Article - your thoughts?

[Danefied]

I'd be really interested in reactions, responses to this article. IMO its the typical "PR works for sweet dogs, but "tougher" dogs require tougher methods" line. I don't agree as it has not been my experience at all with our "tough" cases. IME that's where creatively applied PR really shines.
Anyway, opinions?
“Positive” Bias? Dogs In Training

 

[lizzybeth727]

sorry, double post

 

[lizzybeth727]

Yeah, the author's arguments are not very good.

“Purely positive” or clicker trainers would have you believe that the correct response is a verbal “ah-ah” or simply ignoring the behaviour. ....If you are faced with a determined breed or individual, who is say… constantly eating things on the walk or greeting Grandma at the door by using his 80lbs to rear up and look her in the face just long enough to get a good hold and go in for an epic hump – well then, things are a bit different, aren’t they?

I know a lot of clicker trainers, and I don't know ANY who would tell Grandma to ignore the 80 pound GSD who's trying to hump her.

A lot of "balanced" or punishment-based trainers use this argument against clicker trainers, that clicker trainers think ALL bad behaviors should just be ignored. But this is not true. I'm just about as purely postive in my training as one can be, and there are very few bad behaviors that I simply ignore. Just because I don't use punishment doesn't mean I just let the dog run amok.

In fact, the creators of modern “positive” dog training eventually wrote a paper stating that their original conclusions were incorrect – and that operant conditioning broke down readily in the face of instinctual behaviours.... Of course, you’ll never read that little tidbit in a manual on clicker training, will you?

I think they might be talking about the work done by the Brelands, very interesting psychologists who trained dozens of different species, using positive reinforcement. But their work was done in the 1940's I think.... and most lay traners don't really care enough for it to be in a "manual on clicker training." AND, there ARE books that talk about how to use positive reinforcement training to help curb instinctual behaviors... such as "When Pigs Fly," which I would bet this author has not read.

I have never had a client come to me with a difficult case after seeing a balanced trainer.

This makes sense.... If one balanced trainer doesn't work, why should you go to another?

In the picture of the dog with the head halter, the halter is NOT fitted correctly.

The article seems to really only talk about choke chains and head halters. But personally, these both are pretty much last-resort devices for me. I'd much rather suggest a front clip harness to a client than a head halter, and I'm not opposed to prong collars either.

 

[corgipower]

I'd be really interested in reactions, responses to this article. IMO its the typical "PR works for sweet dogs, but "tougher" dogs require tougher methods" line. I don't agree as it has not been my experience at all with our "tough" cases. IME that's where creatively applied PR really shines.

"Creatively applied PR" is very difficult for people to learn and teach though. It's a lot easier to apply a collar correction.

One of the articles linked to Classics in the History of Psychology -- Breland & Breland (1961) under the premise that

operant conditioning broke down readily in the face of instinctual behaviours

is useless. Yes, the animal might have found something self reinforcing to do instead of the desired behavior, but if the person conducting the training sat there and allowed a self rewarding behavior to continue until it got out of hand while the desired behavior deteriorated, that's poor training. Another reason that article is useless is because the "experiments" were done on chicken, raccoon and pig. Last I knew, my dogs were not chickens, raccoons or pigs. Considering that the article is from 1961, that may have been where training was at when it was written...I have no idea, I wasn't alive yet...but training and specifically PR has come a long way since then.

Hogwash.

Self rewarding behaviors are only rewarding if they're allowed to be. Remove the reward or replace it with something appropriate. And natural instincts are wonderful for channeling into desirable directions and using for premack rewards.

 

[Shai]

IMO that article is so very slanted and written with such an obvious agenda as to render it utter hogwash.

Self rewarding behaviors are only rewarding if they're allowed to be. Remove the reward or replace it with something appropriate. And natural instincts are wonderful for channeling into desirable directions and using for premack rewards.

Well said

 

[Meatos]

That article really gets my goat. The author claims to be "balanced" but I really don't get a sense of "balance" from that article. In fact, it seems pretty one-sided by slamming (and distorting) aspects of PR training such as ignoring behaviours (last I checked, extinction bursts are a scientific fact), and managing behaviours (management is a fact of life - and keeps dogs alive).

I don't like the confusion it presents about head halters and body harnesses. YES these ARE training tools, not management tools, as the article states. Sadly, most people use them as management tools and don't take the training further. But hey, couldn't we say that for prong collars? e-collars? or even...dare I say it...a regular old leash and flat collar?

I'm confused as to why head halters are a bad option because a dog might run after a squirrel and snap its neck, but the article also states that these are an excellent tool for dogs that lunge. I fail to see why one is safe and not the other?

