A Trainers Debate With The Rspca

[Erny]

I think correction training can go terribly wrong for the dog in the hands of the inexperienced dog trainer ....

But there is a point also made in this thread that +R based training CAN actually go wrong for the dog.

It makes my skin crawl to see a novice handler with a check chain, to use it effectively you need some degree of experience, some people have the knack but most don't.

As it does for me when I see head-collars frequently being used incorrectly. Yet RSPCA refer to this style of equipment as "positive". I see them as carrying bias rather than open and fair mindedness when it comes to their own opinions. And given that so many of the unknowing joe average public hang on their words I think they need to be more careful about what opinions they give and how they give them.

I would never suggest correction based cures for inexperienced dog handlers, so in saying that I agree with the post on behalf of RSPCA (even though I hate to agree with them on so many other levels).

Not forgetting Pax, that the RSPCA hold to their opinion REGARDLESS of whether the people they are communicating with have little knowledge and experience in dog training or ALOT of knowledge and experience.

I am cautious about what I will advise/explain etc. etc. to people depending on their relationship with their dog and their ability to understand it. As things progress with their dog training, I'm inclined to 'feed a bit more line' so to speak. However, the RSPCA hold their line hard, no matter what. And they work to force their hard line blanket opinion on everyone by pushing through Government legislations.

PS. I'm glad you posted Pax. Always good to hear other views.

ETA: Was preparing post and in the meantime Cosmolo posted.

Edited December 18, 2007 by Erny

[jesomil]

Great post Cosmolo.

Anita, I have a question regarding your breed specific comments regarding sighthounds. While I completely agree about needing to use motivating and positive methods with basic training with breeds such as Salukis and Afghans, I have never seen one with very high levels of training that has not had corrections. By corrections, I dont mean anything severe, but corrections none the less implemented only at higher levels of training.

Eg, how many of these dogs using purely positive methods have 95% recall when under heavy distraction? Or is able to perform obedience routines with a great deal of accuracy?

Hope Im not too OT, I just had to comment on the breed specific bits.

[PAX]

For example,

I know of a dog, an 8 month old GR. This dog was jumping on visitors, stealing washing from the bedrooms and crying at the door when it wanted company, it was rewarded for this as a puppy as the owners thought it was cute.

They got in the (what they thought were pros) franchise Dog trainers. They were advised to alpha roll the dog when it jumped up, throw chains at it when it cried at the door and I can't remember what was suggested for the stealing.

After one month this rambunctious puppy was fright biting when the owners went to pat it, and basically was a nervous wreck.

Another trainer was called in and taught the dog to go to a crate and pull the door shut when the door bell rang, it was heavily rewarded with a pigs ear or what ever when the door knocked.

It was taught when it wanted to come in and go out to ring a bell, when this wasn't wanted the bell was removed. The washing was kept secure and the dog was taught to retrieve via the two toy game.

All this was achieved in under a month.

I am just saying there are plenty of crap trainers out there and I believe you can't do much damage with +R methods. How do pet people know which method is right for the dog and them, and what methods will help or hinder the relationship that they have or want with the dog.

[PAX]

Great post Cosmolo.

Anita, I have a question regarding your breed specific comments regarding sighthounds. While I completely agree about needing to use motivating and positive methods with basic training with breeds such as Salukis and Afghans, I have never seen one with very high levels of training that has not had corrections. By corrections, I dont mean anything severe, but corrections none the less implemented only at higher levels of training.

Eg, how many of these dogs using purely positive methods have 95% recall when under heavy distraction? Or is able to perform obedience routines with a great deal of accuracy?

Hope Im not too OT, I just had to comment on the breed specific bits.

Jesomil, you should watch the Patti Ruzzo DVD's, you will see a whippet trained without correction. Patti got the dog to prove the point that it can be done.

[jesomil]

Jesomil, you should watch the Patti Ruzzo DVD's, you will see a whippet trained without correction. Patti got the dog to prove the point that it can be done.

The thing is, I have seen a few whippets trained to high levels with no corrections but I havent seen saluki's or Afghans get to the same levels with no corrections. I have seen them trained positively and get a certain level of training but never high.

But of course, every method is as good as the trainer doing it.

[PAX]

But of course, every method is as good as the trainer doing it.

Very true.

I also believe building drive from a puppy is the most valuable tool the trainer will have and makes +R training easy/easier.

[Cosmolo]

I think everyone agrees that you can wreck a dog with poorly applied corrections. But i also see many dogs whose owners went down a positive only path and this caused huge problems- one example that springs to mind was a cattle dog that they had used 'positive' training with to try and extinguish his mouthing as a puppy. By the time i saw him, the mouthing was now serious biting and i honestly believe more could have been done to prevent the problems with this dog if different techniques (including appropriate corrections) had been applied from the beginning.

