Training Out An Unwanted Behavior

[Kavik]

While I generally subscribe to training under threshold and LAT for aggression, it was very interesting that some people who helped me with Zoe's aggression were able to put her in a yard with their GSDs (I boarded her there for a while), and she did eventually stop being aggressive, the other dogs did not react.

Did you have the GSD in your signature with Zoe at the same time and she was ok with your GSD? I have had a GSD who was ok with GSDs but not with other breeds. I thought perhaps because he lived with my other two GSD's and was ok with them, he had a breed tolerance. To expand on this, I took him to the GSD club once and his behaviour was fine until he saw someone with a Staffy in the car park and he lost the plot

LAT works well with dogs who have drive in the right place to achieve good handler focus.....a working Kelpie I imagine would have that inherent drive, but the reactive GSD's I have experienced that didn't respond to LAT type regimes well, were not good at achieving handler focus in other training areas either and were hard work compared with others to train through distractions.

No, at the time I didn't have Diesel, I got him later from the same people who helped me with Zoe :) Yes, Zoe had great food drive and prey drive, and if she hadn't been dog aggressive would have been a great sports dog (which is why I got her), I was able to get good handler focus with her, even though I didn't know nearly as much as I do now.

[Staranais]

Am I right in thinking that a dog can un-learn helplessness if taught (and permitted) alternative options?

That's the strategy you would employ. Thankfully it's not common, more often the reactive behaviour just escalates until the owner or trainer give up on flooding. It's pretty hard to watch.

I had a client who had another trainer come and do this with their dog. The trainer was resolute, but he got bitten and had to stop to deal with the injury. So the dog learned to bite, and after that bit everyone in the house.

Aidan - on similar but different tack I'd love to get your take on something.

I was rereading part of the 3rd Steven Lindsay book yesterday (Procedures & Protocols), and was dumbfounded to find a snippet where he talked about dogs with handling aversion/aggression. He says that in many cases, desensitization and counterconditioning are insufficient to change the dog's attitude towards handling, and negative reinforcement strategies are ultimately needed (basically, to restrain the dog until it "gives up" and experiences relief and relaxation).

I don't have the exact quote with me since I'm at work right now and the book it at home, but I can certainly dig it up if you don't know the paragraph I mean (and also to check that I'm not remembering incorrectly - I may be, but I think I have the gist correct).

I was dumbfounded since my education has always been that D&C are the go with most aggression issues, and what Lindsay was advocating does sound awfully like flooding?

Given your experience, would you agree with Steven Lindsay?

[Simply Grand]

Am I right in thinking that a dog can un-learn helplessness if taught (and permitted) alternative options?

That's the strategy you would employ. Thankfully it's not common, more often the reactive behaviour just escalates until the owner or trainer give up on flooding. It's pretty hard to watch.

I had a client who had another trainer come and do this with their dog. The trainer was resolute, but he got bitten and had to stop to deal with the injury. So the dog learned to bite, and after that bit everyone in the house.

Aidan - on similar but different tack I'd love to get your take on something.

I was rereading part of the 3rd Steven Lindsay book yesterday (Procedures & Protocols), and was dumbfounded to find a snippet where he talked about dogs with handling aversion/aggression. He says that in many cases, desensitization and counterconditioning are insufficient to change the dog's attitude towards handling, and negative reinforcement strategies are ultimately needed (basically, to restrain the dog until it "gives up" and experiences relief and relaxation).

I don't have the exact quote with me since I'm at work right now and the book it at home, but I can certainly dig it up if you don't know the paragraph I mean (and also to check that I'm not remembering incorrectly - I may be, but I think I have the gist correct).

I was dumbfounded since my education has always been that D&C are the go with most aggression issues, and what Lindsay was advocating does sound awfully like flooding?

Given your experience, would you agree with Steven Lindsay?

BAT and CAT use negative reinforcement and seem to be the next step if D&C aren't effective. They are not about just restraining the dog til it gives up though, they are about capturing and rewarding (through negative reinforcement - increased distance from the other dog) an alternative behaviour to reacting. Doesn't sound like that's what Lindsay was talking about though

[Staranais]

Yes, in some ways it sounded like a similar concept to what I have read about BAT/CAT (teaching the dog that appropriate behavior removes pressure, whereas aggression does not). He did not go into the specifics or mechanics, so I'm unsure if he worried about the dog going over threshold while being restrained (compared to both BAT/CAT, where you're supposed to try to stay under threshold if you can).

Also please note that he was specifically discussing dogs that react aggressively to being handled or restrained, hence the reason for enforced restraint until the dog realizes that being restrained is "safe" - I may not have explained that sufficiently well in my last post.