A dog with an odd gait due to the harness (I've never seen this and I walk 18 dogs a day, several of which wear harnesses) is likely a dog who has still not learned how to walk nicely on a leash, despite the tool. This is the fault of the handler, not the tool itself.

I think the point that the article misses is that corrections-based training (done right) builds obedient dogs who suppress behaviours, but PR training (done right) builds obedient dogs who are SAFE. Like Danefied said, this is the area where PR training really shines.

EDIT: Great responses from those who have responded so far. Another forum I'm on has a "thank" button that allows you to thank a useful post. Wish we had that here because I would be thanking them all.

 

[Danefied]

Thanks for the replies. I'm not gonna lie, I'm exhausted with the same old argument over and over. I say you don't have to rough up your dog to get results, and I'm told my dogs are "easy" or my favorite "I'm not going to ignore bad behavior." Seriously, where did that BS get started?!

It IS a lot easier to apply a collar correction, and no, its not the end of the world if that's what you end up doing, but its not the only way, its not always the most effective way either.

I'm gonna have to type up Lunar's story one day. He was our first rehab case after having had children, and that's where it really came home to me that we had to use a method that created not just an obedient dog, but a SAFE dog - a dog I could trust to behave even when DH or I weren't there to enforce commands, a dog with safe default behaviors.

I do tend to agree with a lot of what Dunbar writes about operant conditioning not being all its cracked up to be in the real world with sentient creatures with minds of their own, but that doesn't mean you dismiss it entirely. And operant conditioning DOES work to teach behaviors. Then use use the Premack principle, or real life rewards to proof the behavior under distractions.

 

[corgipower]

I do tend to agree with a lot of what Dunbar writes about operant conditioning not being all its cracked up to be in the real world with sentient creatures with minds of their own, but that doesn't mean you dismiss it entirely. And operant conditioning DOES work to teach behaviors. Then use use the Premack principle, or real life rewards to proof the behavior under distractions.

I haven't read what Dunbar says about operant conditioning, but really, there's no way to avoid operant conditioning. Premacking is operant conditioning. Self rewarding behaviors is operant conditioning. Collar corrections are operant conditioning.

 

[Danefied]

I haven't read what Dunbar says about operant conditioning, but really, there's no way to avoid operant conditioning. Premacking is operant conditioning. Self rewarding behaviors is operant conditioning. Collar corrections are operant conditioning.

Oh, I totally agree! But *most* people think of the four quadrants in very limited terms - usually food rewards and physical punishments. And the truth is many dogs could care less about a hot dog when they're in the middle of a good live chase (many could care less about a punishment either - as seen by the highly prey driven dog gritting his teeth and blowing through the underground fence correction to go after the squirrel).
Add to that how we're so indoctrinated against rewarding "bad" behavior, and things like the Premack principle really blow people's mind. "You mean I stop him from chasing squirrels by rewarding him with chasing squirrels?"

I'll have to dig up some of the Dunbar articles I'm thinking about...

 

[release the hounds]

oh boy, go figure, a dog training article that is biased. Haven't ever seen any of those before

 

[elegy]

“Purely positive” or clicker trainers would have you believe that the correct response is a verbal “ah-ah” or simply ignoring the behaviour. This is fine if the offender is a sweet, stable dog who is sensitive enough to consider this a correction

No. No, dear author, that is not it at all. That is COMPLETELY AND UTTERLY missing the point of positive training altogether.

The positively trained dog is not learning from an "ah ah" or ignoring the behavior as a correction because the positive trainer is not TEACHING the dog through correction. You're missing the entire part of the equation in which the trainer is creating and rewarding the correct behavior so that they don't need to correct the wrong one.

ARGHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

And as for positive training only being for soft, malleable dogs, that's poop. That's the result of somebody who doesn't know how to train dogs without yanking, cranking, and suppressing. That's somebody who hasn't bothered to learn what positive training is all about, but rather, like so many, has learned the stereotypes and dismissed it without bothering to explore and learn further.

 

[Danefied]

Eh, I don't know if it bothers me that its biased. I mean, of course trainers are going to talk up what has worked for them. I know I do

 

[lizzybeth727]

I haven't read what Dunbar says about operant conditioning, but really, there's no way to avoid operant conditioning. Premacking is operant conditioning. Self rewarding behaviors is operant conditioning. Collar corrections are operant conditioning.

^^^THIS. I know a lot of trainers that say "I don't use operant conditioning in dog training." Which only tells me that the trainer doesn't know what operant conditioning IS.

The Brelands were in the 60's? Oops, I guess I need to review my history of psychology.

On a small tangent, their research on training instinctual behaviors was really interesting and not really what the author of this article thinks it is. They called it Instinctive Drift, really interesting, IMGO (In My Geeky Opinion, LOL).

 

[DogsInTraining]

Response1

@Danefied

"PR works for sweet dogs, but "tougher" dogs require tougher methods"

This was not my point at all. My point was that while PR may be effective for many, many things, dealing with reliability and self-rewarding behaviours is simply not one of those things.

My point was that there are four quadrants for a reason, and even the Brelands grew to accept this.

@lizzybeth727

I know a lot of clicker trainers, and I don't know ANY who would tell Grandma to ignore the 80 pound GSD who's trying to hump her.

A lot of "balanced" or punishment-based trainers use this argument against clicker trainers, that clicker trainers think ALL bad behaviors should just be ignored. But this is not true. I'm just about as purely postive in my training as one can be, and there are very few bad behaviors that I simply ignore. Just because I don't use punishment doesn't mean I just let the dog run amok.

This is my all time favourite argument from pure PR trainers. "I don't believe in corrections" somehow also equals "I don't let him get away with that!"

This is where the term "management" applies. Where a balanced trainer would ensure that the dog had a good "sit" and a good "off" trained well and then apply a small number of properly timed corrections to ensure that jumping on anyone was no longer self-rewarding, a PR trainer would develop an entire management scenario. By all means, correct me if I'm wrong.

I think they might be talking about the work done by the Brelands, very interesting psychologists who trained dozens of different species, using positive reinforcement. But their work was done in the 1940's I think.... and most lay traners don't really care enough for it to be in a "manual on clicker training." AND, there ARE books that talk about how to use positive reinforcement training to help curb instinctual behaviors... such as "When Pigs Fly," which I would bet this author has not read.

First, as a PR trainer you shouldn't have to wonder whether I am quoting the Brelands or not. Students of Skinner himself, they are the founders of this training technique, and between them trained tens of thousands of creatures, from ducks to killer whales. The Brelands actually pioneered the techniques that train killer whales and dolphins.

I find it interesting that you can toss the information they gathered aside because it's "from the 1940's." Like it or not, that is when and how the technique that you and your fellow PR trainers are using and defending. If it wasn't for the Brelands taking Skinner's information and turning it into a business venture, it's quite likely that PR training would not exist today - and that's a scary thought.

As for "When Pigs Fly" I not only have read it, I own it. In fact, let me provide you a quote. Not only does Killion admit that positive punishment has it's place in certain situations, here's what she says about punishment in general:

When your dog does something right he is rewarded. When he does something that you do not want him to do, he is ignored, reinforcements are withheld, or he is given a time-out.

If the behaviour is self reinforcing, as the Breland's point out, this method of intervention is not going to work. You are working on a reinforcement schedule that essentially has a red or a green button - and you have a dog whose own behaviour pushes the green button in his mind. Take away all the toys, treats and activities you want, that dog is still going to chase that cat if he is so inclined - simply because he's driven to - and it's fun!

A lot of "balanced" or punishment-based trainers use this argument against clicker trainers, that clicker trainers think ALL bad behaviors should just be ignored. But this is not true. I'm just about as purely postive in my training as one can be, and there are very few bad behaviors that I simply ignore. Just because I don't use punishment doesn't mean I just let the dog run amok.

The fact that you use "balanced" and "punishment based" in the same handful merely points out your ignorance. Just as there are "positive" trainers out there that many of you surely wish would not claim to be so, there are many "balanced" trainers who don't deserve to use this term either. "Balanced" training is not anything-based. It's based on the dog sitting in front of the trainer, the problem, the owner's abilities and environmental factors. First choice should ALWAYS be positive methods, I simply have no problem also throwing in a leash correction. As for how you prevent your dog from "running amok", please do share. I love the vagueness that comes from PR trainers...

This makes sense.... If one balanced trainer doesn't work, why should you go to another?

In the picture of the dog with the head halter, the halter is NOT fitted correctly.

The article seems to really only talk about choke chains and head halters. But personally, these both are pretty much last-resort devices for me. I'd much rather suggest a front clip harness to a client than a head halter, and I'm not opposed to prong collars either.

I wish people thought that way about training methods - sadly, by the time dogs get to me they've frequently been through more than one trainer. If people really looked at the methods I would have far more clients arriving after their first "PR" trainer failed.

Front clip harnesses are a death trap and far more inhumane (IMO) than most other tools available. I'm sure you'll LOVE my article on front-clip harnesses. As for your comment about head halters, you'll also LOVE my article on the Not-So-Gentle-Leader (not being sarcastic this time, you may actually enjoy it).

