I think treat training is cruel. - Page 27

Pedigree Database

Premium classified

This is a placeholder text
Group text

Premium classified

This is a placeholder text
Group text

Premium classified

This is a placeholder text
Group text

Premium classified

This is a placeholder text
Group text

Prager

by Prager on 05 December 2018 - 16:12

So I am little confused here. I being worked over here by more or less harsh statements that "I do not get it' and that  Marker training, for example, is somehow different from a training done before trainers started running around with clickers and started to call it modern revolutionary method.

Everywhere I read and watch videos,  I am getting this about clicker as an example of marker training." Clicker training is a form of marker training.   Marking is when we send a signal to a dog that will be associated with whatever behavior that was just performed.  The signal will be used to either encourage or discourage that behavior.   When the signal is used to encourage a behavior it is called a "positive marker".  Positive markers are usually what is being referred to during marker training.  "Negative markers", also referred to as conditioned punisher." 

 I like this summation ( above) because I think it is the best one I found. You can read here:  

https://dogtraining.world/knowledge-base/conditioned-punishers-vs-unconditioned-punishers-dog-training/

 NOW,....Can someone please explain to me how is this marker training any different from let say Koehler training? The way I see it Koehler training is using 4 quadrants of conditioning. The first - learning - part of Koehler by more or less not unpleasant means explains the dog what command means and then when the dog responds he is immediately rewarded with positive voice ( same as marker training) and pet or food ( basically same as marker training). Now I will admit that learning faze of Koehler can be performed quite harshly it the trainer is an idiot,  but IMO and experience such harshness is not necessary and trainer with an understanding of thresholds will proceed without any harshness.  

Now at first, seemingly the difference seems obvious when the dog does not respond to the command in subsequent stages of training which follow after the learning stage. In Koehler there we use positive and negative punishment where marker I was told by many - uses only negative punishment where in neg. punishment the reward is withdrawn until the dog performs. But then I read JJ ( and others) saying that    and I quote: "I can add a ecollar stim,
I can add a yank on the leash, yank a prong, or yank on a choke collar.
I can add the above and increase them in duration or strength.
I can add my harsh voice."

OK fine. If I consider all this then I do not see any difference in principle and very little difference in methodology between marker training and the Koehler method of training. They both use all 4 quadrants of conditioning in an appropriate time. 

  So my question is, how is this so-called new, modern, revolutionary marker method any different from so-called Old School and by marker trainer despised and described in derogatory terms method?

 I am not interested in being called fool and worse. Just the facts, please. 


by apple on 05 December 2018 - 16:12

I have "The Koehler Method of Guard Dog Training" right in front of me and all the obedience is based strictly on compulsion. There is nothing in the book about teaching the dog how to learn a new behavior via positive reinforcement. For the down, you jerk the leash to the ground and, "Make certain the first correction does the job. For the recall, you pull the dog to you on the leash. For the sit, "One good correction will do the job where many tentative ones fail." For heeling, you change direction 180 degrees with a correction. There is not place in his method for the dog to learn via positive reinforcement. It is all positive punishment followed by praise.

by joanro on 05 December 2018 - 17:12

Research how porposes etc are trained starting in the sixties.
Trainers discovered " marker training" worked like magic on the equatic blowhole mammals.
Then places like Bush Gardens used the method to teach behaviors to pelcans and maccaws, etc. Then they began using the method to teach new behaviors to elephants, in stead of the traditional ropes and chains and forced training.
Then dog trainers picked it up.
Marker training with a clicker is new relative to how long animals have historically been trained.
What you don't seem to grasp, is that the marker training purpose is to " teach the new behavior" to the animal, no matter the species.
Then, and only after the animal has demonstrated an understanding of the behavior to be performed, is the
training with positive negative reenforcment applied.
However, animals such as aquatic mammals and pelicans, the clicker and food reward is continually employed....why? Because there is not an effective way to smack a purpose while it's swimming submerged, or a flying pelican to make it land on a designated spot.

Land mammals can be corrected for errors more efficiently and so other training tools such as ecollar or throw chains or what ever, can be employed during proofing for dogs, horses, mules, etc.after they have thoroughly learned the behavior, but not before they learn it.

As for the difference between Koehler method in TEACHING the new behavior, he is advocating forcing the dog by phisical means, into the desired behavior.
Clicker or marker training, waits for the dog to perform the behavior on it's own....as in trial and error. But never forced during the learning phase.
That is the difference...Koehler is phisically forcing the behavior
Clicker or marker is trial and error by the dog.

Example: teaching the sit.....Koehler method: using chain or prong collar, lift with the leash to cause discomfort and push down on the butt to cause the hind legs to collapse into a sit position...immediately the discomfort is removed by slacking off on the leash ....that is a reward/ negative reinforcement...add praise/ positive reinforcement. Rinse and repeat till the dog sits before discomfort is applied.


Clicker or marker training the sit: hold some treat in the hand that the dog can smell and has already been taught that food comes from the opened hand....hold above out of reach and let the dog or puppy jump around, whine, paw your leg run circles, anything he wants to do trying to get the food....never say a word, don't respond to all the shenanigans just let the pup run his course till he gets befudled and out of frustration and out of ideas, he sits down and looks at you....immediately open hand and give the treat as a reward.
Repeat after a short break, then you'll see next time the pup will not go through his whole repitoire of shinanigans before he sits......each time the pup with sit sooner and sooner.
I've had pups smart enough to sit almost immediately the second time.

So that's the difference...one is phisical force, the other is letting the dog or pup figure out by using his brains how to get what he wants.


