You Don't Understand +R

Hey, my post had the desired effect!!!! Now, we are actually discussing "positive" methods on an intellectual level, not an emotional one.

 
I agree to that and I agree with a lot of what you say. The principles involved are valid. And yes, there are some dogs, IMHO, that are beyond working with, especially for most people. In fairness and safety to all, those should be PTS. Ancilliary to that, I am a proponent of capital punishment. It's not the greatest deterrent, nothing is. Some people are just bad. The value of capital punishment is that it removes a danger from our lives. PTS of an aggressive dog that can't be rehabbed removes a danger and also ends the suffering of the dog in question who, by any one's estimation, must be unhappy. I most heartily agree that most people shouldn't try to physically stop a dog fight. They don't have the training, the physical abillity. Something like DirectStop will cause a momentary physical disability (non-fatal) that allows humans a window of time to gain control. And, per the example of operant conditioning, if you can train a cat, you can certainly train a dog. IMHO, dogs are pack animals and this is what makes them trainable. They want to belong and you, the capable trainer, show them the way. I think some of it is personality. For example, Spiritdog, I imagine you have a natural way with dogs that is subtextual to the training you have or all of the "tools" that you use. Dogs just look to you. Why? Because you are a dominant personality and that's not an insult. You have a clear direction for yourself and know what must be done and expect that it will be done that way. A dog senses your attitude and falls in line (I avoid the word submits) because you know the way and they are following. Occasionally, one will challenge or test the boundaries and you will correct in a way that allows the dog to learn or lead it's way back to inclusion. Not many people have that.
 
Too bad people can't do things the right way all of the time. The friend with the Great Pyrenees has done it right. First, she got the dog from a legitimate breeder and had to wait to get a pup, Lilli. Second, she started on training from day one, at about 10 weeks. Because full-grown, Lilli will be close to 30 inches tall and 140 pounds, outweighing her owner, scruffing won't be much of an option, including the fact that G. Pyr's have an extra layer of hide and fur around the neck that allows them to ward off a coyote even while being bit. She did it all with repetition and the tones of her voice, and setting boundaries that do not change. Lilli never got to sit on the couch or sleep on the bed, even as a little pup. As an adult it would be disasterous, so never allow it to start, ever. And our friend is a natural leader. She just exudes it. Even people, regardless of who they are, tend to follow her lead. And Lilli is quite happy and has turned out to be quite the guard dog. She prefers to sleep outside, as befits her breed. She's well-cared for and adoring. And knows her limits, with out an strenuous physicality from her owner. And, Lilli started out with a good temperment, not trying to mount or lord it over anyone else.
 
FWIW, I think most methods that are gentle can be successful and it gives both dog and human time to learn. As long as we can be that gentle with each other, too.

Actually behavioral psychology does not recognize self rewarding behavior (ie Skinner's original work))  The cat in the box example is a demonstration of positive reenforcement.  The behavior increased in frequency and potentially duration.
 
As to the method used to change aggression or any other behavior  using contingent reenforcement is differentential reenforcement (of one of two types  of DRO  reenforce other behavior, DRI reenforce incompatable behavior).
 
Example:  Dog jumping on visitors.
 
DRI  would be to reenforce a down.  DRO would be to reenforce a spin.  A dog that is lying down can not jump there for incompatable with jumping.  A spinning dog doesnt jump, so some thing else (other behavior) occurs.  In most nuisance behaviors, I would suggest replacing the problem behavior with an incompatable behavior.
 
The primary reason folks have difficulty with applied behavior analysis applied to training is the lack of access to a control situation.  The laboratory or marine tank are one of the only the places you can use the principles fairly cleanly since you have greater control of the environment.  So when training dogs in the typical situations you have to think differently.  You need to find the distance at which the dog can actually attend to the signal you wish to work on.  You need to do that because behavior is a complete sentence:  the Sd (discrete stimulus command)  the behavior, and the conconsequence (reenforcement, punishment or extinction).  Sorry  I think the extra terms are unnecessary but that is just a product of my age and training.  So forget the R+ R- etc as far as I am concerned.
 
If you wish to use positive reenforcement, you must first observe your dog well and determine, what does the dog naturally seak.  These observations help you build a reenforcement library or bank so to speak.  Then you need to find the critical distance, at what distances from a distraction can a dog focus on you for one of those things in your reenforcement bank.
 