Seriously though, the reason that check chains and head halters were the two pieces of equipment I focused on was simply because during the conversation that sparked the article these were the only two pieces that were brought up. Not every article I write or discussion I have about training methods is going to include every possible training or management tool, and the reality is that "positive" training tools are just as likely to be aversive or dangerous as "traditional" tools.

You yourself, Lizzy, use prong collars. You know, in most PR circles this will get you ostracized. I find it very interesting that you can rail against an article in which I suggest that purely positive may not be sufficient and end it by endorsing prong collars (which, while I use them frequently myself, can not be called anything but a very clear aversive). In FACT, with a comment like "I'm just about as purely postive in my training as one can be" followed by an endorsement of prong collars suggests to me that you are NOT a "PR" trainer, and you are most CERTAINLY not "as purely positive as one can be". In FACT, you appear to me from the comments made here to be a very balanced trainer, whether you're ready to admit it to yourself or not.

@corgipower

"Creatively applied PR" is very difficult for people to learn and teach though.

I agree with you 250%. To be a truly talented PR trainer, not only the timing and technique but the problem solving end of things - it ain't easy. And this, I believe, is part of the problem. Average pet owners can't grasp it easily enough and end up bringing more frustration into the relationship. By teaching owners how to use positive methods while simultaneously teaching them how and when to apply a proper correction (and no, I don't mean a nagging correction, nor do I mean a knock-him-over correction. I mean a correction that puts a stop to the behaviour and refocuses the dog's attention back on the other end of the leash) is how we obtain the most reliable results.

Yes, the animal might have found something self reinforcing to do instead of the desired behavior, but if the person conducting the training sat there and allowed a self rewarding behavior to continue until it got out of hand while the desired behavior deteriorated, that's poor training. Another reason that article is useless is because the "experiments" were done on chicken, raccoon and pig. Last I knew, my dogs were not chickens, raccoons or pigs. Considering that the article is from 1961, that may have been where training was at when it was written...I have no idea, I wasn't alive yet...but training and specifically PR has come a long way since then.

Again, some ignorance showing through here that no one is even familiar with the Breland's work. I suggest you all get acquainted, you owe your methodology to these two people. The experiments they keyed in on to illustrate the "Misbehaviour of Organisms" included a raccoon and two chickens, yes, but these two had trained tens of thousands of animals, including dogs.

Your assertion that dogs are not chickens or pigs is interesting, considering that the first argument most PR trainers go to is the dolphin/whale training argument (interesting, huh, that in a situation where the orca trainers literally control every aspect of the animal's environment there are still limits to this training as evidenced by the number of orca trainers who are attacked or killed by these animals... call it a freak accident if you will, but I call it a breakdown in behaviour).

As for the chickens, isn't that the animal du jour for PR seminars? See how fast you can clicker train the chicken?

No, a dog is not a chicken, or a raccoon, or a pig - but neither is he a dolphin, a whale or a laboratory animal. But he is an animal. And either you believe in behaviourism or you do not (likely you do not, because behaviourism contains four quadrants while PR only focuses on two - look forward to a future article on that little topic when I review "Determinants of Animal Behaviour" alongside a PR title that I haven't yet chosen).

 

[DogsInTraining]

Response2

Self rewarding behaviors are only rewarding if they're allowed to be. Remove the reward or replace it with something appropriate. And natural instincts are wonderful for channeling into desirable directions and using for premack rewards.

Uh - no, they are not. They are described as SELF-rewarding, because they are SELF-rewarding. Period. You can not "remove the reward" in a self rewarding behaviour. A retriever gets an internal high from retrieving. Herding dogs from herding. Many sight hounds from chasing, hounds from smelling, the list goes on and on. Unless you apply an aversive or you are investigating canine lobotomies, animals are always going to be self-reinforced by their own behaviour. Sometimes these are perfectly rational behaviours, other times they are obsessive in nature. Sometimes they are beneficial, other times these behaviours interfere with our image of a "well behaved" dog.

@shai (who wins the award for the most thoughtful post, no?)

IMO that article is so very slanted and written with such an obvious agenda as to render it utter hogwash.

Please, do tell. What would be the slant, and what would be the agenda?

For that matter, care to share an article written by a PR trainer about PR training without a slant or an agenda?