Prager

by Prager on 05 December 2018 - 17:12

Joanro said in response to my question why do you think the dog is coming to you now when you have no treats? her Answer: Because I taught him when he was an itty bitty puppy and made it stick by using treats. Teach them very young and they don't forget. The recall becomes second nature and they want to do it. He "learned"...which is the goal of "teaching". The kibble reward was an enforcer for the desired behavior. Once he " learned" the appropriate response, no more kibble necessary. Simple training basics."

Prager: So you are saying this is not a part of conditioning?  Dog just learned and now he does it forever. Yes, I understand. You call it being second nature and you are talking about being imprinted. OK, I get that too. Imprinting is great but it is phase-sensitive learning (learning occurring at a particular age or a particular life stage) which you also have mentioned. Such training is truly wonderful because it is rapid and apparently independent of the consequences of behavior. However, imprinting by observations of scientists and other observant people has a "critical period". This critical period is when the pup is very young ( as you say). After the pup gets older then such learning becomes more difficult if not impossible. 

Konrad Lorentz used basically the same principle sometimes in 1930ies with geese where the gosling imprinted itself onto a first moving object - which was Konrad L. himself.  That is called filial imprinting. I guess you can do the same thing. I have observed mother chicken scratch a groud and then when they find something the mother chicken makes calling sound and her chicks are coming to her running just like your dog to you with a treat. I do the same thing when I feed pups from 3 weeks weaning time I always call the pups to come repeatedly  just like you or the mother chicken. ( I love chickens) I also fire a small caliber blank gun or crack a whip to teach then not to be gun shy and so on.

 I will give you that the dog who is trained this way develops an exceptionally strong bond with the surrogate mother- trainer. Such training creates packing instinct and it is a wonderful way to train a dog.

This is great but unfortunately, an imprinting principle has a "critical period" and thus it is not possibly done with older dogs. There we need to use conditioning, and the problem with conditioning is that if it is not reinforced then the conditioning extinct or is extinguished.  

 Joan, You must be disappointed that my post was not adversary to your post or to you but instead was complimentary. LOL Gotcha! 


by joanro on 05 December 2018 - 17:12

Prgre, there you go, in your first paragraph putting words in my mouth. I never said I did not condition the puppy to respond to his name... As for only being able to do this with puppies at a critical age and not affective on older dogs, I disagree because I have done this very thing with grown dogs that had no clue how to respond to their name.

I have also trained grown horses to come when called usng the same method. However, the difference with horses is that one must get the horse to notice you and same with a grown dog.

Once they look in your direction, it takes a moment till they come to receive the food reward.

And again, they must already be * conditioned * to take food out of your hand.


by joanro on 05 December 2018 - 17:12

And, pragre, I'm not talking about * imprinting* on a puppy...I'm talking about teaching a desired response to the pups name being called which then becomes a trained behavior to a command.
Does not have to be a young puppy, it will work on any age animal, any species...I've done with a camel I used to own that was a year old when I bought him.
He would be out in the big field and when I called his name, he'd come pacing over to me fast as he could, to get his hand full of grain....or Doritos. ( I'd go to his pen with a six pack of beer and a bag of natcho doritos, he'd take a full can of beer a toss it back, washing down a hand full of Doritos...he' d pollish off a half a sx pack in no time and a half bag of chips...the rest were mine!)

by joanro on 05 December 2018 - 17:12

Prgre, saying I'm disappointed that you didn't bash me, is like saying I' d be dissappointed an overhead airplane didn't fall out of the sky, killing everyone on board.

Prager

by Prager on 05 December 2018 - 17:12

Apple obviously you did not read the chapter on how to teach down correctly. You claim that it uses positive punishment to teach where in fact it is using negative reinforcement where the negative is removed if the dog performs. Marker method using treats is also using negative reinforcement which I argued in my OP is negative reinforcement approach if we use treats because treats or lack thereof creates insulin and ghrelin hormone reaction called hunger pangs, which is not pleasant. We remove this negative feeling of hunger if the dog performs the command by giving him the treat. Thus clicker marker training using food and koehler are both using the same negative reinforcement.

Baerenfangs Erbe

by Baerenfangs Erbe on 05 December 2018 - 17:12

To be fair, there were dogs that learned how to get away with bad behavior, dogs with learned aggression, or learned all they had to do was to throw a major temper tantrum and they get away with everything. Dogs with escalating behavior. So I charged the clicker with them and then used straight up negative reinforcement to work through the tantrum and THEN clicked the behavior I wanted.

Because just using positive reinforcement didn't actually fix the issue since they weren't able to think or wouldn't comply to begin with. Like the Koehler method. Apply leash pressure until dog sits, CLICK and treat. It works remarkably well with out of control dogs. And once you got through to them and have a mutual understanding and respect, thats when you can start shaping behaviors and go from there.

So you can combine Koehler and Clicker training and still mark behaviors.

As a trainer, I want to have as many tools in my box as possible so I can adjust my methods. I'm not just limiting myself to just positive reinforcement or just negative reinforcement and positive punishment. I use aaaaaalll the quadrants, and sometimes every single one of them at the same time.

Prager

by Prager on 05 December 2018 - 17:12

Well Joan it was you who used word "Imprinting" and it was you who said that you have done it since the pup was 6 weeks old ( thus during critical period) . Here is your quote.:I imprinted the recall using kibble reward every time he came to me when I spoke his name since he was six weeks old. Have not used any food reward since he was three months old."

 The point is that after the critical period expires the imprinting is impossible. After that we need to use conditioning wnich needs to be constantly reconditioned or it will "extinct".
As far as putting words in your mouth I have said that it seems that you were not using conditioning because in your answer you did not mentioned conditioning. . Nevertheless, that does not make your answer any less useful or correct.





 


Contact information  Disclaimer  Privacy Statement  Copyright Information  Terms of Service  Cookie policy  ↑ Back to top