The RetrieverTraining.Net Forums The Retriever Academy
Total Retriever Training with Mike Lardy
Hawkeye Media Gunners Up Tritronics Outdoor Media
Page 4 of 5 FirstFirst ... 2345 LastLast
Results 31 to 40 of 41

Thread: "Aversives": A little something for the newer trainer.

  1. #31
    Senior Member rmilner's Avatar
    Join Date
    Dec 2005
    Posts
    343

    Default

    Quote Originally Posted by DarrinGreene View Post
    If we want to talk about learning theory then let's talk about it in full.
    I am not a positive extremist, but I am a heck of a lot more positive than I used to be. One of my favorite trainers and a significant mentor to me was a horse trainer named Ray Hunt. Ray was a superb positive trainer of horses. Ray used to say frequently: "I'm here for the horse - to help him get a better deal".

    I try to follow that philosophy.

    A fairly typical method in operant conditioning is to simply observe the animal and when you see a behavior you want, you mark it with a sound. Then you reward it as quickly as you can. The term used is "capturing" a behavior."

    When the trainer "captures" a behavior with a marker and reinforces that behavior with reward, where is the aversive?
    Last edited by rmilner; 09-09-2013 at 01:50 PM.
    Robert Milner
    www.DuckhillKennels.com


    "When he stood up to speak, battalions of words issued forth from his mouth and scoured the countryside in search of an idea, and when they found one, they swiftly and thoroughly beat it to death." ---- -Anonymous

  2. #32
    Senior Member RetrieversONLINE's Avatar
    Join Date
    Nov 2005
    Location
    Lindsay, Ontario
    Posts
    654

    Default

    Quote Originally Posted by rmilner View Post
    I am not a positive extremist, but I am a heck of a lot more positive than I used to be. One of my favorite trainers and a significant mentor to me was a horse trainer named Ray Hunt. Ray was a superb positive trainer of horses. Ray used to say frequently: "I'm here for the horse - to help him get a better deal".

    I try to follow that philosophy.
    Robert

    I am also a huge fan of the late Ray Hunt. I have rode my horse with him several times as well as studied everything he has ever produced. Add to that his mentor, Tom Dorrance and their many disciples like Buck Brannaman, Tom Curtin, Martin Black, Bryan Neubert, Joe Wolters, Peter Campbell, Pat Parelli, etc etc and one can gain a tremendous amount of insight into training. I quote Ray and the others often in my retriever articles because I find the principles universal.

    Ray was always trying to "feel" for the horse. His methods often involved making the right things easy and the wrong things hard for the horse. I wouldn't have labeled him as a just positive trainer. But, he sure reduced the type and amount of pressure and force that was being used conventionally. He has probably had a stronger influence on changing horsemanship methods than any other single individual.
    Dennis

  3. #33
    Senior Member Jerry Beil's Avatar
    Join Date
    Feb 2011
    Location
    North Carolina
    Posts
    389

    Default

    Quote Originally Posted by DarrinGreene View Post
    It's a very worthy discussion Evan, especially when it isn't accurately described in the very first post.

    "Forcing" is far from the exception to the rule in terms of negative re-enforcement strategies used in retriever training. We use both punishment and negative re-enforcement every single time we train. Every single time. Even if we're using steak!

    Here's what I mean, starting with aversives. When a dog is standing and we command sit, followed by pinch collar pressure, we are punishing the previous behavior (standing) and re-enforcing the new behavior (sitting). If we command here, followed with the e-collar, we are punishing whatever else he was doing (sniffing dog poo) and re-enforcing here. On heel with a pinch collar or a stick we are punishing him being out of position and re-enforcing him being in position.

    This is actually true in positive re-enforcement training. If a dog is standing and we command sit, we withhold the reward, creating frustration (a form of punishment), until he sits, relieving the frustration (negative re-enforcement) and earning the reward (positive re-enforcement).

    I could go through all of the other examples as well but it isn't necessary. Hopefully people get the idea, supporting my assertion that we are ALWAYS discouraging a behavior and encouraging a behavior. Regardless of the re-enforcement strategy, for behavior to change we have to discourage the previous behavior and encourage the new (desired) one. Logic says so.

