Quote:

Originally posted by Dennis Hasley:

 Quote:

Hi Ellen,

Can you explain to us how the shepherd sets up the boundry and enforces the rule in most cases?

I can tell you how I do it as I learned from Manfred with dogs of his breeding that were selectively bred for this work. Other shepherds and other trainers do it differently I'm sure. <img src="/ubbthreads/images/graemlins/laugh.gif" alt="" />
I assume you want to know how the starting dog is taught the boundary so I will explain that first. I look for a natural boundary around the area I want grazed. It can be a low stone wall, footpath, road, tire track, rut, change in grass length (mowed strip) -- anything that the dog can pick up as a track. I take the starting dog with me on a shepherd's leash (one that goes over the shoulder so my hands are free) while I lead the flock to the graze. The experienced dog is working free keeping the flock in line/together on the way. Once the sheep are in the area I want them to graze in and the experienced dog has picked up the boundary, I stand with the young dog on the boundary and watch the sheep. I do not pay any attention to the youngster other than being aware of how it is behaving -- ie. is it calm or antsy; is it watching the sheep; is it watching the other dog; is it attentive to what is going on with the sheep & the other dog working the boundary; etc.. When the young dog is settled, is responding consistently to my voice and is watching the sheep, I snap on a long line and let the dog go (only keeping hold of the end of the 50' line). I don't say or do anything. I just stand on the boundary watching the sheep in the graze ignoring the dog. What usually happens is that the dog will start to move back & forth along the boundary watching the sheep & seeing how much freedom it has. I start to watch the dog more closely now -- not staring at it but out of the corner of my eye. If I see the dog starting to move inside the boundary, I use my voice as quietly as possible (it varies with each dog) saying "uh-uh" and indicating with my hand (pointing to the boundary) that I want the dog to move back onto the boundary which is should do. Once back on the boundary I say "good boy/girl". This goes on until I see that the dog is moving on the boundary without my help. It can be as fast as 15 minutes or up to an hour -- some dogs like to test more. Some dogs also like to try to run into the sheep (which is what you are really looking for I would guess <img src="/ubbthreads/images/graemlins/wink.gif" alt="" /> ). If that happens I use a strong voice command "NO!". If that stops the dog, I immediately use a softer voice & say "good boy, come on here, on the border" and indicate with my hand again -- then when on the boundary brief, quiet praise again. If the dog doesn't stop when called, I let it hit the end of the long line good and HARD. I let the dog think the correction came out of the sky by acting as if nothing happened & just call the dog back "good boy, come on here, on the border". If the dog doesn't come back I stand firm & call him back again with a stronger tone. I prefer to let the dog decide to obey on its own rather than make a force issue out of it (dog's restricted on the line anyway). I have never had a dog not respond as described in either the first or second example. I have also never had a dog not pick up the boundary immediately. Once the dog seems to understand that it is not to go inside the boundary (usually within a few minutes to half an hour), is moving back & forth watching the sheep and responsive to my voice direction, I drop the line & let the dog drag it. The dog is free to explore the boundary as long as it respects the rule and responds to my voice directions if I have to give any. If the dog starts getting out of control, it goes back on the short leash to watch -- it loses its freedom. I think I have an article on my web site that explains it in more detail.

You will note that I do not want to stimulate the dog's drive. I want the dog calm and quiet and observant. I want the drive to develop naturally. I want the dog to think about what is going on and what it is doing -- not just react.

Ellen

Thanks Ellen,

This is something I expected you to say given your experience and you references to training and it sort of gave me a way to bring the subject full circle and back the original “original” post of treat or not to treat.

The issue: To treat or not to treat? It is always wise to reward the dog because dogs work cheap but not for free. It is just as important to know what when and how we are rewarding the dog so we can put to use those tools. In the end it really comes down to the individual dog and trainer. Moreover it is about experience and working of methods that have real application to the job or the training you wanting the dog to eventually do. I have found most experienced trainers who can produce consistent on several dog results will not shy away from correcting the dog when warranted. Yet the ability to read the dog is an interesting and intricate part of any good training program I think.

