Simplifying Dog Learning Science -10 FAQs

[DrCharlesBortellPhD]

maybe this will help with terminology...
http://mysite.verizon.net/vzezw4q7/id11.html

[Rainmaker]

Also, for you breeders out there...Dr. Luescher also told us that there has been research that has shown "Mild stress of the neonatal dog (0-10 days) such as induced by handling or by placing the pup on a cool surface [for approx 30 seconds] will increase the pup's ability to cope with stress later in life."
I thought this was very interesting and may be helpful for those who breed to do this for the first 10 days of the newborns life. I'm not sure where he got this information from, but he is an excellent behaviorist and I would trust him with any of my dogs!

Sorry so long....if this was useful to atleast one person then it was worth it!

What it sounds like you are referring to possibly is the Early Neurological Stimulation program used by the U.S. Military in their canine program. There are many references to it and many breeders have been doing it for quite a while. There is a specific start/stop date (day 3-16) and 5 specific steps performed once a day, the last one involving a cold towel. But really don't want to get off topic of Dennis's most excellent thread.

[mbcorsini]

Excellant thread Dennis. Your insight is always welcome. Love your training alone articles.

Mary Beth

[moscowitz]

I guess I'll have to follow Cotton Pershall, "couldn't train a dog by reading a book, that dog training is 99% hard work and persistance, " "That the retriever trainer has to have energy, dedication and sensitivity.

And I think he also said a good trainer knows how to make a correction
A great trainer can read the dog so he doesn't have to make the correction.

So I don't think he was much into make dog training anal.

[Daniel]

Dennis,

I think some of your explanations of the four possible avenues for acquiring behavior aren't quite accurate.

Operant Conditioning, developed by Skinner, describes learning through either Positive or Negative Reinforcement. It occurs when a behavior is either reinforced with a
reward, meaning something valued or liked by the dog, or it's discouraged by punishment, meaning something the dog dislikes.

+R
Positive Reinforcement is something you add that the dog likes, and it always follows the behavior you are seeking. The Positive Reinforcement is something added that will strengthen or increase the behavior it follows. You can add praise, food, play, petting, a retrieve, a fun bumper, etc., as a reward, so long as it is something the dog likes.

-R
Negative Reinforcement is the removal of something the dog doesn't like, and it also comes after the behavior you are seeking. So when you squeeze something sharp into a dog's ear and then jam a bumper into it's mouth when it yells in pain, you're using -R. When the dog will hold the bumper for a second, which is the behavior your seeking, you cease pinching; thereby removing something the dog doesn't want. Removing something that follows the behavior you're seeking, was for Skinner, subtracting something. In this case you're subtracting the infliction of pain.

So Reinforcement, whether positive or negative, both strengthen the behavior your seeking. And both follow or come after the behavior your seeking.

+P
Positive Punishment is something the dog doesn't like and it always come after the behavior. It's purpose is to decrease the behavior it follows. The most common punishment used by the force method is burn the dog. So if the dog is returning from a long blind and tiring decides to swim to shore and run the rest of the way on land, the punishment of shock is added after the dog's undesired behavior. Punishment is something dogs dislike and is intended to decrease undesirable behavior.

-P
Negative Punishment is something subtracted or removed, that the dog does like. This also follows the behavior. The biggest example I can think of is one that many trial and hunt test trainers screw up. At a training session, a trainer brings their dog to the line. The dog acts like a wild bronco, won't heel, and is amped-up on the line, yet the trainer calls for the bird and sends the dog. The dog viewed the retrieve as highly desirable. In that scenario the dog was rewarded with a retrieve for lousy behavior. Being rewarded will increase the likelihood the dog will behave the same way in the future. Negative Punishment, on the other hand, might have been a more effective choice. Subtract something the dog wants or values by removing the opportunity for the retrieve. Skinner called that negative, because he took something away, as occurs in math when something is subtracted, and he labeled it punishment because the dog didn't like the consequences that followed his behavior.

Punishment, whether positive or negative decreases undesirable behavior. Both always follow the dog's behavior, you're not seeking.