    The only question then becomes, how strong the aversive treatment happens to be. Even standing there withholding a treat the dog desperately wants is very frustrating and stressful for the dog. It is, in fact punishment. We're teasing him and no one likes to be teased. I have asked a +r trainer, "Did you like it when your brother ran around with your favorite toy saying "nanny nanny poo poo" when you wanted to play with it?" No, you didn't and you probably chased him, then ran to Mommy to try and get the toy. That may or may not have worked but eventually, you found a way to get that toy back, didn't you? Maybe you ignored him and he got bored with your lack of reaction. In the future then, you immediately ignored him and got your toy back more quickly. He's trained you to ignore him when he picks up a toy of yours now, hasn't he? He give a signal (picking up the toy) and you respond appropriately (ignoring him). This example is no different than a dog running through various behaviors trying to get a reward. You're teasing them, they're frustrated and trying to figure our how to relieve that frustration. This is an aversive experience for the dog, whether these trainers want to admit it or not.

    I have seen "positive only" trainers stand on a leash many times to keep a dog from moving away. They will tell you this isn't punishment, to which I have replied more than once, "OK then put the collar on and I'll tie you to a post". Same concept as the above, only a bit more aversive to the dog.

    We go from there all the way to using a heeling stick! It's all a matter of semantics but there is ALWAYS an aversive of some form used in training, period, end of story. The only question is what level of aversive treatment you're willing to inflict on the dog to get what you want.

    Bottom line, training is training. In order to "change" behavior one has to be discouraged and the other encouraged. You ALWAYS use an aversive to effect that change and some form of reward to re-enforce desirable behavior. It's just a matter of degrees.

    Defining aversives in terms of what we believe to be socially acceptable causes a lot of misconceptions among the general public and I believe is a terrible idea.

    If we want to talk about learning theory then let's talk about it in full.
    Very interesting... Is the practical application of that knowledge that when you are apply the aversive say with here, you are always reinforcing here, but you are not always punishing the same thing. Only the dog knows what it connects the correction to. So, that's why you want to set up specific scenarios for the dog when conditioning to the aversive. It's not automatic for the dog to realize that NOT sitting when given the SIT command is what is being punished, even though he consistently relieves the pressure and thus rewards the sit. Over time, in a controlled environment, he learns that it isn't anything specific that he's doing wrong other than doing ANYTHING besides sitting when told to sit. I don't think I said that very well.

    The problems crop up when the dog isn't 100% clear on that connection, and gets the aversive while he's performing a desired behavior and told to sit. He connects the aversive to the desired behavior in addition to the reward of the sit, and as a trainer you have a losing situation. The dog gets rewarded for stopping the desired behavior and sitting, and if the dog isn't 100% clear on SIT being the target behavior already, there's nothing much you can do about it, and he's less likely to do the desired behavior next time.

    Is that right?
    <><><><><><><><><><><><><><><><><><><>
    Donway's Dixieland Delight - Dixie 2/24/1997 - 2001
    Rebel's Ruffian Hank - Hank - 6/05/2001 - 2/3/2011 - Profile Picture
    Blue Ridge Pot O Gold - Séamus - 1/22/2011 -
    Old North State Queen Anne's Revenge- Annie- 3/21/2013-

  4. #34
    Senior Member Evan's Avatar
    Join Date
    Jan 2003
    Location
    Kansas City, MO
    Posts
    5,140

    Default

    You are not always punishing when using an aversive. Just in case you are getting lost, here are the names of some common aversives used in dog training.

    Leash & choker or prong collar
    Heeling stick
    Rope/check cord
    E-collar

    What makes these tools aversive? They each deliver an unpleasant stimulus, each in a different way, and that stimulus is directed toward a behavior change. Neither the amount, nor the kind of stimulus are what make them aversive. It is that the stimulus they deliver is perceived by the dog as unpleasant, and results in a behavior change. That's it. Don't drown yourself in OC terms.

    Evan
    "Prepare your dog in such a manner that the work he is normally called upon to do under-whelms him, not overwhelms him." ~ Evan Graham

    “People who say it cannot be done should not interrupt those who are doing it.”