I was particularly interested in that he seemed to present the pressure on/off technique as not only an alternative to D&C, but ultimately a more successful method. The man generally seems to be both educated and to have his head screwed on pretty well, so I'd be unwilling to dismiss what he says without further investigation, however it's not something I've heard much about before. :)

[Staranais]

Double post :)

Edited April 12, 2014 by Staranais

[Aidan3]

I think it would depend on the dog and it's history with the handler. I really don't see a lot of handler aggression (outside of rescues) so I may not be the best person to comment on this. Did Lindsay provide a rational argument for this claim, or was it just his opinion? There are no studies that I know of which look at it, but drawing on behavioural principles which are well understood and common sense, I'd say it would be a risky strategy. Although the pay-off for success is high, the price of failure is very steep!

Like most things, I'd split it down, use a high rate of reinforcement, and shape appropriate behaviours under a variety of relevant situations without skipping crucial steps, setting the dog up for success at every step along the way. It's a low-risk, high-reward strategy that doesn't require a great deal of patience usually.

Owners who are in that situation usually need coaching in many other areas and I wouldn't be encouraging them to use risky strategies that require skills that they have already proven not to possess.

[corvus]

Aggression is complicated. The end.

Seriously, this is one of those areas of behaviour where it's particularly critical to understand the function of the behaviour, and that also means considering emotional state and arousal and how they influence how a dog interprets signals and how they are likely to respond to them. As it happens, I wrote an article about that one as well: http://www.creatureteacher.com.au/Click!%20Article.pdf. There is fair argument in the literature to consider emotional state as inherently tied to how close an animal is coming to their goals. If they are getting closer to obtaining their goals, they are happier, and if they are getting farther, they are unhappier. The emphasis is on knowing what the animal's goals are. If they want the scary thing to go away, anything that seems to achieve that will make them feel better. There are all sorts of problems surrounding this. One is that they learn making a big song and dance and aggressive display usually works, but it gets them all worked up and upset and if it doesn't work they basically think they are going to die. So even just making them stick it out until they realise it doesn't work is problematic. Their stress response is likely to be enormous, and even if they do habituate like Zoe did, that doesn't mean they are 'okay' in that they are not stressed or frightened or behaviourally suppressed. They are just coping the best they can. Sometimes that means a successful habituation where they truly can relax and sometimes it just means they are stuck in this terrible place where they don't know how to behave to obtain their goals. The two look pretty similar, at least if you only look short-term in one context. Even if we can use some kind of exposure or flooding to cut arousal down, it doesn't necessarily mean we also handle the fear. Possibly we just teach the dog that there are some problems in life they can't fix, and that is damaging to future training.

Anyway, LAT works fine with the vast majority of dogs. It's cool like that. You don't need handler focus, because you are shaping it in anyway. If it's not working it's because arousal is way too high and you are setting the dog up with a powerful conflict: Attend to me or attend to the extremely arousing, attention-demanding thing. The point of LAT is to try to eliminate or at least greatly reduce that conflict by allowing them to do both, and thereby getting your foot in the door with your dog and shaping a really sensible coping behaviour, but it depends on the arousing thing being not so arousing that it demands their undivided attention. You'll see with a dog that is practiced at LAT that they will zip back and forth at a much faster rate the more aroused they get, so you kind of get an inbuilt meter of arousal. It tells you how demanding the arousing stimulus is for them and therefore how conflicted they are controlling their behaviour around it.

FWIW, when I was having troubles with my hare, I couldn't DS because he would already be too aroused as soon as he even saw me. He did not calm down. He sure as hell would not eat. There's no way I would have flooded him. It would have made the problem worse if anything. I taught him a safety signal linked to my departure. It presumably inhibited his fear enough that I could get some DS done. It took a LONG time. Months. If I had an animal that needed to be handled in that time it would really throw a spanner in the works with a DS/CC protocol. I know people that have used negative reinforcement in similar cases to give a dog control over their exposure and it has worked all right, but with dogs that already had good relationships with their owner.

[Erny]

I like what you've written, Corvus.

Need a "like" button here on DOL.

[Kavik]

To clarify, I used LAT with Zoe with a fair bit of success (considering I started it when she was 7 years old), and it would be one of my go to methods if I get problems like that again (touch wood I never let one get to that point again!). But it was very interesting that these people were able to put her in with their dogs like that, and her dogs didn't react (she was only about a year old when she was put in with the GSDs).

[Aidan3]

I recently used a "CAT-but-with-flooding" approach with an intractable reactive GSD. They'd been seeing me some time back, quick initial progress that stalled out. I sent them off to a vet behaviourist. Meds didn't help much. Back to me after this last Christmas, no improvement by end of March. So we discussed it, laid out the risks, confirmed the function of her behaviour again, and within two sessions we had her sharing treats with my GSD, Toby.

There's still a long road ahead, mostly because it's hard to find more Tobys.

It's definitely not a first option. Or even a second or third. So much depends on the dogs you're working with you have to at least get the human part right, no stuff ups.

After the second session I was starting to think "I should do this with more dogs" but after a few days I was feeling a bit more sensible :laugh:

[Simply Grand]

I like what you've written, Corvus.

Need a "like" button here on DOL.

I was planning to respond with "I love you Corvus" but a like button would work just as well.

And I've said it before, but I still love Aidan too.

In the informative and understandable posts context of the word :)

[Erny]

It's definitely not a first option. Or even a second or third. So much depends on the dogs you're working with you have to at least get the human part right, no stuff ups.

After the second session I was starting to think "I should do this with more dogs" but after a few days I was feeling a bit more sensible.

:laugh:

Edited April 12, 2014 by Erny

[Staranais]

I think it would depend on the dog and it's history with the handler. I really don't see a lot of handler aggression (outside of rescues) so I may not be the best person to comment on this. Did Lindsay provide a rational argument for this claim, or was it just his opinion? There are no studies that I know of which look at it, but drawing on behavioural principles which are well understood and common sense, I'd say it would be a risky strategy. Although the pay-off for success is high, the price of failure is very steep!

Hmmm, found it. :)

He's talking specifically about dogs with fears associated with painful or uncomfortable handling procedures (grooming, nail clipping, veterinary procedures, etc).

To quote the entire paragraph: “Some dogs appear to be more sensitive to touch and prone to develop persistent fears associated with discomfort and painful handling. The usual procedures used for resolving such problems employ some combination of graded interactive exposure with RP-CC. Although conscientious efforts should be made to counter-condition a fearful dog with treats and relaxing massage while it undergoes progressive exposure to the feared activity, it is imperative that avoidance and escape be blocked. Very often in such cases counterconditioning efforts will achieve only a small portion of the desired effect. Response prevention using physical restraint followed by massage as the animal begins to relax can be very useful. It is important for the dog to become relaxed before it is released from restraint. In the case of dogs that become highly reactive, they should be held in restraint (with massage) for an additional 3 minutes after the last strong effort to break free.”

No references given. Page 158 in my hardcover edition of vol 3.

[Simply Grand]

I think it would depend on the dog and it's history with the handler. I really don't see a lot of handler aggression (outside of rescues) so I may not be the best person to comment on this. Did Lindsay provide a rational argument for this claim, or was it just his opinion? There are no studies that I know of which look at it, but drawing on behavioural principles which are well understood and common sense, I'd say it would be a risky strategy. Although the pay-off for success is high, the price of failure is very steep!

Hmmm, found it. :)

He's talking specifically about dogs with fears associated with painful or uncomfortable handling procedures (grooming, nail clipping, veterinary procedures, etc).

To quote the entire paragraph: “Some dogs appear to be more sensitive to touch and prone to develop persistent fears associated with discomfort and painful handling. The usual procedures used for resolving such problems employ some combination of graded interactive exposure with RP-CC. Although conscientious efforts should be made to counter-condition a fearful dog with treats and relaxing massage while it undergoes progressive exposure to the feared activity, it is imperative that avoidance and escape be blocked. Very often in such cases counterconditioning efforts will achieve only a small portion of the desired effect. Response prevention using physical restraint followed by massage as the animal begins to relax can be very useful. It is important for the dog to become relaxed before it is released from restraint. In the case of dogs that become highly reactive, they should be held in restraint (with massage) for an additional 3 minutes after the last strong effort to break free.”

No references given. Page 158 in my hardcover edition of vol 3.

Interested to hear Aidan's take but that's certainly what I've done with dogs that require grooming but are highly resistant to it. A simple example is dogs that don't want their nails clipped for whatever reason, I will hold their paw firmly but not hurting them until they settle down and stop trying to pull it away before releasing it.

[Amax-1]

No, at the time I didn't have Diesel, I got him later from the same people who helped me with Zoe :) Yes, Zoe had great food drive and prey drive, and if she hadn't been dog aggressive would have been a great sports dog (which is why I got her), I was able to get good handler focus with her, even though I didn't know nearly as much as I do now.

Not all dogs share equal ability to focus through distractions even in the same breed which is most definitely a genetic component. Working with GSD's bred on imported Schutzhund titled bloodlines compared to locally produced bloodlines, there is a massive difference in the ability to focus and to the extreme of some pups appearing as if they have done it all before in a former life.....Malinois from sporting lines are much the same in focus ability motivated by reward. LAT type training methods IME are best applied to dogs with enough genetic focus to support the method and with dogs short on genetic ability to focus, other training methods provide better results.

I was rereading part of the 3rd Steven Lindsay book yesterday (Procedures & Protocols), and was dumbfounded to find a snippet where he talked about dogs with handling aversion/aggression. He says that in many cases, desensitization and counterconditioning are insufficient to change the dog's attitude towards handling, and negative reinforcement strategies are ultimately needed (basically, to restrain the dog until it "gives up" and experiences relief and relaxation).

That's basically what Cesar Milan applies irrespective of choke outs and alpha rolls to achieve restraint, once the dog has given up the fight, relaxation techniques are then used to reinforce calm behaviour.

Edited April 14, 2014 by Amax-1

[Aidan3]

I think it would depend on the dog and it's history with the handler. I really don't see a lot of handler aggression (outside of rescues) so I may not be the best person to comment on this. Did Lindsay provide a rational argument for this claim, or was it just his opinion? There are no studies that I know of which look at it, but drawing on behavioural principles which are well understood and common sense, I'd say it would be a risky strategy. Although the pay-off for success is high, the price of failure is very steep!

Hmmm, found it. :)

He's talking specifically about dogs with fears associated with painful or uncomfortable handling procedures (grooming, nail clipping, veterinary procedures, etc).

To quote the entire paragraph: “Some dogs appear to be more sensitive to touch and prone to develop persistent fears associated with discomfort and painful handling. The usual procedures used for resolving such problems employ some combination of graded interactive exposure with RP-CC. Although conscientious efforts should be made to counter-condition a fearful dog with treats and relaxing massage while it undergoes progressive exposure to the feared activity, it is imperative that avoidance and escape be blocked. Very often in such cases counterconditioning efforts will achieve only a small portion of the desired effect. Response prevention using physical restraint followed by massage as the animal begins to relax can be very useful. It is important for the dog to become relaxed before it is released from restraint. In the case of dogs that become highly reactive, they should be held in restraint (with massage) for an additional 3 minutes after the last strong effort to break free.”

No references given. Page 158 in my hardcover edition of vol 3.

Interested to hear Aidan's take but that's certainly what I've done with dogs that require grooming but are highly resistant to it. A simple example is dogs that don't want their nails clipped for whatever reason, I will hold their paw firmly but not hurting them until they settle down and stop trying to pull it away before releasing it.

Response prevention is reasonable in these sorts of situations (after all, we use response prevention in some way in nearly everything we do with dogs). As long as it's safe and well considered. Having trained dogs and other animals to do stuff they really, really don't want to do without any restrictive sort of response prevention many times I'd have to argue that counter-conditioning through differential reinforcement certainly can be effective, so I don't completely agree with Lindsay.

[Staranais]

Thanks Aidan. I wish Lindsay had gone into more depth - it seems like such a brief, throwaway paragraph to assert such a controversial statement in. Even just given us the % of dogs that he has worked with where he's had success with pure D&CC vs needing to resort to enforcing contact and then waiting the dog out would have been interesting.

[Staranais]

Interested to hear Aidan's take but that's certainly what I've done with dogs that require grooming but are highly resistant to it. A simple example is dogs that don't want their nails clipped for whatever reason, I will hold their paw firmly but not hurting them until they settle down and stop trying to pull it away before releasing it.

That's pretty relevant to one scenario, I was considering, Simply Grand - I'm a vet, and one recent case in particular sticks in my mind. Last month we had a fearful and aggressive, extremely ill, chihuahua whom we had in the hospital for several weeks. She screamed and fought and bit with any handling or restraint, no matter how gentle the restraint was, no matter if it was her owner or staff that were restraining her. It wasn't her illness - the owner said she had never been able to restrain the dog for anything the dog disliked. This dog was too ill for food to be a reinforcer. Too upset with us for petting or social contact to be a reinforcer. Too ill for us to dare to chuck sedatives or anxiolytics at her to facilitate handling. She was dangerous to staff and to herself - and unfortunately she was ill enough that we did need to get blood work done, imaging done, and an IV line in immediately.

During her time in hospital the nurses would restrain her - as gently as possible! - for just as long as necessary to do whatever we needed to do, then release her. They'd try to make friends in between times - going up to her and talking nicely, not putting any pressure on her, just trying to get her used to being approached by staff. However by the end of her stay, she was still at least as terrified and aggressive as when she first came in.

In retrospect, perhaps every time she was pinned down, if they waited until she was CALM until they released her, her stay would ultimately have been less terrifying for her?

Idle thoughts. :p

Edited April 18, 2014 by Staranais

[m-j]

Response prevention is reasonable in these sorts of situations (after all, we use response prevention in some way in nearly everything we do with dogs). As long as it's safe and well considered. Having trained dogs and other animals to do stuff they really, really don't want to do without any restrictive sort of response prevention many times I'd have to argue that counter-conditioning through differential reinforcement certainly can be effective, so I don't completely agree with Lindsay.

Not only is it effective but I believe can sometimes be the best way to quieten a large animal (like a young horse that has had very little or no handling) so the trainer stays safe.