RetrieverTraining.Net - the RTF banner

Skinner vs Pavlov

42K views 210 replies 26 participants last post by  Jere 
#1 ·
This ought to be good. I could also have said Operant, vs Classical conditioning.

Which one do you consider "better", when applied to Retriever Training?
Or, if you think that OC is "better" for certain aspects of training, while CC is "better" for others, what would those differences be?
 
#96 ·
I know you're not talking to me Pete, but I think CC is instinctual in all animals as a built in means of survival. We simply leverage this instinctual ability to make associations and anticipate events by applying the principals of OC.

I think all of our dog's conditioned responses are built using this this instinct, whether we as trainers realize it or not

Darrin
There is a lot to it I guess,,,no absolutes in dog training,,,drives influence how dogs learn and assimalate. Timing/motivation and consistency tie in,, ,,but like Jim says its something most don't think about when training. Fun stuff to think about and talk about.

You and Copdoc are Rockin though,,I enjoy trying to understand it better,,,its a good change of pace.
Pete
 
#100 · (Edited)
Thanks Pete, It is an interesting discussion for sure. Jim mentioned having never thought about these concepts, and I don't disagree. I spent a few years training without a complete understanding myself. There's a whole lot more to consider as you cycle through a large number of years and a large number of dogs, for sure.

I decided to learn as I tried to become more proficient at problem solving in a variety of breeds. Supporting that was the need to adapt to various limitations placed on my training by a customer who may or may not approve of, or be physically able to execute a certain method. Having an understanding allows me to better educate my customers, which increases their ability to maintain training. It also helps me when I run into some odd problems we wouldn't necessarily have with a retriever we trained from a puppy.

Fact is, I'm just a babe in the woods in terms of experience, but I am trying hard to accelerate my learning curve through solid study of the basic concepts and handling as many dogs as possible. After a couple of hundred dogs, I feel like I am starting to know a little bit, but only a little bit about training. There are still so many things to learn, and I think there always will be.

This discussion is more about someone's ability to talk the talk than walk the walk, IMHO.
 
#97 ·
Do you give a cold burn or nick with out a whistle sit ...?
No. Dog is whistle sat first...then burn or nick at a higher level than a regular corrrection (what level depends on the dog). The usual protocal in training is attrition first and then indirect pressure (again depends on the dog).

For me, this is more of a "finesse" thing than a "hammer" (again depends on the dog). It is not a "cookie cutter" process.

It should be noted teaching the correct response to indirect pressure does not start with an e-collar.
 
#108 ·
There's something else that I feel is an important distinction between CC and OC.
The dog's attitude about training.

A dog's attitude, is not something that it has the ability to "decide" to have. It's beyond it's control to change it.
That falls under CC, but it can be changed by us through applying OC.

And we most certainly do effect a dog's attitude about training, with how we train.

The problem is, that I don't think that very many people really understand what it is, that conditions and maintains a good attitude about training.

I will say this.
It has a whole lot more to do with the dog's understanding of what everything means, and what predicts what, than how many cookies it gets during the average training session.

A constantly confused dog, is a sad dog.
No matter how much you try to make it happy.
 
#110 · (Edited)
Many years ago I wrote my masters thesis. The title was “Memorize and Perish”. The basic philosophy behind this expression is when memorization becomes the foundation for learning, the skill of thinking is negatively impacted. The sheer magnitude of memorization (this is the way things are) inhibits brainstorming. Critical thinking is stifled.

Another important aspect of critical thinking is taking advantage of how the human mind/brain functions. The ability to think and reason is often impacted by how much stimulation is taking place. The timing of different brain waves is critical. In laymen terms, thinking functions are more effective when you have not been over stimulated. The best time to “get across an idea” (explain something and/or teach) is in the first five minutes of the effort. The subject mind/brain is more receptive. On a side note, my dogs are often taught “new things” near the beginning of a training day.

However, one must always remember that dogs do not think like humans (but that is another topic....or is it?). ;)

Now you might wonder where this is going. Therefore, to clarify. I slept on the thoughts of this thread. My thinking was going nowhere last night. As is my routine (which was discovered many years ago) I gradually woke up this morning with an uncluttered, clear mind. The manner in which to contrast and compare the reward/punishment conflict created by this thread was quickly clarified (for me).