    ― George Bernard Shaw


    The Smartwork System for Retriever Training (link)
    http://www.facebook.com/profile.php?...59&ref=profile

  5. #35
    Senior Member DarrinGreene's Avatar
    Join Date
    Feb 2007
    Location
    Souf Joisy
    Posts
    2,751

    Default

    Quote Originally Posted by rmilner View Post
    I am not a positive extremist, but I am a heck of a lot more positive than I used to be. One of my favorite trainers and a significant mentor to me was a horse trainer named Ray Hunt. Ray was a superb positive trainer of horses. Ray used to say frequently: "I'm here for the horse - to help him get a better deal".

    I try to follow that philosophy.

    A fairly typical method in operant conditioning is to simply observe the animal and when you see a behavior you want, you mark it with a sound. Then you reward it as quickly as you can. The term used is "capturing" a behavior."

    When the trainer "captures" a behavior with a marker and reinforces that behavior with reward, where is the aversive?
    If you're talking about purely standing around waiting for a dog to sit, then marking/rewarding, there is no aversive. That's not how training normally works however. Once the dog realizes you are a source of rewards, they tend to begin making an effort to gain that reward on a repetitive basis. We depend on this repetition to re-enforce behaviors on a reasonable time frame.

    The entire time that dog realizes you have the reward and is trying to figure out how to earn it, your withholding it creating stress, albeit mild stress, and is therefore aversive in nature.

    We humans like to think of this as a simple, stress free little game for a dog but with a highly motivated animal, you frequently see barking and yawning and other signs of frustration and stress during these positive sessions.

    We rationalize this as the dog trying different things to earn a reward. It helps us feel good about teasing them.

    Skinner has a fourth quadrant in the equation for this but it really is just a mild form of punishment.

    As I asked before... If I have something you want and I keep moving it around so you can't get it, is that not negative to you? frustrating? of course it is. It stands to reason then that right up to the moment the light comes on and the dog sits (for example), he's not having a completely stress free experience.

    My point is very simple. All training is stressful and to some degree aversive to the dog. We humans like to draw arbitrary lines to make ourselves feel good about what we're doing but at the end of the day, whether it's simple frustration or a good hard whack with a stick, we are always putting stress on the animal when we are training.

    We may try to choose the least stress method for any training we do but it is ALL stressful.

    If it were not, dogs would not go lay down immediately after 30 minutes of training. My customers are usually amazed that using nothing but a treat, I can make their puppy sleep for an hour or two, but I do it every day.
    Last edited by DarrinGreene; 09-09-2013 at 05:10 PM.
    Darrin Greene

  6. #36
    Senior Member Gun_Dog2002's Avatar
    Join Date
    Apr 2003
    Location
    Mohawk Valley
    Posts
    8,642

    Default

    Anytime you take control away from a dog, or inflict your will as opposed to the dogs will on a dog, that is inversion. Frankly this is why much of young puppy training with clickers and treats work so well as we basically let the dog do whatever he wants and inprint what what want him to do with a reward. Whenever you can make something the dogs idea of what he wants to do you will see success. This principle applies as they get older as well however we can't just let the dog do whatever he wants all the time. Therefore discipline and correction come into play. The key is knowing what to do and have the wisdom to do it properly. That is why these threads get tiresome to some, myself included. I can talk for hours on the principles of training, positive/negative aspects, methods and outcomes and then hand a person a dog on a leash with a heeling stick and they stand there lost. This is why every seminar includes field work with real people and real dogs in real situations. My experience is people learn through action vs lecture best.


    /Paul
    Paul Cantrell
    Black Ice Retrievers
    Marcola OR

    Too many dogs to list (By some Bitch)

    https://www.facebook.com/BlackIceRetrievers
    http://gundog2002.blogspot.com/
    "Helping Hunters Train Their Dogs"

  7. #37
    Senior Member PalouseDogs's Avatar
    Join Date
    Mar 2012
    Location
    Pullman, WA
    Posts
    275

    Default

    Maybe I'm dense, but I can't figure out the original question/comment. Are aversives (or whatever you want to call them) useful in training? Or was the question about examples of aversives?
    Kelly Cassidy (person)

    HR Maple Cassidy CDX JH RE (golden retriever)
    Alder Cassidy CDX RE (standard poodle chipmunk chaser)
    plus whacked-out weird Burka (elderly mix-breed rescue girl)

