Interruptors

[corvus]

We are getting another puppy in about a month and I've been thinking on how I would like to improve on the way I raised Kivi. One of the things I've been wanting to improve on is my heavy reliance on "ah-ah" as a conditioned punisher, or a negative interruptor, I guess. It's not meant to stop KT from doing something forever, just meant to interrupt him and tell him I don't like that and will come and shoo him away if he keeps doing it. It works, but I don't like being so dependent on it and I was wanting to shift my thinking from what I don't like to what I do like.

I came across a video on Youtube about training a positive interruptor. It turns out I've been doing this with Kivi as well as the negative interruptor, but because I wasn't consciously aware that I was trying to train a positive interruptor it's not nearly as strong as it needs to be. And I've got a few other things that act as positive interruptors as well.

I was thinking that one of the things I liked about a conditioned punisher is that it was good for moments when someone's safety was under threat, but then it has occurred to me that I've just spent the last 12 months training an emergency recall almost soley with awesome food treats and it gets that dog of mine moving faster and more reliably than anything else in the world. He flies to our feet through any obstacle when we use that ER. I've also kind of always thought that it's helpful for a dog to know when you dislike something as well as when you like something, but I'm feeling challenged about that at the moment and am wondering if that's just my human way of looking at things. Maybe a dog doesn't need to know when you don't like something if you have a positive interruptor and have taken the time to train a few alternative behaviours.

Would anyone care to comment on negative and positive interruptors? Is there a reason why a dog should know what you don't like as well as what you do?

[poodlefan]

Would anyone care to comment on negative and positive interruptors? Is there a reason why a dog should know what you don't like as well as what you do?

Absolutely. Because dogs need to know the boundary between acceptable and unacceptable behaviour. Unless a dog experiences a negative response, how will it know that the particular behaviour in undesireable?

As you know, all animals will increase behaviour that is rewarding to them and diminish that which isn't. You want the behaviour you don't like (from mouthing, to jumping to peeing on your leg) to diminish. Deflecting or distracting those behaviours doesn't always work, particularly with dogs of tougher character. Sometimes negative reactions are called for.

[Staranais]

Hooray new puppy! What kind?

Yes, I say yes to your question, wholeheartedly. I can't imagine living with a dog without being able to communicate to her that she's displeasing me. Both for very practical reasons (you can't imagine the number of things my puppy comes up to do with that are really destructive... I need to start a new thread titled "things my malinois has eaten"! ) But also, for me, for philosophical reasons too. Ideally, I want to grow to really communicate with my dog, not just train her. And full communication IMO ultimately requires me being able to communicate what I don't like, as well as what I do like her doing.

However, I don't think a "negative interrupter" necessarily needs to be perceived as particularly negative by the dog. A disrupter, where the dog is distracted, and then is redirected to a more appropriate activity which can be rewarded, works really well for some dogs and some behaviors.

[Aidan]

I've also kind of always thought that it's helpful for a dog to know when you dislike something as well as when you like something, but I'm feeling challenged about that at the moment and am wondering if that's just my human way of looking at things.

Dogs do care what we think, but whether we express approval or disapproval isn't necessary to control a behaviour. Lots of people clearly express disapproval while unwanted behaviours are being reinforced. Similarly, lots of people clearly express approval while never getting more of the behaviour that they are approving of!

You can teach a fluent, reliable "Leave It" using positive reinforcement, negative reinforcement or punishment. Using aversives won't make the behaviour any more reliable than using positive reinforcement, so if you are familiar with the positive reinforcement process and are satisfied that you can teach a fluent, reliable response using positive reinforcement, there is nothing stopping you from getting great results this way.

You might want to look up "Doggy Zen", as it is the foundation for a very reliable "Leave It" behaviour using positive reinforcement. You might recognise it as part of the "Triangle of Temptation" which is so popular here.

[corvus]

Hooray new puppy! What kind?

Yes, I say yes to your question, wholeheartedly. I can't imagine living with a dog without being able to communicate to her that she's displeasing me. Both for very practical reasons (you can't imagine the number of things my puppy comes up to do with that are really destructive... I need to start a new thread titled "things my malinois has eaten"! ) But also, for me, for philosophical reasons too. Ideally, I want to grow to really communicate with my dog, not just train her. And full communication IMO ultimately requires me being able to communicate what I don't like, as well as what I do like her doing.

See, that was my thinking exactly...

Don't take this as being argumentative. It's just the argument that's going through my own head. I think, yes, I think that it's fair to communicate when you don't like something your dog is doing, and I think dogs get that when you go "Aght!" for example, that's just you telling them you don't like that, which they ideally care about because they are social animals and one of their aims in life is to avoid confrontations with other animals they are living with. It's funny how you forget how crazy puppies are. They don't know the rules and have no idea that the noise you just made means "don't". You have to teach them that. My thought is, if you have to teach them "don't", then why not teach them "look at me" instead? Does it really make any difference? Like Aidan said, if the action is rewarding enough, the dog will do it anyway. And I guess then you've just decreased the effectiveness of that interruptor.

I go on to think, what's the difference between setting a boundary with a negative action and setting a boundary with a positive action? If you manage things correctly, perhaps you never need to say "don't"? Instead you can say "look at me, now how about you go do this thing that I've reinforced positively like hell?" What's the difference between that and saying "don't do that, now how about you go do this thing that I've reinforced positively like hell?" Just as an example.

Furthermore, I've never tried living with only positive interruptors and no negative interruptors. It could be awesome.

Okay, I also believed that a dog needed a no response so he knew what not to do and then you'd tell him what to do and you would all be happy. But I'm thinking, what if that's actually not necessarily true? What if a dog doesn't need a no response to know what he shouldn't do? If you reinforce what he should do often enough, and then if he ever goes to do what he shouldn't do you just tell him what he should do and reward, then maybe he just sticks to what he's been rewarded for. Say you were freaking awesome at environmental management and/or pre-emptive training, couldn't you control what pup gets rewarded for to the extent where he only develops good habits and you don't need to tell him you don't like something? Isn't that more or less one of the aims of neutralisation?

I'm just bouncing ideas around. If no one blows them out of the water well enough you know what's going to happen! Someone's gonna get fixated...

However, I don't think a "negative interrupter" necessarily needs to be perceived as particularly negative by the dog. A disrupter, where the dog is distracted, and then is redirected to a more appropriate activity which can be rewarded, works really well for some dogs and some behaviors.

Are you saying it makes no difference if it's a negative or a positive interruptor?

Provided everything goes according to plan, new pup will be a Swedish Vallhund. Apparently we only have hunds now. I DESPERATELY wanted an Akita, but alas, OH pointed out that our car is not very big compared to the size of an adult male Akita. It's probably for the best, as we still have those rabbits to look after. One day when I no longer have pet rabbits I can get a hunting dog. Probably never going to happen, though. If it's not rabbits, it'll be ferrets or wildlife or some other small, vulnerable prey animal. Hopefully the Val will be a little more into fun things like frisbees and flirt poles than Kivi Tarro, though. I think that Lapphund is a little broken.

I will look up Doggy Zen. Ironic seeing as I was talking about Zen with dogs just today. ;) I never really got far with ToT as I was like "eh, that sounds like too much hard work" but having said that, life would have been easier if I'd bothered to train "leave it" before it kinda became necessary. Today KT absent-mindedly spat out a gummi snake he found on the footpath when I told him to leave it, then went "Wait, what am I doing? I was going to eat that", and I had to drag him away from it because I had nothing else to offer and for all that he'll spit out some really tasty things on command, he just picks them up again if you're not on the ball enough to move him away from it before he can think about how good it tasted. Another thing to improve on this time 'round.

[huski]

Okay, I also believed that a dog needed a no response so he knew what not to do and then you'd tell him what to do and you would all be happy. But I'm thinking, what if that's actually not necessarily true? What if a dog doesn't need a no response to know what he shouldn't do? If you reinforce what he should do often enough, and then if he ever goes to do what he shouldn't do you just tell him what he should do and reward, then maybe he just sticks to what he's been rewarded for. Say you were freaking awesome at environmental management and/or pre-emptive training, couldn't you control what pup gets rewarded for to the extent where he only develops good habits and you don't need to tell him you don't like something? Isn't that more or less one of the aims of neutralisation?

I use a 'no' marker just like I use a 'yes' marker, IMO it makes communicating to the dog easier and clearer for them. I don't know why using a no marker has to be looked at as something negative, in the instance of just letting the dog know that 'nope, you didn't get it, try again'. I know at least with Daisy, it stops her getting overly frustrated and helps her understand things in a clearer way. What happens if (like you say above) the dog does something he shouldn't, or offers the wrong behaviour, and you tell him what you want him to do but he ignores you - do you keep repeating the command until he gets it right? How do you proof the right behaviour if you remove any opportunity to communicate to the dog that they've got it wrong? Why is giving them a NRM 'bad'?

ETA:

I go on to think, what's the difference between setting a boundary with a negative action and setting a boundary with a positive action? If you manage things correctly, perhaps you never need to say "don't"? Instead you can say "look at me, now how about you go do this thing that I've reinforced positively like hell?" What's the difference between that and saying "don't do that, now how about you go do this thing that I've reinforced positively like hell?" Just as an example.

I don't think you could control the dog's environment so much that they never develop a behaviour that needs correcting or that they never need to be told they've done the wrong thing. Neutralisation as far as I can tell is not about raising the dog so you never have to tell them they've got it wrong.

For example - look at how we teach loose leash walking. Aren't using the 'change of direction' or 'stop like a tree' techniques both ways of using a form of interruptor or "negative" marker because you're teaching the dog that 'no, pulling will not get you where you want to go'. You would also have to assume that you have a dog who cares if you've reinforced something 'positively like hell' and isn't ever going to challenge you when it comes to what he wants to do vs what you want to do.

ETA2: This reminds me a little bit of the conversation I had with the DELTA instructor who took our puppy school class. She, too, told us not to use interruptors and to never say no or ah ah to our pups. I asked her what I should do when Daisy walked up to the coffee table and stole food right off my plate the night before - she was only a little pup so we didn't have a command like 'look' or 'come' that was solid enough for us to use as a way to stop or distract her - I could hardly expect her to know what all the rules were (i.e. that taking food off plates is not allowed) that early on. Oddly enough she couldn't really give me an answer. IMO it was better to use an interruptor like 'no' or 'ah ah' in that situation to stop her than it was to just let her eat food off my plate in the interest of only using purely positive methods.

Edited July 27, 2009 by huski

[Staranais]

However, I don't think a "negative interrupter" necessarily needs to be perceived as particularly negative by the dog. A disrupter, where the dog is distracted, and then is redirected to a more appropriate activity which can be rewarded, works really well for some dogs and some behaviors.

Are you saying it makes no difference if it's a negative or a positive interruptor?

I'm saying I think you can interrupt a behavior without the dog seeing the interuption as being either negative or positive. If my pup is doing something and I clap my hands, she often reorients to me on my clap, then I can send her off to do something else. It's not negative, it's not positive, it's just a noise to get her attention off what she's doing. So to me, an interupter is different to either punishment or reinforcement. It's in a class of its own.

Doggy zen is great, most of my training is based around doggy zen - the idea that paying attention to me instead of to the reward is actually the fastest route to the reward.

But I also like to be able to communicate to my dog "that's not right, I don't like that". Whether by using a "no reward marker" in training, or by growling "uh-uh" in day to day life. It just feels more honest and communicative and less manipulative and convoluted to directly tell the dog what I don't like, rather than just distracting her and hoping she doesn't try it again. I realise that's me putting my human interpretation on things, and that Fledge may not see the difference at all. However, I'm 50% of the partnership here, so my feeling about training matters as much as hers does.

Swedish Vallhund? Interesting! Would love to see pictures when she/he arrives!

[corvus]

Crap, I had this detailed reply all typed out and then I lost it. So here's the short version.

I use a 'no' marker just like I use a 'yes' marker, IMO it makes communicating to the dog easier and clearer for them. I don't know why using a no marker has to be looked at as something negative, in the instance of just letting the dog know that 'nope, you didn't get it, try again'. I know at least with Daisy, it stops her getting overly frustrated and helps her understand things in a clearer way. What happens if (like you say above) the dog does something he shouldn't, or offers the wrong behaviour, and you tell him what you want him to do but he ignores you - do you keep repeating the command until he gets it right? How do you proof the right behaviour if you remove any opportunity to communicate to the dog that they've got it wrong? Why is giving them a NRM 'bad'?

I didn't say anything about NRMs. I used to use them because I thought Kivi needed the extra info, but have since abandoned them because I don't think they actually do give more info than, say, no reward on its own without a marker. But that's beside the point. NRMs are not the same as negative interruptors. The latter is, by definition, negative.

I am guessing that if pup does something he shouldn't and he ignores my suggestion of something better, I do the same thing I do with Kivi in such a situation. Take away whatever it was he was getting into, or remove myself or him from the situation. Or I say "Guess you don't want a reward" and leave it at that if it's not a serious thing. I only use the negative interruptor for things I can't easily move. It's not meant to stop him from doing that thing forever. It's just meant to tell him that right here and now, I'm not keen on that behaviour.

As far as proofing goes, are you asking how you proof behaviour without corrections, or without negative interruptors? Or something else? I'm not entirely sure what you're asking. I can say that I'm not in the habit of using either in proofing behaviour. If I want it to be rock solid I reinforce the crap out of it with rewards. Doesn't everyone? If I was lazy and didn't build Kivi up to higher distractions and just jumped into it, I have lately been using the negative interruptor. It doesn't work! That interruptor is weaker than the command, so it's not much use in proofing.

For example - look at how we teach loose leash walking. Aren't using the 'change of direction' or 'stop like a tree' techniques both ways of using a form of interruptor or "negative" marker because you're teaching the dog that 'no, pulling will not get you where you want to go'. You would also have to assume that you have a dog who cares if you've reinforced something 'positively like hell' and isn't ever going to challenge you when it comes to what he wants to do vs what you want to do.

I would have thought changing directions and being a tree was R-. Well, the latter at least. I've never tried doing the former. It's not quite the same IMO as a negative interruptor or marker.

All dogs challenge at some point. Even Kivi! He may be more teddy bear than dog, but he's a spitz somewhere deep down and he doesn't generally do things unless they obviously benefit him. It behooves any dog to care if you have reinforced something whether it be positively or negatively. The question is have you reinforced it enough to compete with other neat/bad things? And is the response to that reinforcement conditioned? It's amazing what you can achieve with roast beef where chewy junk food treats failed.

ETA2: This reminds me a little bit of the conversation I had with the DELTA instructor who took our puppy school class. She, too, told us not to use interruptors and to never say no or ah ah to our pups. I asked her what I should do when Daisy walked up to the coffee table and stole food right off my plate the night before - she was only a little pup so we didn't have a command like 'look' or 'come' that was solid enough for us to use as a way to stop or distract her - I could hardly expect her to know what all the rules were (i.e. that taking food off plates is not allowed) that early on. Oddly enough she couldn't really give me an answer. IMO it was better to use an interruptor like 'no' or 'ah ah' in that situation to stop her than it was to just let her eat food off my plate in the interest of only using purely positive methods.

Well, I doubt you want to hear my take on this, but you've brought it up so you're going to get it.

I don't know about Daisy, but Kivi didn't come home knowing what "ah-ah" meant, either. Once or twice I caught him doing something I didn't want him to do and said it loud enough for it to be an interruptor by its sound and volume alone. What I got was a puppy that had no idea what I'd been upset about and simply wouldn't even go in that room for the next few hours until I could coax him back in again. Soft dogs come with difficulties of their own?

I never had to tell Kivi not to steal food from the table as a puppy because I was never silly enough to expect him to know that rule and so he was never in a room with food at nose height. So in all honesty, that is a pretty easily solved problem to me. This is why I don't generally give trainers money anymore! No one is perfect and sometimes you break your own rules and let puppies into places where they are inevitably going to make a big mistake. I wouldn't say anything to a puppy in that situation. I'd just take him out again. Doesn't seem worth it to me to use it as a training or communication exercise. Kivi still gets put in another room when we eat as a general rule, although he now knows what acceptable begging looks like.

My point is that I had to teach Kivi what "ah-ah" meant in order for it to be particularly helpful. Why not teach "look at me" instead? I'm not interested in "purely positive". I'm interested in "why use negatives if you could use positives?".

ETA, nope, apparently that was also the long version.

Edited July 27, 2009 by corvus

[huski]

All dogs challenge at some point. Even Kivi! He may be more teddy bear than dog, but he's a spitz somewhere deep down and he doesn't generally do things unless they obviously benefit him. It behooves any dog to care if you have reinforced something whether it be positively or negatively. The question is have you reinforced it enough to compete with other neat/bad things? And is the response to that reinforcement conditioned? It's amazing what you can achieve with roast beef where chewy junk food treats failed.

And what happens when roast beef doesn't work? I can and have many a time shoved food under Daisy's nose when she's on a scent and she doesn't even notice it. I've used it all - raw meat, cooked steak, chicken, bacon, sausage and even donuts. Sometimes, nothing can compete with the urge to scent. In which case I might use an interruptor not as a punishment, but like Staranais described, as something neither positive nor negative, just something to get her attention.

I don't know about Daisy, but Kivi didn't come home knowing what "ah-ah" meant, either. Once or twice I caught him doing something I didn't want him to do and said it loud enough for it to be an interruptor by its sound and volume alone. What I got was a puppy that had no idea what I'd been upset about and simply wouldn't even go in that room for the next few hours until I could coax him back in again. Soft dogs come with difficulties of their own?

But in the situation I described, she didn't need to know what it meant like she would a command like sit. It was used as a way to interrupt what she was doing so I could stop her doing it. If I'd just let it slide, she would have grabbed the food and gotten a reward for doing so by eating it. Sometimes it's quicker to say 'ah' then it is to grab them if you are a distance away from them.

Unfortunately Daisy was never a soft dog, if she was maybe I would have had to find a different way to deal with her It worked for her though (as much as you might believe it's flawed and/or ineffective).

I never had to tell Kivi not to steal food from the table as a puppy because I was never silly enough to expect him to know that rule and so he was never in a room with food at nose height. So in all honesty, that is a pretty easily solved problem to me. This is why I don't generally give trainers money anymore! No one is perfect and sometimes you break your own rules and let puppies into places where they are inevitably going to make a big mistake. I wouldn't say anything to a puppy in that situation. I'd just take him out again. Doesn't seem worth it to me to use it as a training or communication exercise. Kivi still gets put in another room when we eat as a general rule, although he now knows what acceptable begging looks like.

Of course it can be easily solved in theory, but like you said we all make mistakes - leaving something where you don't think the pup is going to be able to reach it yet they can and do, or not thinking ahead and putting them away when you should have. Different things will work for different dogs. Some dogs need to be told 'no' (or an equivalent). Dogs do it to other dogs, yet it's unacceptable or undesirable for people to consider using a "negative"?

My point is that I had to teach Kivi what "ah-ah" meant in order for it to be particularly helpful. Why not teach "look at me" instead? I'm not interested in "purely positive". I'm interested in "why use negatives if you could use positives?".

What is there for the dog to understand about 'ah ah' if you are using it to simply interrupt their action? I've not ever had a dog look confused when I've given a good 'ah' and I've always found it very effective. I've never had a dog look wounded and sulk off and refuse to come near me afterwards. Why wouldn't I want to use something that's worked well for my dogs?

I guess the difference is that I don't necessarily look at "negatives" as something bad or ineffective or something to use as a last result. Negative or positive, I'm much more interested in using what works for the dog

Edited July 27, 2009 by huski

[poodlefan]

All dogs challenge at some point. Even Kivi! He may be more teddy bear than dog, but he's a spitz somewhere deep down and he doesn't generally do things unless they obviously benefit him. It behooves any dog to care if you have reinforced something whether it be positively or negatively. The question is have you reinforced it enough to compete with other neat/bad things? And is the response to that reinforcement conditioned? It's amazing what you can achieve with roast beef where chewy junk food treats failed. :D

What if the dog cares more about a self rewarding behaviour than any positive reinforcement you could offer?

[Quickasyoucan]

All dogs challenge at some point. Even Kivi! He may be more teddy bear than dog, but he's a spitz somewhere deep down and he doesn't generally do things unless they obviously benefit him. It behooves any dog to care if you have reinforced something whether it be positively or negatively. The question is have you reinforced it enough to compete with other neat/bad things? And is the response to that reinforcement conditioned? It's amazing what you can achieve with roast beef where chewy junk food treats failed. :D

What if the dog cares more about a self rewarding behaviour than any positive reinforcement you could offer?

Exactly, with Jake and his skateboard obssession you could waive a ball in front of his face which I believe is close to his favourite thing in the world and that would not distract him. Sometimes, for example, a certain trigger has become so hard wired into a dogs brain that all the temptations in the world will not distract them.

I think timing is crucial too, if you get in at the right time your interruptor or whatever does not need to be forceful, leave it too late and it really is too late for everything.

Corvus when you say dogs didn't come into the world knowing ah ha I don't think that is necessarily true, true in that they don't understand the word, but the energy and the tone in which you convey it would I believe be understood by the vast majority of dogs, same thing with very young children. It is almost instinctual. You could use another word, but it would still be understood.

[corvus]

Roast beef is just an example. I don't need to tell you that you use whatever reinforcement does the job. For my mum's dog Pyry, hunting is the big reward. If you want to overcome the pull of hunting, you have to use it. For her dog Jill it's fetch. She will do anything for fetch. For Kivi, it would have been other dogs, but fortunately we put the work in with other big rewards to condition the response we wanted before expecting compliance around other dogs. If Kivi ever stopped to think whether he wanted to wrestle with dogs or roast beef, he'd probably pick other dogs, at least from time to time. Fortunately roast beef is also high on the list or we'd have to think of something else.

I haven't said anywhere that I think interruptors are universally flawed or ineffective. I said my negative interruptor for Kivi was weaker than a command I had trained with rewards, even if the command is in its early stages. It's quite effective in low key situations, and that's exactly what I aimed to achieve with it.

I think you're getting tangled up in what I have said versus what you think I'm thinking. This is not a positive reinforcement versus correction debate. This is a "do you need a negative interruptor?" musing.

Yes, you can have an interruptor that is neither particularly negative or positive. I'm not especially interested in those ones right now. "Oi!" works pretty decently, though.

To begin with when I said "ah-ah" to Kivi he ignored me because it meant nothing to him. I had to teach him that it meant I was going to back him off whatever he was into. Like I said before, if I did it loud enough to act as an interruptor on its own, it was too harsh for the little man.

"Negatives" as you say, huski, are by their very definition the opposite to positives. If positives represent things that a dog likes, then negatives must represent things that a dog does not like, right? I'm not foolish enough to think I can get through life with a dog without making them experience negative things, but I would just like to minimise them as much as possible, at least with myself. OH likes to growl at Kivi when he does something that OH doesn't like. That's fine by me. It's his relationship with that dog and he can do whatever he likes with it within reason. But it's kinda obvious that when Kivi gets growled at, he comes and quietly lies down beside me. Next time OH tries to call Kivi to him, Kivi doesn't go. At least, not without some encouragement from me. If OH growled at Kivi all the time, it would become less aversive, but what's the point in that?

I've noticed that dogs do a lot of stuff to each other that isn't all that nice. I've also noticed that the consequences for a dog ignoring a threat from another dog is often physical. I have no problem with anyone who wants to base their relationship on the same things, but personally, I'm not a dog and have no intention of acting like one. I have the blessed gift of being able to outsmart my dog (I hope) and reducing the doing of things to my dog that he doesn't like in the process. If that didn't seem better to me somehow, I wouldn't do it, would I? I want to cut down on the use of negatives in training and I don't understand why anyone would particularly care one way or another. Yes, I do think negatives are "bad" because that is, after all, their inherent nature. Certainly "ah-ah" is not in the same ballpark as a boot to the ribs, but I have the luxury of choosing to do neither if I desire.

To me, it's about the attitude. I don't think saying "ah-ah" to Kivi has somehow irreversibly damaged my relationship with him in the slightest, but I would like to concentrate on finding things to reward rather than things to tell dogs not to do.

[huski]

To me, it's about the attitude. I don't think saying "ah-ah" to Kivi has somehow irreversibly damaged my relationship with him in the slightest, but I would like to concentrate on finding things to reward rather than things to tell dogs not to do.

But if it's ok for dogs to give each others 'warnings' and it can be a big part of how they learn appropriate behaviour, you'd (or at least I would) could assume that negative markers can be a useful tool when it comes to communicating with our dogs.

I don't focus on thinking about it being one way or the other, my attitude is I use what works best for the dog in each situation. What works in one situation may not in another. What works for one dog will not work for another. I've done the growl at Daisy when she's been barking at us when we have food and it worked fantastically as a way to tell her to back off. It wasn't so much the growl itself, it was my tone, and the meaning behind it, she knew it was a warning and she needed to hear it as she can be a very over confident dog and sometimes needs to be told no. If I used 'the growl' or 'ah ah' and my dog wouldn't come when I called them as a result I would think twice about using it again - but my relationship with my dogs is one where a correction here or there is not going to affect their bond with me or their willingness or desire to do what I ask them to do.

I guess I don't see why telling the dog no or telling them not what to do has to be perceived as some kind of weakness. I tell my dogs what to do every day, but sometimes it helps to communicate with them so they know when they've got it wrong too. Why would I want to avoid using what has been a useful and successful tool? At least that's how I see it.

[corvus]

All dogs challenge at some point. Even Kivi! He may be more teddy bear than dog, but he's a spitz somewhere deep down and he doesn't generally do things unless they obviously benefit him. It behooves any dog to care if you have reinforced something whether it be positively or negatively. The question is have you reinforced it enough to compete with other neat/bad things? And is the response to that reinforcement conditioned? It's amazing what you can achieve with roast beef where chewy junk food treats failed. :D

What if the dog cares more about a self rewarding behaviour than any positive reinforcement you could offer?

What if? Is it then going to care more about a punishment I can offer than this self-rewarding behaviour? Is an "ah-ah" seriously going to work where a pile of treats would not? I'm talking non-physical interruptors of the negative and positive kind, here. I don't think either of them are going to work in that situation. Do you? Unless it's conditioned, which can go either way, can't it?

[corvus]

I guess I don't see why telling the dog no or telling them not what to do has to be perceived as some kind of weakness. I tell my dogs what to do every day, but sometimes it helps to communicate with them so they know when they've got it wrong too. Why would I want to avoid using what has been a useful and successful tool? At least that's how I see it.

And I don't see what this has to do with you and what you perceive as a successful tool for your dog, particularly. I'm not asking you to avoid using negative interruptors. I'm asking if you think there's a purpose for a negative interruptor that a positive interruptor can't also achieve. You've more or less answered that, thanks.

Until about a week ago I was also thinking that it was important for dogs to know when their people were unhappy with them. Now I'm wondering if that's actually true. I'm challenging it. Why do they need to know? Do they even understand it? Is it really any different in function to a positive interruptor? Pretend for a moment that you didn't know that "ah ah" was a signal of displeasure. Why does it work? What exactly does it achieve?

As far as warnings go... I'm mentally challenging that one, too. But that's for another thread. Let's just say I've had over a year of doggy warnings and I can't say I think Kivi likes them much at all. There are things he would never do around Penny lest he provoke a warning. I wonder why I would need him to feel that way around me when I could just tell him what I do want him to do and give him a big cuddle when he does it instead.

[huski]

I guess I don't see why telling the dog no or telling them not what to do has to be perceived as some kind of weakness. I tell my dogs what to do every day, but sometimes it helps to communicate with them so they know when they've got it wrong too. Why would I want to avoid using what has been a useful and successful tool? At least that's how I see it.

And I don't see what this has to do with you and what you perceive as a successful tool for your dog, particularly. I'm not asking you to avoid using negative interruptors. I'm asking if you think there's a purpose for a negative interruptor that a positive interruptor can't also achieve. You've more or less answered that, thanks.

I was just giving my opinion like you asked at the start of the thread - of course what I know is successful for my dogs is going to impact on whether or not I would use that particular method and if I thought there was a purpose for it. Just as you've given your opinion on why you don't think they work or work well for your dogs

Until about a week ago I was also thinking that it was important for dogs to know when their people were unhappy with them. Now I'm wondering if that's actually true. I'm challenging it. Why do they need to know? Do they even understand it? Is it really any different in function to a positive interruptor? Pretend for a moment that you didn't know that "ah ah" was a signal of displeasure. Why does it work? What exactly does it achieve?

In regards to interruptors - it works because it gets their attention. You mentioned using oi - I use it quite often it too. I wouldn't use a voice correction if it wasn't effective or didn't achieve what I wanted it to. If it didn't work or wasn't suitable for the dog, I wouldn't use it.

As far as warnings go... I'm mentally challenging that one, too. But that's for another thread. Let's just say I've had over a year of doggy warnings and I can't say I think Kivi likes them much at all. There are things he would never do around Penny lest he provoke a warning. I wonder why I would need him to feel that way around me when I could just tell him what I do want him to do and give him a big cuddle when he does it instead.

With the example I gave about giving a warning growl - Daisy was just being a general PITA and decided she wanted to push the limits. She knows what the command quiet means but on this particular occasion she chose to ignore it and keep barking. I gave a warning growl noise and she realised I meant it, stopped barking immediately and gave up. She ignored a known command and had no interest in having a cuddle, this was not something she did on a regular basis. And being an often over confident dog who will challenge you, I would hazard a guess that it doesn't affect her like it you perceive it affects Kivi.

ETA: I could have just put her in her crate - but she would have kept barking. I guess that would be ok if your neighbours don't mind the noise and you can put up with it until she finally gives up but considering that the growl worked very effectively, and quickly, I wouldn't hesitate to use it again if the situation was appropriate :D

ETA2: Corvus, I'm not having a go at you or anything - I'm just giving my genuine opinion. I hope you're not taking my posts as having a narky or negative undertone, because that's not my intention. Or are you provoking me intentionally to see if you're right about my reasoning ability?

Edited July 27, 2009 by huski

[corvus]

I'm just trying to get you to talk about interruptors, huski, rather than philosophies or methods in general.

I haven't once said that the negative interruptor I use doesn't work the way that I want it to. It works exactly how I want it to, and I don't think it has much of an effect on Kivi except that if I sit there just going "ah-ah" every time he goes to do something because he's in one of those into everything moods he starts whining and chomping on things because he's bored and frustrated and I won't let him do anything he wants to do. I am wondering if I would get the same response if I had a positive interruptor instead. I'm wondering if there's any difference in how a dog behaves if you use a positive interruptor or a negative. What do you think?

I'm also thinking perhaps I could have avoided the whining and chomping thing if I'd reinforced similar behaviour with acceptable objects rather than just hopefully waving a chew toy at him and then complaining that he doesn't like them and would prefer to chomp on furniture.

Ultimately, what is troubling me is whether positive interruptors can do the same job as a negative interruptor. Or a better job? Or a worse job? How do they compare and why? Has anyone actually trained a positive interruptor?

[~*Shell*~]

I use both positive and negative interruptors - for example, Zero is nose to the ground in the backyard (or at training) and I want his attention but "oi" didn't work, I would use the command "watch me" for this kind of situation. Generally if he's just having a sniff and isn't scenting a "watch me" gets his attention pretty well.

"Watch me" doesn't work when he's busy doing his own thing though. Say for example, there's been a dog show at the training grounds the day before and has obviously been a bitch in heat so he's scenting like crazy, instead of just casually sniffing around - for a sibe, anything they want to do is more interesting than anything you have for the most part (food is an exception most of the time in Zero's case and even then, there are times when it doesn't tempt him, even when i waft bbq chicken under his nose) so not listening to me is self-rewarding to him - ignoring me means he gets to keep his head to the ground. If he's completely ignoring me then I will use "uh-uh" which usually brings his focus back to me because it's short, sharp and to the point. I've tried using other noises but this one is the best for us - his name or a different command don't work at this stage.

The other interruptor I use is "arrrrrrrrr" (it's low and sounds like a growl) and this is the interruptor I use when Zero is doing something he knows I won't like. For example, for those of you who don't know, Zero used to be dog aggressive. If he's playing with another dog and goes from being playful to dominant (it's very obvious to me when he is because his stance and posture change), I will use this sound to tell him that it's not okay for him to act like that. He hears the sound, stops and assesses and then usually goes back to playing but without the dominance. I will also use the noise if another dog is annoying him and his reaction to it is too much. I don't stop him from growling because I think another dog needs to hear that they're annoying him but there is a difference between a "stop doing that" growl (i will use an "oi" in this situation because it generally gets the attention of both dogs so I can redirect their attention) and a "I'm going to bite you if you don't stop doing that" growl. Luckily i know my dog enough so that i don't hear the "I'm going to bite you if you don't stop doing that" growl anymore - I remove the problem before it gets to that.

Dogs tell each other when they're doing the wrong thing all the time and I believe their brain was hard wired to hear it and respond accordingly, in the same way that a dog like Daisy is hardwired to sniff, Zero and Micha are hardwired to run and Kivi is hardwired to herd. I've never met a dog who didn't know what "arrrrrr", said with the right tone and inflection, meant.

ETA: You were posting at the same time as me Corvus I would consider "watch me" to be a positive interruptor - it's been drummed into Zero for the last 2.5 years that "watch me" means his favourite treats are coming his way. Unfortunately, sometimes I just don't have the trump card for whatever he wants to do and I could wave anything in the world in his face to get him to watch me and he still wouldn't look at me. At that point, the most rewarding thing to him is whatever he's doing whether it be scenting, pawing at the ground, following the bee that's caught under the grass etc.