@everyone

Look, my intent coming here was not to cause a ruckus. I noted the comments on my trackback, and I'm honestly interested in some clarification. I'm absolutely SICK of the infighting between factions. I'm tired of being told that I'm abusive, cruel or somehow a poor trainer because I use aversives, or I choose tools like prong collars over head halters. Instead of our profession learning these new techniques and using them to improve our overall effectiveness, many of us abandoned traditional techniques all together - you have to admit that the far sides of each faction has become, dare I say it, cult-ish in nature.

I happen to believe that it's a combination of the two that produces the best and most reliable results. Which is why while most trainers in our area are PR trainers, there's not a single dog at the top of competitive obedience or protection who was trained in a "purely positive" manner. In fact, Ian Dunbar himself has been quoted as stating that in order to compete at a comparable level one must apply aversives.

My "agenda" as it were, was to respond to the conversation I had witnessed the night before. Nothing less, nothing more.

@meatos

That article really gets my goat. The author claims to be "balanced" but I really don't get a sense of "balance" from that article. In fact, it seems pretty one-sided by slamming (and distorting) aspects of PR training such as ignoring behaviours (last I checked, extinction bursts are a scientific fact), and managing behaviours (management is a fact of life - and keeps dogs alive).

I'm sorry, would you care to point out where I denied the reality of extinction in behaviourism? The issue I have is where the owner is being told to ignore behaviours that should at least be interrupted if not corrected. Management does keep dogs alive, but is it a suitable alternative to avoid giving a dog a correction? Now THAT is hogwash.

I don't like the confusion it presents about head halters and body harnesses. YES these ARE training tools, not management tools, as the article states. Sadly, most people use them as management tools and don't take the training further. But hey, couldn't we say that for prong collars? e-collars? or even...dare I say it...a regular old leash and flat collar?

I disagree. Prong collars and e-collars both apply a stimulation or an aversive when the handler or the dog takes a particular action. The head halter and the body harness are both aversives even when the dog is standing still, or walking at heel. These are management tools, they provide NO feedback. Can you train a dog to walk on a loose leash while using a management tool? Sure! But the tool itself is simply to keep the dog manageable while you do so - the tool is not teaching the dog anything.

I'm confused as to why head halters are a bad option because a dog might run after a squirrel and snap its neck, but the article also states that these are an excellent tool for dogs that lunge. I fail to see why one is safe and not the other?

If you can't see the difference between the ability to control the direction of a dog's face and muzzle to avoid a lunging behaviour (a dog can not lunge left while facing right) and what occurs when a dog hits the solid end of six feet of slack leash absorbing all energy into the head and neck - well, I don't know what to say.

A dog with an odd gait due to the harness (I've never seen this and I walk 18 dogs a day, several of which wear harnesses) is likely a dog who has still not learned how to walk nicely on a leash, despite the tool. This is the fault of the handler, not the tool itself.

I think the point that the article misses is that corrections-based training (done right) builds obedient dogs who suppress behaviours, but PR training (done right) builds obedient dogs who are SAFE. Like Danefied said, this is the area where PR training really shines.

I'm not speaking to all harnesses, rather the front clip variety such as the Easy Walk. If you can't see the altered gait, perhaps you should video tape it and take it to your nearest veterinary specialist and see what they think. They change the dog's gait - the strap goes directly across the dog's humerus bone and can't NOT change the dog's gait.

As for your "obedient" vs. "safe" argument, wow what a giant straw man. Once again, why are these my only options? I agree with you that corrections-based training is not ideal. But I also believe that PR training is not ideal either. I believe that training is best achieved when a positive base is constructed and then enforced using corrections.

Back @danefield

It IS a lot easier to apply a collar correction, and no, its not the end of the world if that's what you end up doing, but its not the only way, its not always the most effective way either.

I do tend to agree with a lot of what Dunbar writes about operant conditioning not being all its cracked up to be in the real world with sentient creatures with minds of their own, but that doesn't mean you dismiss it entirely. And operant conditioning DOES work to teach behaviors. Then use use the Premack principle, or real life rewards to proof the behavior under distractions.

I agree with you. With one caveat - Premack is great, when it's physically possible. Using "ok, go chase squirrels" as a reward for... not chasing squirrels. Great, in theory. Obviously this is not possible (or desirable) for most dog owners. Maybe I live with small animals, and NO chasing/focusing/intensity towards small animals is acceptable. Maybe I live on a busy street and chasing squirrels isn't possible. Maybe I just feel bad for the poor squirrel, whatever.

The point is that real life sometimes gets in the way, and in a single afternoon using both positive rewards and aversives I can have that squirrel chaser CHOOSING to ignore the squirrel.

The point is that there isn't one single BEST possible solution.