  8. #38
    Senior Member DarrinGreene's Avatar
    Join Date
    Feb 2007
    Location
    Souf Joisy
    Posts
    2,751

    Default

    Quote Originally Posted by Gun_Dog2002 View Post
    Anytime you take control away from a dog, or inflict your will as opposed to the dogs will on a dog, that is inversion. Frankly this is why much of young puppy training with clickers and treats work so well as we basically let the dog do whatever he wants and inprint what what want him to do with a reward. Whenever you can make something the dogs idea of what he wants to do you will see success. This principle applies as they get older as well however we can't just let the dog do whatever he wants all the time. Therefore discipline and correction come into play. The key is knowing what to do and have the wisdom to do it properly. That is why these threads get tiresome to some, myself included. I can talk for hours on the principles of training, positive/negative aspects, methods and outcomes and then hand a person a dog on a leash with a heeling stick and they stand there lost. This is why every seminar includes field work with real people and real dogs in real situations. My experience is people learn through action vs lecture best.


    /Paul
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^

    good post /Paul
    Darrin Greene

  9. #39
    Senior Member Evan's Avatar
    Join Date
    Jan 2003
    Location
    Kansas City, MO
    Posts
    5,140

    Default

    Quote Originally Posted by PalouseDogs View Post
    Maybe I'm dense, but I can't figure out the original question/comment. Are aversives (or whatever you want to call them) useful in training? Or was the question about examples of aversives?
    It was just a commentary about a common term. Post #34 is a more condensed description.

    Yes, aversives a are very useful commonly used.

    Evan
    "Prepare your dog in such a manner that the work he is normally called upon to do under-whelms him, not overwhelms him." ~ Evan Graham

    “People who say it cannot be done should not interrupt those who are doing it.”

    ― George Bernard Shaw


    The Smartwork System for Retriever Training (link)
    http://www.facebook.com/profile.php?...59&ref=profile

  10. #40
    Junior Member
    Join Date
    Apr 2013
    Posts
    9

    Default

    Quote Originally Posted by DarrinGreene View Post
    If you're talking about purely standing around waiting for a dog to sit, then marking/rewarding, there is no aversive. That's not how training normally works however. Once the dog realizes you are a source of rewards, they tend to begin making an effort to gain that reward on a repetitive basis. We depend on this repetition to re-enforce behaviors on a reasonable time frame.

    The entire time that dog realizes you have the reward and is trying to figure out how to earn it, your withholding it creating stress, albeit mild stress, and is therefore aversive in nature.

    We humans like to think of this as a simple, stress free little game for a dog but with a highly motivated animal, you frequently see barking and yawning and other signs of frustration and stress during these positive sessions.

    We rationalize this as the dog trying different things to earn a reward. It helps us feel good about teasing them.

    Skinner has a fourth quadrant in the equation for this but it really is just a mild form of punishment.

    As I asked before... If I have something you want and I keep moving it around so you can't get it, is that not negative to you? frustrating? of course it is. It stands to reason then that right up to the moment the light comes on and the dog sits (for example), he's not having a completely stress free experience.

    My point is very simple. All training is stressful and to some degree aversive to the dog. We humans like to draw arbitrary lines to make ourselves feel good about what we're doing but at the end of the day, whether it's simple frustration or a good hard whack with a stick, we are always putting stress on the animal when we are training.

    We may try to choose the least stress method for any training we do but it is ALL stressful.

    If it were not, dogs would not go lay down immediately after 30 minutes of training. My customers are usually amazed that using nothing but a treat, I can make their puppy sleep for an hour or two, but I do it every day.
    In order to limit the 'aversiveness' that the dog is subjected to when a trainer withholds a treat until the desired behaviour is performed, good trainers frequently shape the final behaviour in small approximations; if the dog is showing stress signals such as barking, yawning etc, then the trainer is trying to change the behaviour in a much too big a leap. Training should be structured so that the dog is able to easily shift behaviour resulting in as little stress as possible.

    But of course Darrin is correct in saying that training by its very nature, no matter what method or technique is used is punishing some behaviours and reinforcing others. Personally I try to make that process as least aversive as I possibly can for the dog while aiming for clear communication of what is required.

    Wendy

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •