Topic For Discussion - Keep It Nice, Folks ;)

[corvus]

Lets assume that I walk Kivi instead of Corvus and Kive notices a thread...according to you Kivi will feel safe because he knows 'that he can control his fear' in heel...so he will just choose heeling to feel safe.

Kivi would probably struggle to make the generalisation. He does when my partner walks him. He responds to the heel cue, but not as well. The behaviour was trained with me as the handler all the time, so he needs more practice to generalise to other handlers. It's a different behaviour with someone new.

My other dog, Erik, is a different story. His preference is to direct his safety behaviours at me, but he's crazy good at generalisation and has in the past directed them at other people if I'm not close by. He has also run about 20m to find me so he can start some kind of safety behaviour. Safety behaviours are good if they bring the dog close to you, so it's hard to tell what role closeness plays.

[corvus]

Corvus taught Kivi a behaviour that she initiated when his level of perceived threat was low and demonstrated that if he kept up that behaviour throughout the time the threatening stimulus was present that he would remain safe. Kivi therefore learnt that if HE initiated that same behaviour he would remain safe in the presence of a threatening stimulus. So Kivi now makes the choice to initiate that behaviour when he feels a level of threat, because experience has shown him that he then remains safe.

I don't think Corvus was going to abandon Kivi if he didn't walk in heel position, she would let him close to her no matter what if that was all it took to make him feel safe and calm. But instead of relying on Kivi feeling scared then scrambling to get close to her however he could, she taught him an alternate specific behaviour he could choose to initiate when he felt he needed to.

I've done the same with Riley, his is jumping into my arms, and as I mentioned in a previous post, Saxon has taught himself his own, to put himself in between my feet.

Exactly! :)

And, as mentioned, Kivi wasn't much interested in coming close. I would walk him past while holding him tight against me, so he got that closeness anyway, but he wasn't in a state to particularly notice.

The eye opening moment was when he accidentally triggered another dog at the river one day and the dog forced him to the ground and pinned him there. He yelped and struggled, and when the dog let him up, all I said was "Kivi..." and he trotted into heel and stayed there until I was able to get us clear of the other dog. It was a very different scenario to the dog-behind-a-fence scenario, but I guess he perceived it as the same. It's quite cool when you have behaviours triggered by how the dog perceives the environment. You learn a lot about how they see the world and what happens to them. The only downside is he's stopped helping Erik when Erik is having a problem with another dog. He used to intervene on Erik's behalf, but now he falls into a heel instead and it's on Erik to get himself to safety between my feet.

[Willem]

FWIW, my Kivi learned his safety behaviour of a heel when HIGHLY aroused. Exclusively. Not remotely ideal, but it worked, and only took a few weeks. He would get hysterical when certain dogs behind fences barked at him. I just kept cueing a heel until he was able to do it. He evidently associated the heel with his decreasing arousal and increasing sense of safety, and started doing it as soon as he detected the dog rather than after he'd gone mental and I'd frog-marched him past. Arousal aids learning, but you have to pick the right consequence sometimes.

Right. Not real sure how it's possible for an on leash dog to only achieve closeness via a heel.

Kivi is a super pro-social dog. I suspect he was so upset because the dogs were out of sight and inaccessible. His usual means of dealing with scary dogs (engage with them in a friendly manner) was not available to him. He was reacting by trying to rush to the fence and sniff the dog or see it. Lots of barking and bouncing. If he wanted to be close to me I wouldn't have had a problem. All I wanted was for him to stay close to me. Instead, he had no clue I was even there. He couldn't respond to the heel cue until he was calm enough to process it and act on it - i.e. once he was already starting to calm down after passing the dog.

so he just had a bad recall?

[Willem]

Lets assume that I walk Kivi instead of Corvus and Kive notices a thread...according to you Kivi will feel safe because he knows 'that he can control his fear' in heel...so he will just choose heeling to feel safe.

Kivi would probably struggle to make the generalisation. He does when my partner walks him. He responds to the heel cue, but not as well. The behaviour was trained with me as the handler all the time, so he needs more practice to generalise to other handlers. It's a different behaviour with someone new.

My other dog, Erik, is a different story. His preference is to direct his safety behaviours at me, but he's crazy good at generalisation and has in the past directed them at other people if I'm not close by. He has also run about 20m to find me so he can start some kind of safety behaviour. Safety behaviours are good if they bring the dog close to you, so it's hard to tell what role closeness plays.

...not if he 'evidently associated the heel with his decreasing arousal and increasing sense of safety'...

[Willem]

Corvus taught Kivi a behaviour that she initiated when his level of perceived threat was low and demonstrated that if he kept up that behaviour throughout the time the threatening stimulus was present that he would remain safe. Kivi therefore learnt that if HE initiated that same behaviour he would remain safe in the presence of a threatening stimulus. So Kivi now makes the choice to initiate that behaviour when he feels a level of threat, because experience has shown him that he then remains safe.

I don't think Corvus was going to abandon Kivi if he didn't walk in heel position, she would let him close to her no matter what if that was all it took to make him feel safe and calm. But instead of relying on Kivi feeling scared then scrambling to get close to her however he could, she taught him an alternate specific behaviour he could choose to initiate when he felt he needed to.

I've done the same with Riley, his is jumping into my arms, and as I mentioned in a previous post, Saxon has taught himself his own, to put himself in between my feet.

Exactly! :)

And, as mentioned, Kivi wasn't much interested in coming close. I would walk him past while holding him tight against me, so he got that closeness anyway, but he wasn't in a state to particularly notice.

The eye opening moment was when he accidentally triggered another dog at the river one day and the dog forced him to the ground and pinned him there. He yelped and struggled, and when the dog let him up, all I said was "Kivi..." and he trotted into heel and stayed there until I was able to get us clear of the other dog. It was a very different scenario to the dog-behind-a-fence scenario, but I guess he perceived it as the same. It's quite cool when you have behaviours triggered by how the dog perceives the environment. You learn a lot about how they see the world and what happens to them. The only downside is he's stopped helping Erik when Erik is having a problem with another dog. He used to intervene on Erik's behalf, but now he falls into a heel instead and it's on Erik to get himself to safety between my feet.

...as a response of the positive punishment by the other dog he wanted to be close to you and of course followed your recall...just proofs the effectiveness of positive punishment.

[Simply Grand]

Now I think you're just trolling Willem.

You are under no obligation to use the methods we are talking about if you don't understand them. As long as you are happy with your results with your dog of course you don't have to do what anyone else does.

Edited January 6, 2016 by Simply Grand

[Willem]

Now I think you're just trolling Willem.

You are under no obligation to use the methods we are talking about if you don't understand them. As long as you are happy with your results with your dog of course you don't have to do what anyone else does.

I'm not trolling and I might not have studied dog behaviour for years like you guys, but I can recognize that Corvus description what was taught to the dog is somehow contradictive....and the incident with the other dog was just that: positive punishment by the other dog, which resulted in a very obedient Kivi.

[Willem]

... But she only chases the ball with enthusiasm about three times and then she loses interest. Fetch takes a lot of work too.

...same here, can't say that she gets overexcited when we play ball or fetch...but a rag on a rope and she goes nuts. Maybe the flapping rag reminds her of a bird. Rag, rope and plastic pole - running the rope through the pole allows to accelerate the rag by pulling (a little bit like fly fishing)...cheap as chips and easy to replace if it breaks and it allows to train also the drops and sits in a controlled environment while she is high aroused...

...was contemplating about why she was a 'bad' fetcher...well, I have to admit I never really taught her, when we played fetch everything happend more or less by accident. So today I took a stick and started the training. First, na, she wasn't interested in the stick at all. I put a rope on the stick so I could pull it and voila, different story, at least she took it in her mouth (reward)...repeated it approx. 5 times, then I removed the rope...moved the stick with my hand...she chased it and took it in her mouth (reward)...repeated it approx. 5 times. Dropped the stick and kicked it with my feet...she went for it (reward)...kicked it further away and run away when she had it in her mouth so she followed me with the stick (reward)...after a few repetition she fetched it pretty good; next step was to throw something else...same procedure, lots of rewards ...same result...tomorrow again.

She doesn't get high aroused as it is the case while playing Leave-it-Take-it with the flirt/flick pole...but she knows now it earns her treats :) ..

Edited January 6, 2016 by Willem

[corvus]

...not if he 'evidently associated the heel with his decreasing arousal and increasing sense of safety'...

All right, I'll humour you.

"The heel" was specifically "heel with corvus". "Heel with whoever has my leash" is a DIFFERENT BEHAVIOUR that has NOT been negatively reinforced in THIS SCENARIO OR ANY LIKE IT. When he's with the OH, the cue is present, but the behaviour is not possible. So he gets a bit confused trying to figure out how he should handle this. You know, because he's a dog. They are commonly crap at that kind of generalisation.

Positive punishment results in a DECREASE IN BEHAVIOURAL FREQUENCY, DURATION, OR INTENSITY. If Kivi is pinned by a dog at the river, what behaviour do we see decreased? If we see an increase in a behaviour, what does that tell us about the salient consequence? If you are arguing that we see an increase in heeling because the dog punished Kivi, then the mechanism can only be negative reinforcement. Kivi escapes the aversive dog by heeling. That is exactly what I'm arguing happened. The question is, is the scary dog a consequence or an antecedent? I would usually argue both, but in this case, the consequence is unimportant because Kivi didn't see the dog change its behaviour, so he didn't know what caused the tackle and pin. The aggressive behaviour was therefore uninformative for him and it is doubtful he learned anything from it. He knew how best to respond, though. Seek safety. Not a secure base (attachment figure), although that is a strategy he would often adopt, but a very specific safe behaviour.

Kivi is a big cry baby and certainly comes looking for cuddles when he gets upset. Often. That is not heeling. There are a LOT of ways Kivi gets close to us, and it's not hard to tell the difference between "I'm scared, hold me" and "Eeek, scary dog. Oh! I can fix this!" It's harder to tell the difference between "I would like treats, now" and "That dog is kinda stressing me out", but the latter has a different quality to it if you watch closely. He comes faster and sticks tighter and he's not as relaxed. A safety behaviour doesn't cure a dog of being scared. It just gives them a way to handle it (calmly).

[Mrs Rusty Bucket]

Anyone correct me if I'm wrong, but IME truly confident dogs just don't ever do the crawling approach, roll on back, "grovelling" thing.

You'd think so. But the rest of evil hound's behaviour is not consistent with this idea of an insecure dog. She is insecure - sometimes. And other times she's super confident. Sometimes she even breaks up play or the prelim to a fight between other dogs by running between them. I don't know why humans don't try this more often, just blast between two dogs eyeing each other off...

Ok I know (bite risk). She always gets away with it.

She will cross two ovals to do the grovelly greeting with a new dog or human. And they say "ooh look she's so timid and submissive". I say - if she was "timid" why did she approach in the first place. And submissive? look at them now and she's herding the deerhound or ridgeback any way she wants it to go.

I think most of the time it's her way of making a totally non-threatening safe greeting for the other dog. She gets greetings and play out of dogs that otherwise refuse to make friends by being completely non-threatening. Works well on fear aggressive border collies and other herding dogs but she doesn't push it. If they don't come in for the sniff - she doesn't chase them.

But some dogs - get it all wrong and she leaps up and scolds the hell out of them. Sometimes they've just been clutzes and stepped on her, and some do that lean over and bully thing and as best I can tell - she says "no way". I don't know how dogs choose what they're doing sometimes. Cos she's occasionally scolded some dog that has come barrelling up to her and rolled over. That dog has obviously "done it wrong" in her opinion. And I can't tell - apart from it barged too close before rolling over ie no slow down, look away lip lick grovel crawl first.

And she rolls over for certain humans. Usually ones she thinks will give her a belly rub. She's really good at picking the right humans. And they say - ooh she's so submissive and I say - no - you're special, she doesn't do that for everybody.

[Willem]

.....Positive punishment results in a DECREASE IN BEHAVIOURAL FREQUENCY, DURATION, OR INTENSITY. If Kivi is pinned by a dog at the river, what behaviour do we see decreased? If we see an increase in a behaviour, what does that tell us about the salient consequence? If you are arguing that we see an increase in heeling because the dog punished Kivi, then the mechanism can only be negative reinforcement. Kivi escapes the aversive dog by heeling. That is exactly what I'm arguing happened. The question is, is the scary dog a consequence or an antecedent? I would usually argue both, but in this case, the consequence is unimportant because Kivi didn't see the dog change its behaviour, so he didn't know what caused the tackle and pin. The aggressive behaviour was therefore uninformative for him and it is doubtful he learned anything from it. He knew how best to respond, though. Seek safety. Not a secure base (attachment figure), although that is a strategy he would often adopt, but a very specific safe behaviour.

Kivi is a big cry baby and certainly comes looking for cuddles when he gets upset. Often. That is not heeling. There are a LOT of ways Kivi gets close to us, and it's not hard to tell the difference between "I'm scared, hold me" and "Eeek, scary dog. Oh! I can fix this!" It's harder to tell the difference between "I would like treats, now" and "That dog is kinda stressing me out", but the latter has a different quality to it if you watch closely. He comes faster and sticks tighter and he's not as relaxed. A safety behaviour doesn't cure a dog of being scared. It just gives them a way to handle it (calmly).

you mentioned before that ...'All I wanted was for him to stay close to me. Instead, he had no clue I was even there. He couldn't respond to the heel cue until he was calm enough to process it and act on it - i.e. once he was already starting to calm down after passing the dog'...I think he still had a clue that you were here, he just ignored you. When the other dog pinned him down for sure it was something that didn't make him calmer, still, when the other dog let him go and you said 'Kivi' he came straight away...suddenly he could hear you very well. So the 'decreased unwanted behaviour' was 'ignoring you' or 'selektive hearing' :D. In this moment, the unintended positive punishment worked in your favor and the result was that Kivi was suddenly very obedient.

Look, I see this as a discussion, I might be right or wrong, and as this is not a lab environment where everything can be repeated and analysed, there will be always different views...and lets not forget that we are all 'conditioned' and see what we want to see hence we all fall into this trap of subjective perception all the time.

[corvus]

you mentioned before that ...'All I wanted was for him to stay close to me. Instead, he had no clue I was even there. He couldn't respond to the heel cue until he was calm enough to process it and act on it - i.e. once he was already starting to calm down after passing the dog'...I think he still had a clue that you were here, he just ignored you. When the other dog pinned him down for sure it was something that didn't make him calmer, still, when the other dog let him go and you said 'Kivi' he came straight away...suddenly he could hear you very well. So the 'decreased unwanted behaviour' was 'ignoring you' or 'selektive hearing' :D. In this moment, the unintended positive punishment worked in your favor and the result was that Kivi was suddenly very obedient.

You know, this idea that dogs are always aware of what their owners are doing and asking of them and can always choose how to behave in response is so problematic that I actually wrote a paper about it and got it published in a journal. My volatile dogs classes pretty much start with an explanation for why people's dogs don't respond to them very well when they are aroused, because this is such a prevalent and damaging belief that the dog should listen and be obedient regardless of what is happening to them. The dog CAN'T respond to you. Even if they can hear you, you are asking them to do something that runs counter to their needs and goals. It's VERY hard for them, like asking you to sit down and work on a maths problem when you are pacing around fretting about a loved one in hospital. You're not going to do it. Who cares about maths right now? And you wouldn't be able to concentrate anyway! If the waiting room then goes dark and a fire breaks out and I'm yelling "I said sit down here and do this maths problem!", you probably won't even hear me. Blaming your inability to do a maths problem when you're sick with worry or outright panicked on your selective hearing is unfair and inaccurate. The very fact that Kivi could respond to the heel cue once his arousal had decreased is a pretty good testament to this phenomenon as a likely explanation.

There's a saying that goes, if a dead dog can do it, it's not a behaviour. Can a dead dog ignore you? Hell yes. "Ignoring corvus" is not a behaviour, therefore, it cannot be decreased. Particularly in the case at the river where Kivi was not actually given a cue from corvus at all. He wasn't "suddenly obedient", because he hadn't been given any specific instructions to follow obediently in the first place. Saying his name is not a recall.

Look, I see this as a discussion, I might be right or wrong, and as this is not a lab environment where everything can be repeated and analysed, there will be always different views...and lets not forget that we are all 'conditioned' and see what we want to see hence we all fall into this trap of subjective perception all the time.

I can help you, here. You're wrong. I might not be right, but you are definitely wrong. Hope that clears a few things up that a collection of evidence apparently did not. Not going to restate the evidence, because you ignored most of it the first time.

Discussions that depend on an alternative interpretation of a dog's behaviour by someone that was not there and does not know the dog and can't produce a cohesive argument that is consistent with theory or even logical pretty quickly become a waste of time. I have a lot of more important and constructive things to do. The paper has a ridiculous title I can never remember, but if you Google "conceptualising operant conditioning arousal starling", you should get it as the first hit. It's free and has funky interactive graphs.