The point is that I'm not "evil" or "cruel" or "abusive" because my dogs receive corrections, or because I teach my clients when and how to properly dole one out.

The point is that I'm not against the methods of PR - I'm against the use of PR to the exclusion of every other possible solution - the same way I'm against the exclusive use of ANY methodology.

My apologies for the length, when I started writing there were only a few comments.

Ok, fire away.

 

[Shai]

Uh - no, they are not. They are described as SELF-rewarding, because they are SELF-rewarding. Period. You can not "remove the reward" in a self rewarding behaviour. A retriever gets an internal high from retrieving. Herding dogs from herding. Many sight hounds from chasing, hounds from smelling, the list goes on and on. Unless you apply an aversive or you are investigating canine lobotomies, animals are always going to be self-reinforced by their own behaviour. Sometimes these are perfectly rational behaviours, other times they are obsessive in nature. Sometimes they are beneficial, other times these behaviours interfere with our image of a "well behaved" dog.

Or, you can prevent or allow access to the self-rewarding activities, turning them to your advantage. Of course you have to be willing the set up situation to your advantage in order to do this.

My terrier and my herder learned to stop chasing rabbits/deer on command by...being allowed to chase rabbits :yikes:.

My retriever returns birds, which she would rather keep and nom, because she has been trained to. How? By earning the right to do another retrieve by completing the current retrieve according to my criteria. Same for being steady at the line. There is no reward in the world to which she will give the time of day while doing field work except for the chance to do more field work.

@shai (who wins the award for the most thoughtful post, no?)

You're too kind.

Please, do tell. What would be the slant, and what would be the agenda?

The agenda would be painting those who use clickers as killing sweet grandmas, clearly.

Just kidding. Mostly.

You're essentially picking an extremist sect and using that as an excuse to marginalize a training method that does, in fact, work. Painting all clicker trainers as simpering fools pleading with dear Fido to please, please stop doing that dearie while dangling haplessly from a head halter is not exactly original, but repetition doesn't make it truth.

As you yourself said, any good trainer studies and understands all methods and tools, and chooses that which works best for them in a given situation. No matter which end of the spectrum you consider, stupidity occurs. I would contend that this stems from a lack of understanding of the method, NOT from the method itself.

For that matter, care to share an article written by a PR trainer about PR training without a slant or an agenda?

The argument "well...they do it too!" has never really held much sway with me, honestly.

 

[corgipower]

They called it Instinctive Drift, really interesting, IMGO (In My Geeky Opinion, LOL).

I will agree with your geeky opinion.


First, Welcome to Chaz, DogsInTraining. I hope you stick around for more than just this thread. We may or may not agree with you, but we're always up for training discussions.

This was not my point at all. My point was that while PR may be effective for many, many things, dealing with reliability and self-rewarding behaviours is simply not one of those things.

Reliability?
How reliable is correction based training? Or "balanced" training?
Training is NEVER going to be 100% reliable. No matter what the methods.

This is my all time favourite argument from pure PR trainers. "I don't believe in corrections" somehow also equals "I don't let him get away with that!"

No, it doesn't equal. I can believe in corrections and still allow my dog to get away with all sorts of stuff.

Heck, if you walked into my house, you'd probably be horrified at my dogs' behavior. Are they trained? Absolutely! Do I allow them to jump and bark and mouth? You betcha I do. Can I stop them mid behavior? Most definitely. It's my choice to allow it and not at all dependent on the method I'm using.

This is where the term "management" applies. Where a balanced trainer would ensure that the dog had a good "sit" and a good "off" trained well and then apply a small number of properly timed corrections to ensure that jumping on anyone was no longer self-rewarding, a PR trainer would develop an entire management scenario. By all means, correct me if I'm wrong.

Management still plays when doing balanced training. Until the dog DOES have a good sit or good off, I'm managing him and his environment regardless of what method I use to train him. Otherwise, those "properly timed corrections" aren't going to happen either.

First, as a PR trainer you shouldn't have to wonder whether I am quoting the Brelands or not. Students of Skinner himself, they are the founders of this training technique, and between them trained tens of thousands of creatures, from ducks to killer whales. The Brelands actually pioneered the techniques that train killer whales and dolphins.

:rofl1:
So all PR trainers should know Brelands and Skinner? opcorn:

I find it interesting that you can toss the information they gathered aside because it's "from the 1940's." Like it or not, that is when and how the technique that you and your fellow PR trainers are using and defending. If it wasn't for the Brelands taking Skinner's information and turning it into a business venture, it's quite likely that PR training would not exist today - and that's a scary thought.