How you establish behavior is a whole 'nuther discussion.

You need to find the distance at which the dog can actually attend to the signal you wish to work on.

For me, it was about 20 feet but I have extended it, over time, to 30 feet. During a barkfest with the neighbor's border collies, if I get too close, Shadow will escalate in order to defend me. At 20 feet to 30 feet, the pressure is off to immediately defend me. I can recall, though it may take a few times since he might not hear me. I have to time it when they stop to breath. What's-his-name would just wade in and start grabbing dogs. I prefer my method. I can grab my dog and he won't bite me but another dog might. So, my recall has a better chance of breaking up a situation and Shadow will recall, at least long enough for me to re-leash and lead away. And they're not always in confrontation. But it does provide a good exercise for me to work on recall on my dog, who is mostly Husky. They are known to "forget" their training when it suits them. Secondly, I don't want Shadow to think he's being punished or corrected, especially if he didn't start it and is simply defending as he sees fit. That is, I am re-directing the behavior, not trying to stop it. Re-directing allows him the room to make the choice to heed the recall, which leads to reward, so, I suppose that is providing an operant circumstance. It allows me to increase the range of recall and allows him to learn that heeding my recall is a great thing to do. The thing that he will want to do, something much better than arguing with another dog, even if only for a moment, the golden moment when, if necessary, I can re-leash or physically hold him, etc.

Also, with regard to "guardy" dogs, I would rather feed my dogs separately than have to damage my relationship with them by physically grabbing at them.

I just wanted to clarify, she wasn't guarding her food - she was guarding my children's food. They were eating a snack at the coffee table when my brother and his Golden came over. She has had many previous contacts with this dog, in our home and others, all of them positive. So I never expected this problem. I couldn't let her put a 14 year old dog down on the ground until she didn't obey a "leave it" so I pulled her away and asked for a down-stay. She was rewarded with praise after obeying the down-stay command - end of story. No yelling, screaming, hitting or kicking. I just removed her like I would if my kids were fighting and I needed to put them in their time out spots. Not positive, but certainly not abusive. If stopping an animal (or kid) from destructive behavior is abusive, I guess we better turn in Super Nanny LOL!

In order to prevent that behavior from occurring again, she is currently being trained to go to her "place" when we eat so we. We've only had her about 5 months, so we are still working on many basic commands. I have learned a ton here and the posts here have introduced me to the clicker training that is WAY better than the obedience classes we took this summer. That was "yank & crank" in the true sense of the word. This guy made the CM look like Mother Theresa. I would never criticize anyone for an occasional leash pop if it works, but it was not working for my dog. Now she is doing great with a prong that we are using on the dead ring and will gradually transition to a flat collar with the clicker. Now even my kids - who weigh less than she does - can walk her easily to and from school and they LOVE to use the clicker to practice tricks that I have already taught her.

Positive is the best way to go - I can't argue against that. But I honestly don't believe that properly used negative reinforcement is always wrong. Without getting into the confusing -P, +P, etc jargon, I guess what I am trying to say, is that I think it is impossible to use purely positive methods. Having a leash on a dog physcially restricts them. Withholding a treat for not performing a behavior is negative. Even turning your back and ignoring a dog that jumps on you is negative because you are withholding what the dog wants - your attention. No matter how positive we try to be positive, negative behaviors will get a negative response. Negative does not always equal abusive. Would you consider a parent abusive if they picked up a kid having a temper tantrum and took them to a quiet area? Heck no. But if they slapped or spanked them, that's another story.

ORIGINAL: spiritdogs
Someone suggested this thread, but I don't see the point.  

See it now?  An informative discussion about +R, but it could use more hands on advice for us less informed owners. I'm still waiting for the success stories from owners who have overcome aggresion through strictly R+ methods. For the record I do have a clicker that I have used succesfully, my Bella does the cutest sit up and beg when I tell her to "say please" which is really funny for a Pit.

Jenhuedepohl,
Instead of the "down-stay" with the clicker might I suggest something else to keep her back when you are eating. Use your body language. a stern look and tone of voice to get her to back up while you are eating. Wait for her to get all the way to the table and then move her back. It is a mental excersize for you to think that this is your space and your food and she needs to stay back. She will know she should stay back because her leader demands this of her, not because she is waiting for a click and a treat. You don't punish her, don't yell at her, don't touch her, don't get mad at her just through force of will get your message across.