[Willem]

wrt 'the dog CAN'T respond to it': just not true, a dog can respond if it's high aroused and trained to respond at this arousal level....it is just much harder to train. High aroused dogs respond to clicker if conditioned....or my dog to 'Leave-it' when training with the flirt/flick pole. The problem is to maintain high arousal in a controlled environment so you can train the dog. Think about Schutzhund training - even Karin Pryor states this ...'The dog must be able to hear and respond to its handler when in full fighting drive'...

[Willem]

[There's a saying that goes, if a dead dog can do it, it's not a behaviour. Can a dead dog ignore you? Hell yes. "Ignoring corvus" is not a behaviour, therefore, it cannot be decreased. Particularly in the case at the river where Kivi was not actually given a cue from corvus at all. He wasn't "suddenly obedient", because he hadn't been given any specific instructions to follow obediently in the first place. Saying his name is not a recall.

...and that is just nonsense, a dead dog can't ignore you. Ignoring means that the dog knows the cue but decide to ignore you and does something else (and that's just undesirable behaviour). You tell us that your dog ran to a fence and barked at another dog...and you didn't make an effort to recall him? ...if you didn't it would be pretty slack, if you did your dog ignored you obviously because he decided that it is more fun to bark at the other dog....not much room for other interpretations. Your statement that your dog couldn't hear you because it was in an aroused state ...well, would be plausible from someone with little experience in dog training...

[JulesP]

[There's a saying that goes, if a dead dog can do it, it's not a behaviour. Can a dead dog ignore you? Hell yes. "Ignoring corvus" is not a behaviour, therefore, it cannot be decreased. Particularly in the case at the river where Kivi was not actually given a cue from corvus at all. He wasn't "suddenly obedient", because he hadn't been given any specific instructions to follow obediently in the first place. Saying his name is not a recall.

...and that is just nonsense, a dead dog can't ignore you. Ignoring means that the dog knows the cue but decide to ignore you and does something else (and that's just undesirable behaviour). You tell us that your dog ran to a fence and barked at another dog...and you didn't make an effort to recall him? ...if you didn't it would be pretty slack, if you did your dog ignored you obviously because he decided that it is more fun to bark at the other dog....not much room for other interpretations. Your statement that your dog couldn't hear you because it was in an aroused state ...well, would be plausible from someone with little experience in dog training...

And that is just plain rude. Keep this up and everyone will have you on ignore.

[Willem]

And that is just plain rude. Keep this up and everyone will have you on ignore.

...I can live with this...

[JulesP]

And that is just plain rude. Keep this up and everyone will have you on ignore.

...I can live with this...

Have fun chatting to yourself then.

[Willem]

I guess there are some merits in the following quote:

...'Too often, however, modem dog owners try to solve behavior problems by analyzing or explaining why the dog is misbehaving. With the exception of genetically induced behavioral anomalies (which are hard to diagnose and pin down) it is rarely productive to ask why a dog does what it does. Identify the behavior (not the cause of the behavior). Then identify something you can reinforce that will replace that behavior-and the stuff you don’t want will go away by itself'.

[Kavik]

Arousal is very important! Dogs certainly can become too aroused to listen and respond to cues. You can train them to work in a highly aroused state, but it does not happen automatically and takes time and effort. Some dogs are naturally more easily overaroused. One of my struggles with Nitro at the moment :laugh:

Dogs also can be too worried/anxious to listen and respond to cues.

[Willem]

Arousal is very important! Dogs certainly can become too aroused to listen and respond to cues. You can train them to work in a highly aroused state, but it does not happen automatically and takes time and effort. Some dogs are naturally more easily overaroused. One of my struggles with Nitro at the moment :laugh:

Dogs also can be too worried/anxious to listen and respond to cues.

I understand that some people prefer that their dog doesn't get overexcited or reach a high aroused state (and they may have good reasons for this), but I don't want a 'robot' dog, so if my dog gets overexcited it is o.k....as long as I still can control it and can break it up if required. Hence the training with the flirt / flick pole...when the rag moves in front of her she trembles ...everything in her is just waiting for the 'Take-it'...and then bang she goes nuts....while she is chasing the rag in a very high aroused state I can train LEAVE-IT (she comes back to me in heel position), DROP...STAND, SIT...and all happens in a controlled environment....that works for my dog, other dogs might find this game boring so or their only play it on the 'usual treat earning scheme' at a lower arousal level (which doesn't give the same results).