See if this is “reasonable”. Take a simple situation where indirect pressure works - the mouthing dog at the handler's side. The dog has heeled properly and sits without any verbal command. The dog's OB in this respect is “solid”. The word “solid” means the same thing as “excellent conditioned responses”. However, the dog begins to mouth/roll the bumper and the trainer does “sit/stick/sit” and he stops. This is classic and simple indirect pressure.

Most can describe the situation with anecdotal observations – the dog is “glassy eyed”, apparently unaware of his behavior, does it more when excited or “amped” and the ever popular doesn't know I'm even there. Some say it is a subconscious behavior. The general consensus is indirect pressure works (often). It seems that the difficult part (in this thread) is explaining why.

Apparently (and I use this word because I presently don't understand the explanations), the “stick” enforces sit and punishes mouthing. Here is my first question from this morning's moment of clarification. Are punishing, extinguishing, distracting, re-focusing and/or becoming more responsive all the same thing?

My second question is, since mouthing is more of a subconscious behavior (conjecture on my part), the dog does not actually think “I need to stop mouthing” and stops doing it because with sufficient reps the behavior is sub-consciously extinguished and/or punished enough?

In comparison, my training has been greatly influenced by the term balance. The dog that mouths is out of balance. How so? because he is not responsive/focused.....enough. I can see it. The dog is not paying attention to me and locked into his own somewhat compulsive behaviors. This is probably because I have given him too much to handle and poor coping skills allow him to manage this induced anxiety.

Rather than thinking that indirect pressure has punished the behavior, my “sit/stick/sit” application refocuses the dog in the area of responsiveness. If he learns to and/or is conditioned to pay more attention to me (enhanced responsiveness), his need to develop poor coping skills for his apparent anxiety is reduced. In essence, I don't think in terms of Pablov and/or Skinner terminology. It is more of “what you see is what you get” kind of thing as opposed to “what you see is not what you get”.

Which brings me to the present linguistic impedance. Does punishing “mouthing” indirectly while enforcing sit directly mean the same thing as “Good! You are sitting, but clear your mind and pay more attention to me?” Am I enforcing two things “sit” and “pay attention” and not really punishing anything?

OK, now I've reached the point of “cluttered thinking” which is the cue to stop (brain waves need resetting plus my dogs need to “air”).
 
#112 ·
Are punishing, extinguishing, distracting, re-focusing and/or becoming more responsive all the same thing?
The words mean to you, what you make them mean.

But what they mean to you, doesn't help me understand what you are trying to say.
You can make it work for you, but you can't communicate to me how you make it work, because I have no idea what you mean.


If nothing else, studying theories like OC and CC, allows us to standardize at least some of the definitions, and effectively communicate with others that are also familiar with those standardized definitions.


Now, if you are actually thinking about fire extinguishers, plus peas, and minus arghs while you are working a dog, you are going to be way behind the ball, with every swing you take!

But, when you have time to sit down and really think this stuff through, it helps you be able to form training plans that make sense, and are effective.
 
#111 ·
Here is a short introduction to the E-Collar by Michael Ellis. In it he discusses how the collar lends itself to superstitious associations if not properly introduce. I though it was worth the watch and pertained to this discussion some what.

http://michaelellisschool.com/vid_properuse.htm
 
#117 ·
Here is a short introduction to the E-Collar by Michael Ellis. In it he discusses how the collar lends itself to superstitious associations if not properly introduce. I though it was worth the watch and pertained to this discussion some what.

http://michaelellisschool.com/vid_properuse.htm
I don't care about Pavlov vs Skinner, I don't know exactly what classical conditioning and operant conditioning are.

But everybody who trains with, is thinking about training with, or has ever heard of or has an opinion on ecollar training should watch the above video. That guy is a good communicator.
 
#113 ·
copterdoc said:
The words mean to you, what you make them mean.

But what they mean to you, doesn't help me understand what you are trying to say.
You can make it work for you, but you can't communicate to me how you make it work, because I have no idea what you mean.

If nothing else, studying theories like OC and CC, allows us to standardize at least some of the definitions, and effectively communicate with others that are also familiar with those standardized definitions.


Now, if you are actually thinking about fire extinguishers, plus peas, and minus arghs while you are working a dog, you are going to be way behind the ball, with every swing you take!

But, when you have time to sit down and really think this stuff through, it helps you be able to form training plans that make sense, and are effective.
That's kind of the response I expected. Thanks.
 
#114 ·
Here is another Michael Ellis video that is part of a lecture on OC and also talks about how an e-collar can play into all of this. It also helped me understand some of the things that Copterdoc has stated. If you are interested it is worth a watch.

http://michaelellisschool.com/vid_opperantcondit.htm
 
#118 ·
Howard said:
I don't care about Pavlov vs Skinner, I don't know exactly what classical conditioning and operant conditioning are.

But everybody who trains with, is thinking about training with, or has ever heard of or has an opinion on ecollar training should watch the above video. That guy is a good communicator.
I totally agree! :)
 
#131 ·
If you consider the dog's work as a series of smaller behaviors where one behavior leads to another, you have a behavior chain. Completing one behavior leads to the next behavior and so on until the task is completed. When the dog does something wrong it has departed from the correct behavior chain and has started a different set of behaviors. In theory, any correction for a preceding behavior will be indirect pressure for a behavior following it in the chain.

I am surprised that it works but it does.

Luckily a trainer doesn't have to understand the theory but just has to know correct application.
 
#132 ·
If you consider the dog's work as a series of smaller behaviors where one behavior leads to another, you have a behavior chain. Completing one behavior leads to the next behavior and so on until the task is completed. When the dog does something wrong it has departed from the correct behavior chain and has started a different set of behaviors. In theory, any correction for a preceding behavior will be indirect pressure for a behavior following it in the chain.
I am surprised that it works but it does.

Luckily a trainer doesn't have to understand the theory but just has to know correct application.
This is the way I understand indirect pressure too..the direct pressure effects the next command indirectly ....Not the preceding action being punished..as in the example give before...Steve S

John , we have all been wanting to know the same thing since the begining of this threead....Steve S
 
#134 ·
The correction tells the dog "You have started on the wrong behavior chain. Let me cast again and so that you can do the right behavior chain." I think that understanding behavior chains is important to understanding why this works. Maybe a good model would be to think of the behavior chain like a chain reaction. You are telling the dog that he is doing the wrong chain reaction.

These are not my own thoughts. I credit Lorie Jolly and Marilyn Fender.
 
#135 ·
For a cast refusal, why wouldn't Steve's "NO", sit whistle , cast... or a "cold burn", sit whistle, cast work at least as well as IP.

Better for me in fact, since I can understand why I got the compliance.

john
 
#149 ·
I've spent a good bit of time trying to read each and every post in this thread right up to #144 which currently precedes my reply.

Frankly, this stuff really muddies the water for me.

Copterdoc. If all of this talk and theory and syntax helps you and your training, that's great. What matters is what works for you and your dogs.

I personally un-benefit from all of this stuff.

Here's my analogy. I could study germination, GMO seed development, photosynthesis and meteorology to plant a killer crop of tomatoes. Or I could watch for the right temperature and the calendar, buy some good seeds, plant them and water the garden if it doesn't get enough rain. In the end, I'll have some nice tomoatoes if I do it right.

All this theory and definitiion and discussion may benefit some of you. If it does, that's awesome. For me and my dogs at this point in time, it is overload and overcomplication.

Chris
 
#159 ·
I've spent a good bit of time trying to read each and every post in this thread right up to #144 which currently precedes my reply.

Frankly, this stuff really muddies the water for me.

Copterdoc. If all of this talk and theory and syntax helps you and your training, that's great. What matters is what works for you and your dogs.

I personally un-benefit from all of this stuff.

Here's my analogy. I could study germination, GMO seed development, photosynthesis and meteorology to plant a killer crop of tomatoes. Or I could watch for the right temperature and the calendar, buy some good seeds, plant them and water the garden if it doesn't get enough rain. In the end, I'll have some nice tomoatoes if I do it right.

All this theory and definitiion and discussion may benefit some of you. If it does, that's awesome. For me and my dogs at this point in time, it is overload and overcomplication.

Chris
Chris some of us want to know why it works ..INQUIRING MINDS...I agree it is hard to get a complete understanding of all the stuff ( gdg) but the more one understands the why they can appreciate the out come.. ..I'm not saying I have the complete understanding but struggling to get there...Steve S
 
#155 ·
"Whistle sits re-enforced with pressure always has a moment where we +p running and -r sitting. "


As in the case of a dog already in a sit position or in the process of sitting and pressure is applied ( as talked about in the use of indirect pressure to get the next command ) where is the +p ? If you only stimulate ( to reinforce ) while the dog is doing a command correctly where is the +p ? Heel ,dog in motion pressure ...back dog in motion ...Improper use of the collar scares me a lot...With out proper training in its use it can and will cause more problems than can be fixed ....Steve S
 
#157 · (Edited)
Steve I don't ever nick or burn a dog whose ass is already on the ground. That serves no purpose I know of.
Nick/burn on sit is either avoid a nick or escape a burn. The dog is out of position then the pressure is applied, even in the case of indirect pressure. The way it's timed the last behavior the dog did just before the pressure is moving, so that is punished and the next behavior (sitting) in re-enforced.
 
#176 · (Edited)
Here's a test you can try.

See if you can get the dog to go back through the same spot you burned it from.
I guess it depends on how much pressure you use. More than once I have gotten the same CR. I am trying to communicate no, let's try again, with just enough pressure to get a change in behavior. In the case of a CR on a long angle entry,the dog is choosing to not get in when it has been trained to and has done it correctly many times. It's cold or windy, or there is a PB lying up on the bank. I don't believe I have created a hot spot on the bank by correcting the dog for a CR and not getting in. I have corrected the dog for non-compliance ( lack of effort) to a known command. Taking the correct cast and allowing the dog to carry it for sometime reinforces the desired behavior.

I could try to anticipate the instant the dog makes the decision to not get in. It could be on the line, or in route. Rather than give the dog a freebee cast, I could stop, correct while the sit is in progress, then cast. Or, call back with or without pressure and resend.

I will add that in the past I have created hot spots the dog learned to avoid. That does not always work so well when you try to generalize the behavior to new locations. I ended up with a dog out of balance to the watery side.

I try to give the dog the benefit of the doubt and err on the side of no pressure if I am not sure. Learning as I go...
 
#177 ·
Testing Dogs and 'Trialling' Dogs are two different things (in my world)!...Maybe that IS! the difference between Pavlov and Skinner?...and a few more besides?
I'm off' you guy's now,and leave you to your 'games'!...Shame!...
..............
 
#179 · (Edited)
@ Steve B. Hi! All I am referring to there in terms of "new methods" is that I'm now of the mindset to avoid collar pressure in all but the most clear cut situations. Go/stop/come, for instance. As you mentioned, sometimes pressure can be interpreted incorrectly, thus, I would rather avoid it unless I'm pretty sure the dog is going to understand it 100%.
 
#180 ·
@ Steve B. Hi! All I am referring to there in terms of "new methods" is that I'm now of the mindset to avoid collar pressure in all but the most clear cut situations. Go/stop/come, for instance. As you mentioned, sometimes pressure can be interpreted incorrectly, thus, I would rather avoid it unless I'm pretty sure the dog is going to understand it 100%.


Very good ..Welcome aboard....Your " new method " is what I have been doing since the 70's...Steve S ...
 
#183 ·
But will not learn from the info passed back and forth ...That is what an open forum is for ...open debate and discussion...All we need to do is keep it civil and respect others views even if they are different from ours...After all there is more than one way to train a dog.....Steve S
 
#184 ·
Pavlov is always on your shoulder. Classical conditioning is always in play. Classical conditioning is why your dog gets excited when you get out the guns and bumpers and e-collars.