The technique Ellen described:
************************************************************************
“If I see the dog starting to move inside the boundary, I use my voice as quietly as possible (it varies with each dog) saying "uh-uh" and indicating with my hand (pointing to the boundary) that I want the dog to move back onto the boundary which is should do. Once back on the boundary I say "good boy/girl". This goes on until I see that the dog is moving on the boundary without my help. It can be, as fast as 15 minutes or up to an hour -- some dogs like to test more. Some dogs also like to try to run into the sheep (which is what you are really looking for I would guess ). If that happens I use a strong voice command "NO!". If that stops the dog, I immediately use a softer voice & say, "good boy, come on here, on the border" and indicate with my hand again -- then when on the boundary brief, quiet praise again. If the dog doesn't stop when called, I let it hit the end of the long line good and HARD.”
*****************************************************************

This demonstrates what most of us involved in this discussion and others have seen on other post. It is just in a different context. Ellen lets the dog make the error so a MOTIVATIONAL correction can happen. The dog is not focused on the handler so the “God given correction” while directional is not associated to the handler as much as it is to the behavior. The dog “feeling the force of compliance” is rewarded by no further corrections and verified no more are coming in the safe zone signaled by the handlers “Good boy!” All trainers know to much of anything is never good for their training goals

Because of genetic inclinations some dogs are more compliant to a specific task and some dog are a little harder to convince there is any task other then their own agendas.

At least this is how I see it.

 Quote:

Originally posted by Dennis Hasley:
Thanks Ellen. This is something I expected you to say given your experience and your references to training and it sort of gave me a way to bring the subject full circle and back the original “original” post of treat or not to treat.

I know -- you probably expected me to illustrate a correction <img src="/ubbthreads/images/graemlins/wink.gif" alt="" /> and Laura probably expected me to illustrate praise. I try to please everybody. <img src="/ubbthreads/images/graemlins/laugh.gif" alt="" />

The P+ P- R+ R- and primary/secondary reinforcement formulas make sense, but IMO there is a "relationship" issue involved which is an unavoidale and very important part of training also. IMO one can not have a working dog without a working relationship. IMO a purely command/control dog is more a machine than a working dog -- but that is a whole other discussion. To illustrate this point I will comment on your analysis of my correction on the boundary.

 Quote:

Ellen lets the dog make the error so a MOTIVATIONAL correction can happen. The dog is not focused on the handler so the “God given correction” while directional is not associated to the handler as much as it is to the behavior.

First, I want to make it very clear that I do NOT believe in setting up a dog to make a mistake so that I can give a "motivational correction" or any other kind of correction -- certainly not in the beginning stage of training for sure which was the example here. Setting up a dog to make a mistake is just a formula for teaching a dog not to trust you or, just as bad, to be dependent on you. I only try to set up training situations in which the dog will SUCCEED -- achieve success on its own.

Second, in my example you are correct the dog is not focused on the handler, but the dog is not focused on its behavior either. The dog is focused on the sheep when it gets the correction. Now here is where it gets interesting so I have a question for you. <img src="/ubbthreads/images/graemlins/smile.gif" alt="" />

Problem: (1)In the learning stage which was the example here, the smart shepherd does NOT want to inhibit the dog's drive/interest in the sheep so, therefore, the smart shepherd should know better than to correct the dog for the behavior that comes naturally out of that drive (in this case the running in to chase and/or grip); and (2)the smart shepherd does NOT want to lose the dog's trust or respect either so proper corrections are crucial.
Question: What does the shepherd want to correct the dog for? Or what does the shepherd hope the dog will learn from this experience?

In the early learning stages, when the dog runs to the flock and hits the end of the line, he's going to associate the aversive with his immediate behavior. It's simple +P. You said "I do not want him to feel he is being punished for chasing the sheep when he is being corrected for not listening." But he doesn't even really know yet what it is you are asking of him. What he does find out is that there are unpleasant consequences for going THERE, while there seem to be positive reinforcers for staying out further. Better to stay out further. Even if it's only a verbal "uh-uh" when he shows an inclination to get closer, that's still +P. I don't believe that after a dog hits the end of a line hard, he thinks to himself, "if only I had listened to her back there...". He associates it directly to the behavior at that moment - chasing the flock.

Problems 1 & 2

Proper Time & Proper Timing = Goal

 Quote:

Originally posted by Lee Baragona - Sch3FH2:
In the early learning stages, when the dog runs to the flock and hits the end of the line, he's going to associate the aversive with his immediate behavior. It's simple +P.

I disagree. He associates the aversive with the sheep. He "feels" the attraction to the sheep and is reacting instinctively(behavior) to that feeling. When he feels the correction his positive feeling toward the sheep is suddenly shocked into negative. The behavior is merely incidental at this stage since he could be expressing his attraction(prey drive) in a variety of other ways besides chasing as well.

 Quote:

You said "I do not want him to feel he is being punished for chasing the sheep when he is being corrected for not listening."

That's right. That is why that kind of correction is counterproductive IMO. I don't want him to associate bad feelings with the sheep which this kind of correction does. I want the sheep to increase in their attraction value, not decrease. At this stage he is probably not even thinking about his behavior -- he is just doing.