It's not a matter of tossing it aside. It's a matter of over the course of the last 70 years, a lot more has been learned about dog training and we have much more current information to cite.

Would you go to a doctor who was still using medicine from 70 years ago? Or would you prefer one who was a little more up on today's choices?

Your assertion that dogs are not chickens or pigs is interesting, considering that the first argument most PR trainers go to is the dolphin/whale training argument.

Some do...I'm not one of them. Dogs are not dolphins, whales, chickens, pigs or wolves. They are a unique species in so far as having evolved alongside humans for thousands of years. That evolution enables them to have a working relationship with us far beyond what chickens, whales, dolphins or any other species can have.

They are described as SELF-rewarding, because they are SELF-rewarding. Period. You can not "remove the reward" in a self rewarding behaviour. A retriever gets an internal high from retrieving. Herding dogs from herding. Many sight hounds from chasing, hounds from smelling, the list goes on and on. Unless you apply an aversive or you are investigating canine lobotomies, animals are always going to be self-reinforced by their own behaviour. Sometimes these are perfectly rational behaviours, other times they are obsessive in nature. Sometimes they are beneficial, other times these behaviours interfere with our image of a "well behaved" dog.

If my dog is jumping on me as a self rewarding behavior -- keep in mind, it's not the jumping that's the end reward here, it's the me -- I can absolutely remove the reward, namely myself. Very quickly the jumping on me becomes a whole lot less rewarding. If, simultaneously I reward the dog with myself when he offers all four feet on the floor, that behavior replaces the jumping.

If the act of jumping truly is the reward he's going for, I can redirect that into jumping over an obstacle or onto a bench or through a hoop.

In herding training we teach the dog to lie down. We do so by immediately rewarding the lie down with a chance to go run the stock again.

It's not a matter of the behaviors not being self rewarding, it's a matter of putting the behaviors on cue, teaching the dog that he can earn a chance to perform them, and in some cases teaching him that there are better ways to get what they're really after.

Dogs are such opportunists, we can use that to steer their behaviors in the direction we want.

The issue I have is where the owner is being told to ignore behaviours that should at least be interrupted if not corrected. Management does keep dogs alive, but is it a suitable alternative to avoid giving a dog a correction?

Telling owners to ignore behaviors is valuable in so far as most of the time, they inadvertently reinforce behaviors. Dog jumps, owner pushes it away, dog thinks this is an awesome game and jumps again. Ignoring the jumping would be better than reinforcing it unintentionally.

 

[DogsInTraining]

Shai, you failed to answer my most basic questions, again sticking to the vagueness that PR trainers seem to adore.

Or, you can prevent or allow access to the self-rewarding activities, turning them to your advantage. Of course you have to be willing the set up situation to your advantage in order to do this.

My terrier and my herder learned to stop chasing rabbits/deer on command by...being allowed to chase rabbits .

Preventing access. Back to management we go!

My three herding dogs do not chase vermin of any kind because they have been trained - using both positive and aversive methods. Choosing to chase the animal results in a correction and an end to the outing. Choosing to return to me results in a reward. We do not live in an area that permits allowing our dogs to chase wildlife - in fact, not only is it frowned upon, it's extremely dangerous (in most areas, not just my own). It's kinda like saying I prevent my kid from going out and getting drunk by getting him drunk at home. If your end goal is to have a child that chooses NOT to drink, you're going about it the wrong way.

The agenda would be painting those who use clickers as killing sweet grandmas, clearly.

Again, point out where I said that and please stick to the facts. The number one response by PR trainers dealing with dogs who jump up is to ignore the behaviour. Stand like a tree, cross your arms, turn your back, etc. If you'd like me to quote a half dozen PR books, I most happily will. Obviously, this is not acceptable in certain situations.

You're essentially picking an extremist sect and using that as an excuse to marginalize a training method that does, in fact, work. Painting all clicker trainers as simpering fools pleading with dear Fido to please, please stop doing that dearie while dangling haplessly from a head halter is not exactly original, but repetition doesn't make it truth.

As you yourself said, any good trainer studies and understands all methods and tools, and chooses that which works best for them in a given situation. No matter which end of the spectrum you consider, stupidity occurs. I would contend that this stems from a lack of understanding of the method, NOT from the method itself.

My issue was not with clicker trainers. It was with those trainers who identify themselves as "purely positive." They are not a rarity, and those who claim that practitioners of all other methods are cruel and confused are not a rarity either. In fact, one organizing body actually states that it is against their code of ethics for a member to even recommend ANOTHER TRAINER who uses aversive techniques.

Interesting accusation, as well, this stereotyping - did I not previously read a sentence in which balanced and force-based trainers were lumped together?