I do this when I go to the door and my dogs back up and sit down when I open the door now. They know why they are doing it and they make a choice to do it because it is what I demand of them. They do it when I feed them as well although that might be operant conditioning because they know they won't get fed until they sit back and calm down and wait for me to say OK to eat. It's just that I didn't have to teach them to calm down or back up or sit or wait. They did that all at once when I used my mind and body without saying anything.

I'm gonna pick up that clicker and teach Macie to "shake hands", people love that one.

My only exposure in using professional  training methods has been Command Performance (CP) by David Dikeman and attending Obedience and Agility class at local Training Center.  CP required multiple length leashes, a choker, and throw at the dog chains (did not use these).  Its teaching was reward based (only praise) and the dog would self correct with the choker for non-compliance.  I used this method on one of my Danes but can#%92t help but think that the choker snapping contributed to the dog#%92s diagnosis of megaesophagus later in life.  The local Training Center also used reward based (food treats) and quick snapps of the choker for non-compliance.  My problem with this training was my Danes would not take ANY food treats during the training sessions so it was CP but less on the choker snapping.  Both methods resulted in teaching the Danes the basics and advance obedience.  For the agility training there was no corrections applied for non-compliance.  It was a gradual exposure to the equipment and practice, practice, and practice.
 
My only exposure to Caesar Millan (CM) was at a Petsmart foster dog showing.  One of my fosters, Sassy was being handled by another.  Another dog got too close to Sassy and Sassy growled and snapped at the other dog.  The handler immediately reacted by grabbing Sassy#%92s neck and pushing it to the ground.  I was taken back by this and before commenting I asked the handler next to me what she was doing with Sassy.  I was told she was applying CM#%92s technique in training.  I was open if this would help Sassy be cured of her fear aggression.  It happened again and the handler reacted in the same way.  It happened on more time but with a puppy and this time Sassy broke skin.  I removed Sassy from the situation.  Sassy does not behave this way at my home.  She lives with 4 other dogs and when new dogs come over for a visit, she is fine.  I am not going to create the situation just for training purposes.  The approach that I am taking with Sassy is awareness, warnings, and removal from the situations.  On another thread it was stated that dogs react from fear aggression in one of 3 ways-run, attack, or stand still (I hope I did not misquote).  Applying this knowledge, on a home visit to a potential adopter, Sassy was kept off-leash when meeting the family dog.  I took precautions just in case, but for Sassy it worked.
 
I am not sure if Clicker training would work for me since it seems to me more one on one and not for a pack situation.
 
I am exploring a better understanding of the Pack dynamics.  My two old Danes are very well trained and when I brought in a young Dane into their pack I did not have to do anything other than teach NO.  Blizzard and Drizzle showed Petro the ropes.  I don#%92t have a lot of expectation in my house other than the dogs be well socialized, housebroken, respectful of household items, and know how to play.  These are things that are a must for adopters and I believe a pack can teach.  I usually have 2 or 3 fosters that revolve in and out of the pack.  For most of the fosters, they eventually come in line.   My latest one, Marvin is having a hard time here but the pack has definitely had an influence on his behavior.  In the picture below how can Amy not come into compliance?

Well, I'm not sure how many places I have to post this, but fine, here's a good example of overcoming agression with +R only.
 
Sheba became very fearful because of a bad experience in a camp ground when she was in a (unknown to us at the time) fear stage.  The jerk in the next camper didn't like dogs and took particular delight in getting tanked up and teasing her.  Fast forward to a neighborhood drunk that she flat out did not like.  One nite, totally sloshed he kept approaching us after being warned by her and me.  He stumbled and fell into me.  She bit him.  One small puncture, but massive bruising from the power of her jaw.
 
After that nite, she became terribly fearful and terribly agressive because of that fear.  We worked with a behavorialist who used ONLY +R methods and today she is a totally different dog.  It took a LONG time and a lot of work but today she is a confident girl who will LOOK to me for reassurance when she is uneasy...but she no longer FEARS everyone and everything.
 
There's your success story.  No clicker involved.  No tricks either and minimal treating.