Building drive is great- but there are some people who i don't encourage to build drive because they are likely to build the drive, never use it and this can create unnecesary problems. I also see alot of adult dogs who don't care much for food, pats or toys to begin with. Applying appropriate corrections to supress other behaviours (that the dog normally finds more rewarding than the motivation the owner is offering) can then increase the value of what the owner has so for me, its never 'correction only' training. I think this is a misconception that some people have that if you use corrections you don't/ can't/ won't use positive reinforcement.

I have heard RSPCA before make the comparison between training dogs and dolphins. Aside from the obvious difference in terms of the animals themselves, the dolphins environment is also completely controlled and they don't have access to fish at any other time! Dogs have access to food, affection and toys much of the time- and are required to work within the constraints of society and behave a certain way- something a dolphin would never be expected to do.

[SkySoaringMagpie]

Anita, I have a question regarding your breed specific comments regarding sighthounds. While I completely agree about needing to use motivating and positive methods with basic training with breeds such as Salukis and Afghans, I have never seen one with very high levels of training that has not had corrections. By corrections, I dont mean anything severe, but corrections none the less implemented only at higher levels of training.

Eg, how many of these dogs using purely positive methods have 95% recall when under heavy distraction? Or is able to perform obedience routines with a great deal of accuracy?

I don't think I know anyone who has an Afghan or Saluki that will recall under heavy distraction (heavy distraction being a roo or bunny on the run). Their breeding runs directly counter to it. What kind of correction are you suggesting would work in this situation?

One Saluki trainer I have spoken to has said that if you want bomb proof sighthound recall it is necessary to ensure from an early age that they never get the opportunity not to recall - to me that's about controlling the dog and the environment, not correction. I have seen a BC that I know has been purely positive trained called off a mob of running roos very accurately, so that also says to me that a bombproof recall can be trained w/out correction in the right breed.

Personally, I don't try and recall my guys in high distraction unfenced environments - so they are never off lead in those environments. Not all solutions are training solutions. At the end of a lure coursing run, we combine the basic training with careful positioning on our part, and we don't usually have a problem. And in ordinary low to moderate distraction environments, my guys recall fine and that recall was built on reward and release (not luring).

As to obedience routines, my other half has two CCD passes on the Afghan. I would not claim that the Afghan will ever reach the levels of accuracy that a High in Trial UD Border Collie can achieve, but the CCD passes were not achieved with correction. It is true that we had to remove some R+ that was counterproductive but that wasn't food. Believe it or not, we had to learn to poker face and ignore him when he did the Afghan clown thing, instead of reinforcing him by laughing (even tho' it was very funny, the bastard). That is what I mean when I talk about real accuracy and observation - knowing what you are reinforcing is very important and we are still learning.

There are show people with these breeds who correct and who (usually) get the in ring performance they want. However, there is no-one that I have observed who uses correction who always gets the compliance they want in the ring, including the big winners. So I am still not persuaded that correction breeds accuracy in these breeds - especially as in the obedience ring you can no more correct than you can reward, and sighthounds are nothing if not ring-wise. They do better when they are outsmarted than when they are corrected in my experience.

Edited to add: One thing that has been missing from the debate so far is discussion about what it means for a technique to "work". For some that is about the dog doing what they want. For others, it is about maintaining a certain kind of relationship with the dog. For some it is a mix of both, but in varying quantities/qualities. My definition of a good outcome and someone else's definition of a good outcome will not always be the same.

Edited December 18, 2007 by anita

[Rom]

For example,

I know of a dog, an 8 month old GR. This dog was jumping on visitors, stealing washing from the bedrooms and crying at the door when it wanted company, it was rewarded for this as a puppy as the owners thought it was cute.

Thats an example of how +R can go terribly wrong.

They got in the (what they thought were pros) franchise Dog trainers. They were advised to alpha roll the dog when it jumped up, throw chains at it when it cried at the door and I can't remember what was suggested for the stealing.

After one month this rambunctious puppy was fright biting when the owners went to pat it, and basically was a nervous wreck.

.....and an example of how +P can go terribly wrong.

Like Mark said, +R doesn't work for all dogs. The RSPCA disputes this and I guess that they are referring to instances such as the above to support their claim. The thing that bugs me is that they fail to recognise that +P can be added to training without making a mess of a dog in the same way that +R can be used in training without making a mess of a dog and their stance hinders the education of using +P effectively. They don't even recognise that the training tools they promote can and do apply +P stimulus.

[jesomil]

Hey Anita, well done for your OH getting the 2 passes. Thats really great .

I wont say anymore about salukis/afghans and corrections as i am taking it way off topic. But several methods can work successfully on these breeds depending purley on the skill of the trainer. With a skillful trainer, I see nothing wrong with using corrections.

[Rom]

Another thought on this topic and a scenario that I think that most can relate to......just to clarify my own thoughts on the punishment issue and hopefully learn something. The scenario I'm using is one that I see a lot of at the local training club.

Handler 1 has a dog who he/she believes will not handle punishment well and they opt for reward based training alone or they are at a club that only promotes reward based training. In the initial training the dog is full of beans, acting the clown and is way over stimulated by the environment. During the course of this, the dog frequently hits the end of the lead at force.

Handler 2 has a dog of similar temperament but is open to the use of punishers. The punisher that he uses is of similar intensity to the stimulus that the dog in the above felt when it hit the end of the lead.

Why is it that a dog can deliver that level of correction to itself and get over it...sometimes to the point that it continues the behaviour and yet if the handler delivers that level of correction, many see and I've seen it myself, that the dog shuts down?

Its a case in point with my own dog and the truth is that both of the above examples could have been us at some point in time or another.

But the thing is, I can deliver a correction to my dog now if it is needed and she doesn't shut down and through study and practice, I can take a totally strange dog and put a correction on them and not shut them down, ....the major thing that has changed, I believe is my knowledge about punishers and how and when to apply them while also working to protect the dogs anticipation of reinforcement.....and that I think is the difference with the first dog....he anticipated some reward from the environment that was not perhaps the reward that the owner was planning on. So, he had a different perception of the level of correction he put on himself and some dogs think that the level of correction that they put on themselves is a worthwhile inconvenience to be tolerated in the persuit of percieved reward.....but, if you show them how to avoid the correction and still get a reward, they do change their behaviour.

Another thing is that in the early days, I was frightened of putting a correction on my dog and tried to avoid it....if a correction was delivered at all it came when I felt so pushed that I was angry and frustrated....and I did shut my dog down, then I would go through an attack of the guilts because of it. I believe now that the corrections were shutting my dog down not just because she got a correction, but because there was an extreme pattern of inconsistency.....she could never predict what actions would attract a correction so she couldn't learn to avoid those actions.

The other important thing that I believe is that I was delivering the correction in a way that the dog could pair it with me and my mood, rather than its own actions and behaviours because of shyte timing and my emotional, physical and vocal involvement in the correction because I was pushed to the point where I felt angry and frustrated. Since I've overcome my fear of applying a correction I've learned that if I prevent the dog from pairing the correction with both me and the training tools and if I move the dog quickly past the correction into a reinforcing situation, then the correction does not shut the dog down.

Was I so different to the average dog owner? Is average human nature that different to mine? Is it possible that the above pattern is being repeated out there between thousands of dogs and their owners who are persisting with reward based training alone because its politically correct and there are large animal welfare organisations promoting it as the 'only' humane way to train a dog? Is the real evil here not that punishments are used, but that there is a lack of education about how to use them because of policies of welfare organisations.

The dog in the first example is still getting the same intensity of correction and is at the same risk of injury. How is it that that is acceptable by welfare organisations, but using that stimulus in a manner that will actually help the dog learn is not? When the policies alone do not prevent or protect the dog from delivering that level of correction to itself?

Edited December 19, 2007 by Rom

[ShellyBeggs]

Great post Rom....

I also wonder why some people, and especially the RSPCA see things as either black or white; all or nothing.

It doesn't have to be all positive and it doesn't have to be all corrections/punishment.

I think there can be balance of positive&correction for the training of each dog and that this ratio can vary between individual dogs.

You can most certainly correct a dog without causing any physical or psycological damage, just like you can a toddler.

I mean if kids were dogs, the RSPCA would have that Super Nanny on TV locked away.....because even time out is a corrective punishment.

[leopuppy04]

I also wonder why some people, and especially the RSPCA see things as either black or white; all or nothing.

It doesn't have to be all positive and it doesn't have to be all corrections/punishment.

I think there can be balance of positive&correction for the training of each dog and that this ratio can vary between individual dogs.

You can most certainly correct a dog without causing any physical or psycological damage, just like you can a toddler.

I mean if kids were dogs, the RSPCA would have that Super Nanny on TV locked away.....because even time out is a corrective punishment.

Just a few things - The post written by the RSPCA did say that they use punishment in one form or another (including time outs). Any form or training/ behaviour modification without punishment of some form will get you nowhere (as all of us are aware).

As the post also said "being positive is not being permissive" - a positive trainer is not stepping on egg shells around their dog too afraid to do/ say something wrong. A dog, regardless of training method still needs to know what the guidelines are and what is expected of them.

Just like with anything - I do get the feeling sometimes that +ve training is often misunderstood as an 'airy fairy' training where the dog is never told what is wrong........

That's all I'm gonna say

Edited December 19, 2007 by leopuppy04

[Erny]

Just a few things - The post written by the RSPCA did say that they use punishment in one form or another (including time outs). Any form or training/ behaviour modification without punishment of some form will get you nowhere (as all of us are aware).

Yes - although they have limited "punishment" to :

RSPCA : ... the use of passive punishment (e.g. ignoring unwanted behaviours, removing attention, time outs etc).

my highlight

For some dogs and in some situations, "passive" punishment simply doesn't cut it and the dog can care less.

IMO the pendulum needs to swing in the centre - however RSPCA's push is to swing it a good way to the "left" (so to speak) and if that doesn't work or if problems associated with it being ineffective, pts.

It is clear by their letter to Mark that the RSPCA denies that anything beyond what they have suggested above as "passive" punishment is acceptable - regardless of whether it is well applied by a person of experience and knowledge or not and regardless of liklihood of success.

Edited December 19, 2007 by Erny

[Rusky]

I am not a lover of the RSPCA but I agree absolutely with the response from them, well apart from the wolf bit... If Mark wants to do a bit more wolf research he can find a whole lifestyle with dogs based on wolf pack behaviour on Jan Fennels website.

I also think and have for many years that a treat delivered with bad timing is much better option for the dog than badly and incorrectly delivered positive punishment.

I live in WA all affiliated all breed clubs now use positive reinforcement training. Some breed clubs are still using positive punishment but the numbers decrease constantly. I am happy that the RSPCA has finally embraced positive training.

[Kelpie-i]

Disclaimer:

Firstly, my comment below is made on a curious note only and not to flame what anyone beleives and/or has written on this thread. Secondly, as most of you already know that I consider myself a 'balanced' trainer who uses whatever is needed to obtain the required results. So here is my comment/question.....

not all dogs are motivated or driven by Food. Tony

I hear the above comment made many times by various people. My question is this.....why would we think that a dog is not motivated by food, after all it requires food to survive, right? No eat, no live! I think the statement should be "some dogs are not motivated by food that is not placed in their food bowl".

Getting back to the thread subject, the RSPCA has not met my old dog "Blue", probably one of the meanest and nastiest working kelpies to ever set foot on this earth . No amount of withholding food, love, attention or whatever would make that bastard listen.

Edited December 19, 2007 by Kelpie-i

[Guest Tess32]

I am not a lover of the RSPCA but I agree absolutely with the response from them, well apart from the wolf bit... If Mark wants to do a bit more wolf research he can find a whole lifestyle with dogs based on wolf pack behaviour on Jan Fennels website.

I also think and have for many years that a treat delivered with bad timing is much better option for the dog than badly and incorrectly delivered positive punishment.

I live in WA all affiliated all breed clubs now use positive reinforcement training. Some breed clubs are still using positive punishment but the numbers decrease constantly. I am happy that the RSPCA has finally embraced positive training.

She's the last person I'd call an expert on wolf behaviour. She thought it was some amazing epiphany to realise that dogs don't get as excited when you ignore them on your arrival home

Personally I think it all depends on what part of training and behaviour someone is talking about. I believe some trainers DO tend to put everything down to pack behaviour, when it's clear that the exercise wasn't even taught properly to begin with. I don't need to know much about dog behaviour to teach a sit using operant conditioning, but I DO need to know dog behaviour in order to live with my dog harmoniously.

[Tonymc]

Kelpie-I, your spot on!!!!Thats what I meant but my statement was a bit lacking.You have put it much clearer. Tony

[Kelpie-i]

LOL, no sweat Tonymc. It's just one of those statements that really urks me and I always feel I need to say something about it.....perhaps I should just crawl back in to the Kelpie cave and shut up!!

[PAX]

I think we (or I) are also getting caught up in 'what is a correction'.

For me, I think of a correction in the worst sense of the word which I know is wrong but we really need to say what we think a correction is, as it will vary greatly from person to person.

I do give my own dogs corrections but it might be a subtle 'ah' or it could be a time out. I think my dogs are beautiful happy, well behaved members of my family and I do not wish to hurt them as I know they always try to do the best for me.

So this is really open to personel interpretation of what a correction is.

How can you correct a behavoiur that hasn't been learnt?

I make sure that my dogs truely know a behaviour before I would ever correct it. I know my dogs always want to choose the correct behavoiur as they are so heavily rewarded for that choice. If they did choose the wrong behavoiur my subtle disapproval would be enough of a correction to punish them.

That is where my input to this discussion is, What is a correction?