asking a retriever to retrieve, that's work and the reward all in one. Same with herding dogs and herding, spaniels and flushing game, bloodhounds and trailing, etc. You don't have to teach the basic behavior. The dog does the work because it gives him pleasure. Rewards and aversives come into play when you need to modify the behavior to fit more specific needs.

In the case of the latter, I'm with Richard and the others.

Hmmmmmmmm. . . .

Ellen what are you thinking here? Does you question have some root in your experience in herding? Just curious. . .

Originally posted by L Swanston:
Or do you mean a dog that will perform even when the rewards or aversives that have been used in his training up to that point become absent?
Yes. I mean a dog that will work after it has been trained for whatever work it has been trained for.

 Quote:

Originally posted by VanCamp:
Ellen what are you thinking here? Does you question have some root in your experience in herding? Just curious. . .

I am basically just curious about "where people are coming from" when they train their dogs. Like most things, this curiosity has evolved -- it has its roots in all phases I've worked thru with my dogs from pet to obedience to schutzhund to herding. But most of all, I just think it is an important state of mind to be aware of when one trains a dog for a variety of reasons.

Maybe this will clear up a few confusions. Learning a behavior and correcting a behavior are accomplished in two different ways. Learning a behavior is accomplished through reinforcing a desired behavior. Corrections or correcting a behavior is accomplished by either using physical force to compel it perform (compulsion) or by withholding reinforcement. They should never be confused. Compulsion reduces drive and works along with your dog’s avoidance thresholds. This is why it is always paired with a motivator or drive building technique. The phrase Drive-Compulsion–Drive comes from this. Done correctly it looks like a series of peaks and valleys similar to an EKG heart reading machine. While the concept is easy to understand, learning to work within the boundaries or thresholds of avoidance takes many years of experience to master. Click here to view a prior post on giving a proper correction (8th post down). Withholding reinforcement on the other hand does not work with your dog’s avoidance thresholds so it does not reduce drive. Why even consider compulsion then? Well for many reasons. All dog can be worked in compulsion, not all dogs can be worked by withholding reinforcement. The motivator (drive) for the reinforcement must be greater than the distractions that are present or else the distraction becomes the motivator. Also behavior extinction by withholding reinforcement is a very time consuming and frustrating method. Lastly remember that dogs are pack animals. They work for the pack. Compulsion is a pack-oriented exercise where the dog is working to please its pack leader. In withholding reinforcement the dog is working for itself (drive satisfaction). I have seen dogs that lack focus while working with handlers they do not respect turn into a new dog in the hands of respected alpha handler. This is because they are now working for the pack. Hope this helps a bit.

 Quote:

Originally posted by Vince P.:
Learning a behavior and correcting a behavior are accomplished in two different ways. Learning is accomplished through reinforcing a desired behavior. ...correcting a behavior is accomplished by either using physical force to compel it perform (compulsion) or by withholding reinforcement. They should never be confused. Compulsion reduces drive...

Vince, it sounds like you are saying that learning ONLY occurs thru positive reinforcement, and not thru "corrections" or compulsion. Or am I misunderstanding - are you just RECOMMENDING that people follow that route: positive reinforcement first, then compulsion after the behavior is learned? Cuz learning absolutely DOES occur thru negative reinforcement (escape training) and avoidance conditioning, which most people would refer to as compulsion. The key is that all of these paradigms do reinforce the desired behavior; some are just more pleasant for the dog than others, but learning will occur with all of them. And compulsion or avoidance conditioning doesn't necessarily decrease drive, it may simply change which drive the dog is working in. An animal in a negative reinforcement or avoidance paradigm can be HIGHLY driven to perform and thus get rid of/avoid the aversive stimulus. A dog trained really well thru negative reinforcement and avoidance can look every bit as flashy and motivated as the ball crazy dog, and in fact he IS just as motivated - he's just working in a different drive.

I got that Vince was saying that learning occurs through reinforcement; he did not use the terms positive nor negative. I tend to agree with what he's saying, as I understand it.

I agree that learning occurs through reinforcement (both postive and negative). But not all dogs learn equally well with each method.

My young bitch simply does not respond well to correction. In fact, it seems to make her more stubborn. She simply refuses to be jerked around and refuses to avoid a leash correction.

When she was a pup, I tried to get her to down (platz) using various coercive techniques...leash guidance, re-positioning the front legs, even light pops on the leash. The stronger the coercion, the greater the resistance... using a treat gave immediate results.

It seemed as though she was saying you can't MAKE me do anything... unless you make me WANT to do it.

Sch3FH2:

“Vince, it sounds like you are saying that learning ONLY occurs thru positive reinforcement, and not thru "corrections" or compulsion.”

I feel compulsion should never be used to teach a behavior. Giving a correction before the dog understands what to do is not conductive to learning.

“compulsion or avoidance conditioning doesn't necessarily decrease drive”

Compulsion always reduces drive. The result of the dog’s temperament or properly pairing a drive motivator with the compulsion can increase the overall drive in the behavior. But the act of compulsion always reduces drive.

Dave Lilley:

“My young bitch simply does not respond well to correction. In fact, it seems to make her more stubborn. She simply refuses to be jerked around and refuses to avoid a leash correction.”

This goes back to my threshold statements. The compulsion must be hard enough to compel compliance without crossing the threshold into avoidance.