Puppy Forum and Dog Forums banner

1 - 14 of 14 Posts

·
Registered
Joined
·
5,406 Posts
gotta say, she doesn't really address the issue of where the gentle leader falls. I can understand that, as she uses the head halter frequently. I've found that a lot of dogs seem to consider it aversive.
 

·
Registered
Joined
·
7,262 Posts
Discussion Starter #3
gotta say, she doesn't really address the issue of where the gentle leader falls. I can understand that, as she uses the head halter frequently. I've found that a lot of dogs seem to consider it aversive.
She does in the sense that she explains there are several ways to use it and how it's defined depends how it's used (but, yeah, it does seem as though she's dodging the question).

I was more interested in her explanation of the terminology. It's one of the most accurate and understandable I've seen on the web. Thought it might be interesting to some. One of my pet peeves is using "negative reinforcement" to mean "positive punishment."
 

·
Registered
Joined
·
5,406 Posts
She does in the sense that she explains there are several ways to use it and how it's defined depends how it's used (but, yeah, it does seem as though she's dodging the question).
"
Actually how it's defined depends on the learner's perception. If the dog perceives something as aversive, it doesn't matter that you think it should be rewarding. For instance, I've seen people use rough play with sensitive dogs. The human obviously thinks the play is rewarding. Sometimes the dogs are obviously not loving it.
 

·
Registered
Joined
·
7,262 Posts
Discussion Starter #5
Actually how it's defined depends on the learner's perception. If the dog perceives something as aversive, it doesn't matter that you think it should be rewarding. For instance, I've seen people use rough play with sensitive dogs. The human obviously thinks the play is rewarding. Sometimes the dogs are obviously not loving it.
Certainly whether something is pleasant or aversive is subjective. Yin seems to begin with the assumption that the pressure from the gentle leader is aversive and that may not be true in all cases. Given that assumption, however, she then further classifies it as positive punishment or negative reinforcement depending on how it's actually used: is the pressure applied following an undesirable behavior with the intent to decrease it (positive punishment) or is the pressure felt until a desirable behavior is exhibited with the intent to increase it (negative reinforcement).

With your example of rough play, once we know a dog enjoys it, we can use it as positive reinforcement (dog greets us nicely, we wrestle) or negative punishment (dog gets nippy, play stops). If a dog doesn't enjoy it, we could use in the opposite way, but that would be cruel.

Really, though, I wasn't as interested in her classification of the gentle leader as I was in her clear definitions of the basic terminology and examples. Most examples I've read use rats bar pressing for chow or pigeons pecking for seeds; it's nice to read dog training applications.
 

·
Registered
Joined
·
3,722 Posts
She does in the sense that she explains there are several ways to use it and how it's defined depends how it's used (but, yeah, it does seem as though she's dodging the question).

I was more interested in her explanation of the terminology. It's one of the most accurate and understandable I've seen on the web. Thought it might be interesting to some. One of my pet peeves is using "negative reinforcement" to mean "positive punishment."
They go hand in hand though. To use R- you have to apply an aversive in the first place, which is P+. So technically, it's very difficult to use R- without having used P+ as well.
 

·
Registered
Joined
·
5,406 Posts
Really, though, I wasn't as interested in her classification of the gentle leader as I was in her clear definitions of the basic terminology and examples. Most examples I've read use rats bar pressing for chow or pigeons pecking for seeds; it's nice to read dog training applications.
I'm very interested in how Yin defines the GL/Head Halter. She claims to be a positive reinforcement based trainer. And I think to a large degree she is. But this video disturbs me as much as watching someone do the same thing with a prong collar. Does it work? Yes. Are there other, gentler handling methods that work as well? Yes. Could she have been as successful without all that jerking around? I think so. http://www.youtube.com/watch?v=EUCl6ndLN7Q
 

·
Registered
Joined
·
7,262 Posts
Discussion Starter #8
They go hand in hand though. To use R- you have to apply an aversive in the first place, which is P+. So technically, it's very difficult to use R- without having used P+ as well.
So, negative reinforcement and positive punishment are the same thing? Dog jumps on me, I knee him in the chest - I can call that negative reinforcement?
 