The retriever community overwhelmingly uses the force method and relies on -R, +P, and -P. Notwithstanding it's popularity, it has built in short comings that require trainers to spend more far time than necessary, often for poorer results, than would be the case if they used the entire theory. As such we lag behind almost all other breed sports in our training method. All things being equal +R takes less time to teach behaviors, is more motivating, generates fewer training problems, and creates a far deeper bond between trainer and dog.

I surprised, but glad, there has been so much interest in your post.

Daniel

[Daniel]

I'm saying it's positive reinforcement also.

You guys are leaving out an important criteria; to be positive reinforcement it must be something the dog likes or wants. I never met a dog that liked getting juiced.

Daniel

[Bridget Bodine]

Daniel I believe you are forgetting about all of the retrieving these dogs get to do!! THAT is +R......Why don't you leave your agenda off the keyboard

[RetrieversONLINE]

Dennis,

I think some of your explanations of the four possible avenues for acquiring behavior aren't quite accurate.

Operant Conditioning, developed by Skinner, describes learning through either Positive or Negative Reinforcement. It occurs when a behavior is either reinforced with a
reward, meaning something valued or liked by the dog, or it's discouraged by punishment, meaning something the dog dislikes.

The retriever community overwhelmingly uses the force method and relies on -R, +P, and -P. Notwithstanding it's popularity, it has built in short comings that require trainers to spend more far time than necessary, often for poorer results, than would be the case if they used the entire theory. As such we lag behind almost all other breed sports in our training method. All things being equal +R takes less time to teach behaviors, is more motivating, generates fewer training problems, and creates a far deeper bond between trainer and dog.

I surprised, but glad, there has been so much interest in your post.

Daniel

Daniel

Since I have no objection to your definitions, I am not clear what you find inaccurate. Perhaps, one of the issues with examples is that in actual field application, they often employ more than one of + or - R or P. This isn't a rat lab!


More importantly, I am not sure who in the retriever training community you train with but the use of positive reinforcement is extremely common in the training I am familiar with. I'll wager that +ve R is used 5-10 times more than -ve R or P in training advanced dogs. In Basics and Transition, there is a lot of conditioning and reinforcement especially with -ve R. But, some advanced dogs may go weeks without an aversive but receive all kinds of rewards-most notably as Bridget said-retrieves and "good dogs". Having said that there is no compelling evidence that +ve reinforcement is more powerful than -ve reinforcement to increase behaviours but I do agree both are required for the best partnership. What distinguishes the retriever community is that they DO use all 4 unlike a lot in other breed sports.

I would hope that you don't think a high desire retriever can be only trained with +ve R for succesful competitive field sports like field trials. If so, I have not seen the evidence.

[Daniel]

Collar Conditioning:
Dog knows sit and here. Now we are adding pressure to reinforce sit and here and make it more solid by saying sit "nick" sit, or Here ' nick' Here is this positive reinforcement because we have added some pressure? or is it positive punishment???

Debbie,

As trainers we can get so locked into what were doing that 'we sometimes don't the forest for the trees'. The modern force method uses an e-collar as it's primary training tool. So in order to make the e-collar effective a dog has to be taught to understand the collar's significance, and how to turn off the collar. It's not about enforcing sit or here. Don't you agree? Rather than get bogged down in an endless discussion about which element of Operant Conditioning is being engaged, like so many others have done, in this instance why not just give it a pass, because the use of a collar is always aversive to a dog, so it's a bit of a moot point.

Instead, once the dog fully understands the collar, take a behavior, any behavior, and then pose your question. Then it will be a much easier for you to answer your question. Because dogs always dislike even the mildest nic, it will never be +R (positive reinforcement) therefore it will be one of the remaining three choices.

The larger question that's not been asked is: why is it that the hunting community is so far behind other dog breeds, and in fact all other animal training, in adapting newer training technology?

Daniel

[Ken Bora]

.... The modern force method uses an e-collar as it's primary training tool......
Daniel

sure of this you are?