In Bob Bailey's Fundamentals of Animal Training DVD he explains gives an example of how to tell if a behavior is operantly conditioned or classically conditioned.

"I want to go to the movies. There is no place you can hit me to make me want to go to the movies."

(Well he phrases it better than that I don't have the disc on hand to check)

Classical conditioning is always in play. But I'd say 90% of retriever training is operant conditioning, particularly because in so many cases we're fighting against the dog's natural inclination (prime example: cheating water/cover).

Most dog training is primarily operant. Only with aggression/fear/behavior problems of that ilk do you work primarily with classical conditioning, because you're working to change the dog's internal state.
 
#185 ·
....I'd say 90% of retriever training is operant conditioning, particularly because in so many cases we're fighting against the dog's natural inclination (prime example: cheating water/cover).
I agree. But, the foundation of the training for a handling Retriever, relies on Classically conditioned responses.

We can't get to the 90%, until we have established those responses.
 
#188 · (Edited)
The stim never means sit... unless it comes directly after a sit command...

It never means anything really. The commands we give, be they verbal or via body language are meaningful. The stim is simply there to be either escaped or avoided with compliance.

It is a re-enforcer and a punisher, but never, does it ever come to be a cue to an particular action.

At least not the way we train.

It most certainly can be used that way. If I only ever wanted one command I could turn the stim on, lure the dog into a sit position and turn it off... If I did this enough times without ever saying anything there would coma a time when I hit the button and the dog did the only thing he knew, in this case sit. We don't do it that way (hopefully). I suppose you could probably equate it to a number of commands just like we do treats, and occasionally get a dog that cycles through all known behaviors trying to turn the collar off. I think this is neat when it's to earn a treat, but I would never do it with a collar.

Because we use the collar to punish and or re-enforce a whole variety of behaviors, it never really signals anything, except that the dog needs to expediently do what he was commanded.
 
#189 ·
......The stim is simply there to be either escaped or avoided with compliance........
That's how we use the collar to shape future behavior.

Collar conditioning, is not about shaping behavior.
It's about conditioning responses, to the specific aversive used.

The dog is already performing the behaviors, in response to the commands, prior to starting collar conditioning.
But, it won't respond correctly to the pressure applied by the collar, until it is classically conditioned to.
 
#190 ·
Dogs can be conditioned to respond a certain way in a certain context with the push of a button without a command or cue. When a dog is made responsible for a specific behavior and that behavior fails on cue with the context,,, a well conditioned dog will do what is needed to be done ,,,silently,,,with only stimulous to cause him to do the correct behavior
Pete
 
#191 ·
That's true.

When a dog knows sit-to-flush, and a bird gets up, it can be held to a high enough standard that sit is enforceable with the collar, without the command being given.
 
#192 ·
Dogs can be conditioned to respond a certain way in a certain context with the push of a button without a command or cue. When a dog is made responsible for a specific behavior and that behavior fails on cue with the context,,, a well conditioned dog will do what is needed to be done ,,,silently,,,with only stimulous to cause him to do the correct behavior
Pete

When a dog knows sit-to-flush, and a bird gets up, it can be held to a high enough standard that sit is enforceable with the collar, without the command being given. copterdoc..

If the dog associated the bird flush with the stimulation in a positive manner it works ...but,if in a negative manner you get a dog that is bird shy now...Just as a dog can be conditioned to sit on a call or shot...the correct association is a must to have the proper out come....Steve S
 
#193 ·
If the dog associated the bird flush with the stimulation in a positive manner it works ...but,if in a negative manner you get a dog that is bird shy now...Just as a dog can be conditioned to sit on a call or shot...the correct association is a must to have the proper out come....Steve S
It's called chaining.

If the dog knows that "B" predicts "A", and is conditioned to perform an action in response to "B" in preparation for "A", we can then condition another link in the chain.

So, we can create a new "C", that predicts "B", and have the dog perform the same behavior that it was conditioned to perform in preparation for "A".

That's all Classical Conditioning.

And it's how we condition a complete retrieve, from a series of conditioned behaviors.
We condition them all separately, and then each one "commands" the next in sequence.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top