Edited July 27, 2009 by ~*Shell*~

[Staranais]

Ultimately, what is troubling me is whether positive interruptors can do the same job as a negative interruptor. Or a better job? Or a worse job? How do they compare and why? Has anyone actually trained a positive interruptor?

If by a positive interrupter, you just mean something like a "watch me", then yes, I have trained one. When my old boy started to posture at other dogs at obedience schools many moons ago, I'd tell him "watch" (trained with food treats), and if we were far enough away from the other dogs and he wasn't too far along in firing up, then he'd generally stop what he was doing and watch me to get the treat he knew would be forthcoming. We did a similar thing later, with drive training, which was much more effective.

Anyway, my thoughts on your questions about the benefits of positive vs negative interrupters:

Advantages of positive interrupter:

1. Teaches dog what it can actually do, so can result in less frustration than just telling the dog "don't do that" and leaving it just sitting there still wanting to chew something (for example) without knowing any legal options for satisfying that urge.

2. It is arguably nicer for the dog not to get growled at by the owner (though I think most dogs cope with this very well, and possibly even like being given clear boundaries in this fasion).

Disadvantages of positive interrupter:

1. A smart dog can learn to engage in the undesirable behaviour in order to get the positive interrupter (i.e "Corvus is being really boring right now. Hmmm, perhaps if I start to chew the furniture she'll start paying attention to me like she always does when I chew the furniture, and we'll have one of those fun training session together with the treats I like, like we did last time when I started chewing the furniture.")

2. I just don't always want to be telling my dog what to do. It got kind of old at dog obedience school, having to be perpetually telling my dog to "watch" and pumping treats into him to ensure he didn't engage in the unwanted behaviour - since he did invariably revert back to his old behaviour if I wasn't constantly interrupting it with commands and food treats. I wonder if it's nicer for the dog too, since negative interupters give them more freedom to choose their own behaviours? i.e, they have more behavioural freedom if they can do anything EXCEPT what you say, rather than ONLY do what you say.

Is that the kind of answer you're after? I personally think that there is a place for a positive and a negative interrupter in training, it gives the maximal amount of information to my puppy if I growl at her or time out her for chewing the furniture, then redirect her to a more appropriate activity with heaps of praise, rather than just one or the other technique.

[Aidan]

Doggy Zen is the concept of "to get the [insert desirable thing here], I must first show some self-control"

It is a foundation behaviour and, thankfully, it generalises quite well if you give it a chance. You start off with some food in your closed hand, present it to the untrained dog. He will no doubt lick, paw, nudge etc to get the food, but you wait until he backs off, then open your hand and let him have the food. Pretty simple stuff, not particularly useful on it's own - but it's a starting point and you build from there.

There is nothing "wrong" with using an interruptor IMHO, even many of the most staunch "purely positive" trainers will use an "interrupt and redirect" approach to some behaviour problems and manners training. An interruptor is something that interrupts the behaviour but has no future effect on it.

Once an interruptor diminishes future behaviour, or becomes attached as a cue, it is no longer an interruptor. For e.g, it would be unwise to verbally interrupt the dog to speed up the "doggy zen" exercises, because then you miss the whole point - the dog starts to look for your cue, rather than adopting a "doggy zen attitude" of self-control and restraint.

Personally, I like my dogs to have that attitude, at least some of the time! I do enjoy teaching a dog self-control, I do not enjoy micro-managing.

I don't like my dogs to rush out an open door, steal party snacks, jump the baby gate, chase the cats around the house, jump on me for attention, bark to come inside/outside etc etc - It's all "zenned". Not interrupted, not punished, not put on cue. If you want to go out the door, sit while I put your leash on. If you want a piece of cabana, wait over there and I will give you a bit soon. If you want to come into the living room, sit behind the baby gate until I'm ready to let you in... I start off simple, manage the environment, then proceed from there. A small initial investment leads to long term dividends.

I do use interruptors, I do use punishers, and I certainly use cues - but if I can "zen" it, that will be my first choice. Why? It has carry-over into real life situations. All through this thread I have seen people talking about "what if your reinforcer isn't strong enough?" or "what if the behaviour is self-reinforcing?" It doesn't really matter. An interruptor isn't going to work either in those cases, and a negative reinforcer is under the same limitations as a positive reinforcer. You can turn the stim up on an e-collar, but if the dog doesn't know what to do it's just a shock, even in skilled hands. What is important is your reinforcement history. If the dog has learned what works, that is what he will tend to do. If he's out of his depth, you're not necessarily going to fix it by switching to a "negative".

If a dog has self-control then it all becomes much easier. I'm not saying my dogs have monk-like powers of self-control, let's be realistic, but I do know they are easier to manage around things they want than dogs who have not learned these lessons.