More than any other branch of training, the "purely positive" group seems to truly believe that their method is superior in all ways in all situations. The facts are that it is not.

The argument "well...they do it too!" has never really held much sway with me, honestly.

Perhaps not, but I'll give you a dollar if you can find a comparing article written by a positive trainer that doesn't at least insinuate that the rest of us are uneducated brutes. To hold my feet to the fire in a situation where you would not do the same to one of your own shows... what were the words? Oh yeah, a "slant" or an "agenda."

 

[Linds]

Well, let me just say that I thought your two posts DogsInTraining were very very well worded, very nicely put especially when your own article was getting lambasted and I pretty much agreed with everything you said. I know, I know, shocker to everyone on here *snort*

There are things I would do different and things that I might not be completely insync with but that's training for you

I'm honestly interested in some clarification. I'm absolutely SICK of the infighting between factions. I'm tired of being told that I'm abusive, cruel or somehow a poor trainer because I use aversives, or I choose tools like prong collars over head halters. Instead of our profession learning these new techniques and using them to improve our overall effectiveness, many of us abandoned traditional techniques all together - you have to admit that the far sides of each faction ahas become, dare I say it, cult-ish in nature.

I especially agree with this little tidbit here.

I honestly don't care how you train for the most part but I am sick and tired of the extremes and using said extremes to prove points.

There is no always in training, there is no never in training, there are no absolutes in training and there is very rarely a one size fits all answer to training and I think at the end of the day pretty much all of use can agree on that

Welcome to the forum, I would love if you stuck around because I really like the way your think and write

 

[Danefied]

@Danefied
This was not my point at all. My point was that while PR may be effective for many, many things, dealing with reliability and self-rewarding behaviours is simply not one of those things.

My point was that there are four quadrants for a reason, and even the Brelands grew to accept this.

Tell me, where in PR training does it say you cannot use all four quadrants? I use P+, I just don't believe P+ has to be painful, scary, or intimidating to be effective. And say what you will, popping a check chain is not fun for any dog - if it were it wouldn't work now would it?

If you know your dog and their motivations, build your relationship, you can use a "look" as P+ never have to even touch the dog.
I smack my dogs on the butt all the time and they think its a play invite, I can look away and you'd think I had hit them.

Back @danefield
I agree with you. With one caveat - Premack is great, when it's physically possible. Using "ok, go chase squirrels" as a reward for... not chasing squirrels. Great, in theory. Obviously this is not possible (or desirable) for most dog owners. Maybe I live with small animals, and NO chasing/focusing/intensity towards small animals is acceptable. Maybe I live on a busy street and chasing squirrels isn't possible. Maybe I just feel bad for the poor squirrel, whatever.

The point is that real life sometimes gets in the way, and in a single afternoon using both positive rewards and aversives I can have that squirrel chaser CHOOSING to ignore the squirrel.

Aversives work, true but how far are you willing to go?
My mutt dog would regularly get the $hit beat out of him for killing chickens in his previous home - like 7 month old puppy had to get his wind back before he could get back up recovering from the kick with a steel toed boot. And when he got back up guess what he did? Went right back to the chickens.
I use the elevator button analogy when it comes to punishment. When the elevator doesn't come right away, what do we do? Push the button harder, more often - even though intellectually we KNOW the elevator isn't going to come any faster. We still push and push and push. That's what happens with punishment. Collar pop doesn't work? Try a prong. Prong doesn't work? Go to an e-collar. Dial up the e-collar. MOST dogs will respond to the collar pop, and many the prong - these tools DO work, and used properly they are not inhumane at all. My issue though, is the mentality it creates - that elevator button mentality. And I for one refuse to teach "weekend warrior" type dog owners that mentality.

Back to my mutt dog. I used McDevitt's exercises to teach him to RELAX while his prey drive was stimulated, and then yes, also used the Premack principle. Not always with allowing him to chase what he originally wanted to, but giving him an appropriate substitute of equal value to him - a game of tug instead of pouncing on the yorkie. Some owners use flirt poles as rewards... You can make Premack work if you're creative enough See there's where the creative part comes in
I accidentally taught our feral rescue recall by falling on my azz in a ravine. This was after months and months of using all the tricks to no avail. Finally I realized that putting myself in a vulnerable position allowed him to get over his fear of getting "caught" and learn what I meant by "come".

And honestly, that's a huge draw for me in PR training - the figuring the dog out part. Figuring out what motivates them, how you can make rewards (and punishments) meaningful for that dog, how to alter the motivation for those self rewarding behaviors... Its good stuff. And it works