Of course he hasn't learned yet that the correction is for not listening -- only *I* know that at this stage. The dog will only understand what I want from the correction (voice or restraint) when he figures out that my voice "doggie, no" preceeds (is an early warning) that a P+ (to use your term) is about to happen if he doesn't stop what he is doing and listen. Pretty quickly he learns that the "uh-uh" means "wait a minute (stop motion) let's look & see what she wants me to do" and that the "good boy" means "you're doing fine keep it up".

 Quote:

But he doesn't even really know yet what it is you are asking of him.

Exactly! Again that is why this kind of correction is counterproductive IMO.

 Quote:

What he does find out is that there are unpleasant consequences for going THERE, while there seem to be positive reinforcers for staying out further.

Again, I disagree. The dog feels initially that there are unpleasant consequences for being attracted to the sheep. That is what he feels at this stage from the correction. To the dog there is no THERE there -- yet. He is REacting purely on instinct with probably no awareness of anything other than what he feels -- and what he feels is attraction to the sheep.

 Quote:

Better to stay out further.

IMO he doesn't even know that yet. All he knows from that correction at that moment he feels it is that his atraction to the sheep which he felt as positive is now felt as negative. As I said -- this is counterproductive for my purposes. The dog has NO feeling about here, there or anywhere as it applies to this lesson -- YET.

 Quote:

Even if it's only a verbal "uh-uh" when he shows an inclination to get closer, that's still +P.

Not really -- it's a preliminary +P meant to get the dog to start "thinking" a P is coming if he continues what he is doing so "I better stop what I'm doing & pay attention". I don't care if it is a P+P-R+R- as long as it allows the dog to figure out how to stop & pay attention.

 Quote:

I don't believe that after a dog hits the end of a line hard, he thinks to himself, "if only I had listened to her back there...".

Of course he doesn't think that. He is not thinking at all -- YET. He is reacting out of instinct and that is what I want to KEEP. I want him to LEARN to think(make choices).

 Quote:

He associates it directly to the behavior at that moment -- chasing the flock.

Again, I disagree. He associates it directly to the positive feeling he has to the sheep -- the behavior is secondary.

Teaching the dog about accepted behavior and acceptable working distance is about timing. A person exerting any type of influence the dog feels is negative is a correction to him.

Just because the sheep are there and the dog likes to work them dos not mean he will outright reject his nature and not like them when give one or two negative influences. Instead he adapts and listens to the directives of the handler/shepherd because he does not like to be in disfavor. To some dogs an “uh-uh” is enough to inhibit the behavior but this influence is learned somehow and often it was a point of conflict that was well timed and motivational enough for the dog to think twice about not obeying the word of the shepherd. Does this mean the dog lives in perpetual fear.lol…of course not. I correct my dogs I have my clients correct their dogs too and all the dogs are happy workers with their tails wagging. It is all about timing and associations thy made during the learning phases. In this phase they learned right/accepted behavior also = rewards. Multilevel rewards I might add even though they were corrected for misbehaviors. Dogs as a social pack animal understands control and hierarchy so as long as the time and timing are correct the dog can associate it behavior. Only when no direct association is made or the amount/type of aversive (correction) is used does the dog have signs of fear and confusion. This ultimately leads to a drop in drive because the dog sees no way to avoid his situation and to be able to continue to work his desire. Granted the dog could have a negative association to the sheep but dogs being resilient are slaves to their drives. The right dog with the drive will work through it and if another association can be made such as proximity or location it will. Consistency is the key there and the signals without the correction help make the association simply easier as long as the dog understand what “uh-uh” means.

Now about motivational corrections…
People all the time misunderstand motivational corrections so to define:
A Motivational Correction is enough influence/correction the dog would avoid it again. This is individualized and based on the dog’s own temperament. Does it mean a hard snap on the leash? . For some dogs, yes it does, but for others not at all. Each dog has its own level of what it deems motivationally defining IN PART its tractability.) So when you prevent the dog from working because of an action you took; either a corrective jerk or a signal that a correction is coming if the working distance is not adhered too.

Lets not forget “dog language” in the form of posturing and proximity and even if the handler does nothing to the dog but come close to the dog for a beech of behavior the handler is speaking pooch 101 with domination that the dog can associate to the behavior or in time the area it is working it. Boundary training

The dogs used in this case are no doubt tractable and wanting to please the owner. It was even state they are specially bred to be such animals. So naturally the whole process is made easier with les conflict due to the nature of the animal. Like all jobs selection of the right dog with the right drive and temperament to do the work is the “real” magic of any specialized area in the dog world.