Jenhuedepohl,
Instead of the "down-stay" with the clicker might I suggest something else to keep her back when you are eating. Use your body language. a stern look and tone of voice to get her to back up while you are eating. Wait for her to get all the way to the table and then move her back. It is a mental excersize for you to think that this is your space and your food and she needs to stay back. She will know she should stay back because her leader demands this of her, not because she is waiting for a click and a treat. You don't punish her, don't yell at her, don't touch her, don't get mad at her just through force of will get your message across.

Actually, that is what we had been doing - and it worked quite well. She was actually on the other side of the room when she went after my brother's dog. She was upset because that dog was close to the kid's snack and knew that was against the rules!

We are now using the clicker to teach her to go to a spot in another room. Her old spot to go to while we ate was within sight, so I didn't have to leave the table. I'm using the clicker to put her in her new spot from anywhere in the house. With this dog, I can get most everyday behaviors with just body language. She sits and waits at the door because she knows that is the only way I will let her out. She lays down quietly and waits for her supper, because she won't get any unless she is calm. She sits and waits for me to put on her leash because she knows the longer she does the "happy leash dance" the longer it will take to get outside. I've been pulling out the clicker for new stuff, like roll-over, touch and place.

I'm really enjoying this thread - I like personal stories about what has worked for others.

Glenda,
That's good to hear you have worked through that issue but I think we are agreeing more than disagreeing. I am not against positive only training if that is what works for a particular dog and a particular behavior. I am totally against punishing a dog or causing it pain, dogs don't teach each other that way so why should we?

Your dog should look to you for leadership and I'm glad the behaviorist taught you how to do that. I suspect he used more of the same methods Cesar uses than you realize ( notice I didn't say "Cesar's methods" or "methods he invented"). That your dog became fearful after two incidents with drunk men is no big surprise and I bet you realize that your reactions to the situation and subsequent treatment of your dog is what caused the problem. Your same ability to instill feelings in your dog is what solved the problem and makes you able to avoid repeating that scenario. That is not +R training but very close to what many others see as good leadership.

DPU,
People should not just copy what they see on TV either and do alpha rolls which can cause more problems than they solve and if someone was doing that it doesn't mean you were watching "Cesar's methods".

ORIGINAL: jenhuedepohl
She was upset because that dog was close to the kid's snack and knew that was against the rules!

I think you are giving her to much credit there. She didn't want the other dog to get the food is all, she isn't enforcing your rules on the other dog.

We are now using the clicker to teach her to go to a spot in another room.

That is a great combination of all the tools available to us since no one way is totally right.

I will agree that my reaction to her behavior AFTER the bite did in fact fuel her behavior. And yes, I did need the help of a behavorialist.  My boys have never had issues or behaviors like Sheba did.
 
However, based on what I have read and seen of CM, no the techniques to help Sheba were nothing like the techniques that he uses.  These were based SOLEY on building her confidence in herself and on her trusting me to be the protector.

ORIGINAL: glenmar
However, based on what I have read and seen of CM, no the techniques to help Sheba were nothing like the techniques that he uses.  These were based SOLEY on building her confidence in herself and on her trusting me to be the protector.

You are probably right and I apologize for bringing him up in a thread about +R training. Perhaps Ann can explain better how leadership and +R go together because I don't really get the connection. I have a feeling there is more of a cross over in techniques than people seem willing to admit and these discussions are often in black and white terms.

As I stated before, pain, punishment, cruelty, anger...none of it is appropriate to dealing with dogs and has the opposite effect of  convincing them that we are a good leader to follow. I just don't think that with four options available we should limit ourselves to +R and even those that say they do use other forms as well. I think +R has become as much a marketing tool because it sounds nice and politically correct but doesn't cover all the things people who work under that title do. If you ignore a behavior aren't you using -R?

It seems to me that you are intent on baiting me and others.  Yes, I imagine that Anne could likely explain far better than I.  I am not a trainer, nor am I an expert on anything, other than what works for me.  As I have said, I could never even bring myself to rub a pups nose in a mess, despite the prevailing wisdom at the time that this was THE way to housebreak.  And I would rather train than to break.

I am totally against punishing a dog or causing it pain, dogs don't teach each other that way so why should we?

 
In my many decades I too have evolved into using a gentle approach.  But my pack disciplines fosters causing discomfort and pain.  I believe Danes have radar in their paws so when I see one step on foster or brush their paw on the other, there is a message going on.

I suppose the dog a size of a Dane stepping on a smaller dog would cause pain but I don't think that is the intent.