·
Registered
Joined
·
206 Posts
So, negative reinforcement and positive punishment are the same thing? Dog jumps on me, I knee him in the chest - I can call that negative reinforcement?
No, you can look at it as an equation. In positive reinforcement you add something good to an equation to reinforce a behavior (ie a treat). In negative reinforcement you remove something bad in order to reinforce a behavior (ie leash pressure). In positive punishment you add something bad (ie hit the dog). In negative punishment you remove something good (ie withold a treat).

Therefore, positive reinforcement and negative punishment go hand in hand so to say...
 

·
Registered
Joined
·
7,262 Posts
Discussion Starter #10
No, you can look at it as an equation. In positive reinforcement you add something good to an equation to reinforce a behavior (ie a treat). In negative reinforcement you remove something bad in order to reinforce a behavior (ie leash pressure). In positive punishment you add something bad (ie hit the dog). In negative punishment you remove something good (ie withold a treat).

Therefore, positive reinforcement and negative punishment go hand in hand so to say...
Yes, it's all related and nuanced. My point to lil_fuzzy is that most people I've encountered say "negative reinforcement" when they mean "positive punishment" (e.g., they'll refer to a knee to the chest for jumping or scolding for crossing a boundary as negative reinforcement). Perhaps I've been in the company of the wrong people if no one else has ever encountered this type of confusion.

In operant conditioning, the stimulus follows the response: R --> S. So, in positive punishment, the aversive stimulus is presented after the undesirable response (e.g., knee in the chest for jumping, verbal scolding for crossing a boundary). In negative reinforcement, the aversive stimulus is removed after the desired response (e.g., leash pressure or electric shock removed for moving towards handler). In negative reinforcement, the aversive stimulus is not presented as a result of the dog's actions.

As dogs learn, a positive punishment can become negative reinforcement. The first time I knee my dog in the chest for jumping, it's positive punishment; if on subsequent greetings, the dog no longer jumps because he's avoiding my knee, it's negative reinforcement. Of course, going back to Pawzk9's point, we're assuming here that the dog finds the knee in his chest, verbal scolding, etc. aversive.

There's also some difference in how training is defined. Do I want to stop a behavior? Then I use punishment. Dog jumps when I come home, I knee him in the chest (positive punishment). Do I want to encourage a behavior? Then I use reinforcement. Dog sits nicely when I come home, I give him attention (positive reinforcement).

Really, my original intent was to share a bit of information that others may find useful. I'll refrain from doing so in the future.
 

·
Registered
Joined
·
5,406 Posts
So, negative reinforcement and positive punishment are the same thing? Dog jumps on me, I knee him in the chest - I can call that negative reinforcement?
positive punishment and negative reinforcement go hand in hand - just as positive reinforcement and negative punishment go hand in hand. For instance, I have my dog a prong collar (which would not happen in real life - it's a hypothetical). When the dog pulls into the collar, the prongs poke him. When the dog ceases to pull, the pokes goe away. So, are you punishing pulling with positive punishment or are you rewarding LLW with negative reinforcement?
 

·
Registered
Joined
·
5,406 Posts
Yes, it's all related and nuanced. My point to lil_fuzzy is that most people I've encountered say "negative reinforcement" when they mean "positive punishment" (e.g., they'll refer to a knee to the chest for jumping or scolding for crossing a boundary as negative reinforcement). Perhaps I've been in the company of the wrong people if no one else has ever encountered this type of confusion.
<snip>Really, my original intent was to share a bit of information that others may find useful. I'll refrain from doing so in the future.
Please don't (refrain) in the future. I think it's been an interesting discussion. Your focus was more on the information given. I would have like to see Yin really talk about the GL (which was supposedly the topic) because I'm not sure (given video) that what she does with it meshes well with the philosophy she claims. While I"ve seen some really great information from her, that bothers me a bit. I think generally the people who really muddle the terms are people who don't really understand them, though they can get sticky even for those who do.
 

·
Registered
Joined
·
7,262 Posts
Discussion Starter #13
positive punishment and negative reinforcement go hand in hand - just as positive reinforcement and negative punishment go hand in hand. For instance, I have my dog a prong collar (which would not happen in real life - it's a hypothetical). When the dog pulls into the collar, the prongs poke him. When the dog ceases to pull, the pokes goe away. So, are you punishing pulling with positive punishment or are you rewarding LLW with negative reinforcement?
Prong collars, head harnesses, and choke chains absolutely work on both principles and there's an invisible line between P+ & R- in its use. In general, though, whether something is classified as P+ or R- depends on the sequence of events: the stimulus always follows the response. If you start your walk with the leash and collar loose and it tightens / pokes when the dog pulls, it's P+. If you begin your walk with pressure on the leash so that it pokes the dog's neck and then release the pressure when the dog is in a proper heel, it's R-. And yes, as you continue the walk you alternate between P+ and R- partly because the behavior is continuous. Again, we're assuming that the pressure from collar / harness is aversive, simply wearing the collar / harness isn't aversive, and that the dog hasn't been classically conditioned to have a negative reaction to the collar / harness.

Similarly, if you're trying to teach your dog to stay within a specific area, you can combine P+ (dog crosses border, dog gets scolded) with R- (continue to scold until the dog re-enters the allowed area). Eventually the dog learns he can completely avoid scolding by staying within a specified area (more R-). The key is that the reinforcement or punishment follows the behavior. You can't negatively reinforce a behavior by presenting an aversive stimulus following a behavior - and this is the way most people I've heard misuse the term "negative reinforcement."

You're right that it can be confusing and there are situations where the classification of the stimulus is difficult, if not impossible. Is my paycheck R+ (yay! more money for dog toys) or R- (yay! escape from poverty)? And yes, I realize this depends on the specific state of my bank account.

Please don't (refrain) in the future. I think it's been an interesting discussion. Your focus was more on the information given. I would have like to see Yin really talk about the GL (which was supposedly the topic) because I'm not sure (given video) that what she does with it meshes well with the philosophy she claims. While I"ve seen some really great information from her, that bothers me a bit. I think generally the people who really muddle the terms are people who don't really understand them, though they can get sticky even for those who do.
Sorry for being snippy; I was feeling attacked for everything I posted. Conditioning and learning are fascinating topics and it's so much more complex than most of the dog training books I've read. Skinner is one of my favorite psychologist, mostly because of his invention of the air crib. I don't really know enough about Yin to have an opinion of her. I have her book, How to Behave So Your Dog Behaves, but lost interest when it was repetitive of the other books I've read.

To be completely honest, I didn't see the video of her using the gentle leader, just videos of other LLW techniques. The GL scares me a bit. It seems as though a rambunctious dog could sustain severe injuries from it.
 

·
Registered
Joined
·
5,406 Posts
Again, we're assuming that the pressure from collar / harness is aversive, simply wearing the collar / harness isn't aversive, and that the dog hasn't been classically conditioned to have a negative reaction to the collar / harness.

Sorry for being snippy; I was feeling attacked for everything I posted. Conditioning and learning are fascinating topics and it's so much more complex than most of the dog training books I've read. Skinner is one of my favorite psychologist, mostly because of his invention of the air crib. I don't really know enough about Yin to have an opinion of her. I have her book, How to Behave So Your Dog Behaves, but lost interest when it was repetitive of the other books I've read.

To be completely honest, I didn't see the video of her using the gentle leader, just videos of other LLW techniques. The GL scares me a bit. It seems as though a rambunctious dog could sustain severe injuries from it.
I think for a number of dogs the head halter is naturally aversive - I don't know many dogs who find a prong (or even ecollar) aversive just to wear. I'ts when people use it that things get dicey. On the other hand the majority of dogs when you put the head halter on, do their level best to get that thing off their face, and have to be conditioned to accept even wearing it. Plus the pressure on the top of the nose has a "calming - read depressing effect on many dogs, which is why some behaviorists like it so much

I'm sorry you felt attacked. It certainly wasn't my intent - just to present another opinion, which is what discussion is about, right? Sophia Yin has put out some very good stuff on how to train dogs without force or violence, and it just saddened me a bit to see that video of her fighting with a dog like that. Especially when there are very elegant, non forceful methods to work with a reactive dog - Grisha Stewart's BAT (which technically involves R- because the reward is release from social pressure) and my favorite - Leslie McDevitt's Look At That game. I agree with you that a dog could do damage to itself with a head halter
 
1 - 14 of 14 Posts
Top