Rehabilitating An Aggressive Dog

[abed]

Hello to everyone.

I am a working dog trainer and I don't generally enter into training discussions on pet dog forums due to the fallout and arguments that occur in the process and personally I am not interested those debates. Having said that, I deal with dog and human aggressive dogs on a regular basis and although you can never trust a DA or HA dog to behave in all circumstances, you can infact raise the reactive thresholds and with a good handler who can read the dog well, you can generally improve the behaviour to the point that the average person seeing the dog out and about would never know the dog had DA or HA tendencies underlying and in most circumstances the dog will behave well on leash. Part of the success is the handler knowing when to back the dog out of over confronting situations that is likely to light the dog up being a process of gradual exposure and desensitisation that takes time.

The dog the OP describes seems more like a dog that wants to get the other dog before the other dog gets him given that he dragged the handler 3 metres to mount the attack. With dogs who react in that way, I use a stabilisation collar or a Dominant Dog Collar as they are also known for the purpose that if the dog does light up in the rehabilitation process, the dog can be restrained and corrected with the drive to attack taken away from the dog. I agree that a physical correction can escalates the problem when the correction is not hard enough which is usually the case. A hard enough correction for the dog to remember the consequence of mounting an unwarranted attack will cause the dog avoidance of the action it was corrected for which avoidance of aggressive lunging is what the handler is looking to achieve. If the dog tucks it's tail seeing another dog and behaves by fearing a correction if deciding to lunge, I am personally happy with that, as it's a step in the right direction. Also the dog learns that the other dogs are not a threat and aggression wasn't necessary as the other dogs didn't do anything and they were nothing to fear in the first place.

With the worse cases of lunging aggression towards people or other dogs that I have worked with, have stopped the lunging altogther after 3 corrections with the DD collar and the lesser aggressive types have stopped the behaviour after one correction. It's true that DD collar corrections masks the behaviour temperarily, it doesn't fix it but povides the foundation to work with for long term rehabilitation which is basically finding a job the dog will focus on best whilst in the proximity of other dogs during desensitisation and using known non reactive calm dogs as the decoys when reducing distances and exposure to other dogs.

My worse case of late was a GSD pet dog that was both HA and DA, a dog that reacted from insecurities but was civil enough in drive to have a go and reduce it's stress and was a very dangerous animal to anything or anyone unfamiliar to the dog. After initial training with me, the owners worked with the dog to reshape the behaviour over about 6 months and what resulted is this:

You can't trust the dog off leash in a public place and would never be a consideration with this dog. The dog uses a martingale collar and behaves well on leash. Strangers can't pat the dog and the dog would light up if another dog got in his face or acted aggressively towards him, but providing the owners don't place the dog in these positions, he is calm and obedient. It was 12 months in November since the dog has aggressively lunged, but the owners have done a great job in his rehabilitation and reading the dog and being on the alert that he could react. Serious DA and HA I don't believe can be fixed to the point of having confidence as you would with a dog that has never displayed the behaviour and reactivity, it's a condition with the right owner dedicated to manange the situation can result in a good pet in the right circumstances as far as a sporting dog or a dog that must be dog and stranger freindly, it has the wrong temperament to push the envelope that far.

I might add that the use of DD collars effectively should be used in consultation with a trainer experienced in these methods of correction.

[abed]

'Cosmolo' date='6th Dec 2010 - 08:27 AM' post='5005350']

I used to think aggression could be cured and i used to think using physical aversives was almost essential at some point in the process. Experience has changed my views significantly and while i still believe that in some cases application of a correction can be helpful, i know it is a simple behaviour block at a point in time- sometimes i believe this is helpful so that the other desensitisation processes can take place more easily, in other cases it is not needed.

In the desensitising process, unless the dog lights up in an aggressive reaction there is no reason for any type of correction to be applied, so contemplating a correction or aversive measure in the process, the dog must be lighting up or acting aggressively which provides us with the options below:

1. Drop the leash and let the dog go.

2. Drag the dog backwards and heighten defence drive through agitation

3. Correct the dog and take the drive away

I don't agree that either dropping the leash allowing the dog to attack or ramping it up in collar agitation is beneficial in the rehab process. Obviously no one will apply option 1, drag the dog back giving commands the dog ignores which is what most people do is bad IMHO especially if the other dog retreats allowing the aggressive dog to win will bring a large proportion of the desensitsing program undone. If in the situation during desensitisation that the dog does light up in aggression, correction is the only beneficial option to keep the work done intact.

Edited December 6, 2010 by abed

[corvus]

In the desensitising process, unless the dog lights up in an aggressive reaction there is no reason for any type of correction to be applied, so contemplating a correction or aversive measure in the process, the dog must be lighting up or acting aggressively which provides us with the options below:

1. Drop the leash and let the dog go.

2. Drag the dog backwards and heighten defence drive through agitation

3. Correct the dog and take the drive away

Or wait it out. In CAT (Constructional Aggression Treatment - like BAT but slightly different) I think they wait it out. Sometimes increasing distance is reinforcing, so they just stop to make sure the dog isn't reinforced one way or another.

You should see my hare after his safety signal therapy. I recently went on holidays, usually a surefire way to end up with a skittish hare that hides from everyone for at least a couple of days. This time I had his sitter use the safety signal and came back to a hare more bold and relaxed than I'd seen him in years. It was completely incomprehensible to me. In the 6 years I've had him I've never seen him like that after a holiday. He normally hates it when I mess up his routine by going away, and gets progressively panicky until his routine goes back to normal. I don't know if I'm doing the right thing, but I've started using his safety signal kind of as a marker for responding to pressure the way I want him to - mildly and stopping in the open to assess the situation further before running blindly into the side of his cage. I think it keeps him calm and thinking. There are definitely limits, especially with a wild hare, but coupled with operant conditioning I think it has a lot of promise.

Steven Lindsay uses a relaxation training protocol he suggests is good to teach a dog before treating fear or aggression. He couples it with a scent, a verbal cue, and a visual/tactile cue (a blanket).

Edited December 6, 2010 by corvus

[Aidan]

1. Drop the leash and let the dog go.

2. Drag the dog backwards and heighten defence drive through agitation

3. Correct the dog and take the drive away

There are more than 3 options.

[raineth]

I really enjoyed your post too Kelpie-i ;)

from my understanding, I think rather than the focus being on greater production of certain neurotransmitters; what would be occurring is a strengthening of the connections between certain neurons.

With every repetition of the dog learning to have a different emotional reaction to dogs certain connections between neurons are strengthened.

what can also occur is that the other connections that were resulting in a negative emotional (and behavioural response) can become weaker. As Aidan mentioned though it is not unlearned. But the connection can be weakened.

Some research has found a certain number of repetitions (or really certain neurons always firing before other neurons) will always produce a strong connection (known as long-term potentiation).

I guess this is why you can often see little progress, then suddenly its almost like a switch goes on in their head and you get much better progress

Wow, thanks for that explanation raineth and Aidan...raineth I hope you don't mind if I cut/paste it for future reference

When you say the "connection can be weakened", this would be the dog regressing due to lack of regular exposure...because the brain no longer needs to fire the necessary neurons....is this correct?

Hi Kelpie-i, uou can use it- but really its just my musings on the subject and should be taken with a grain of salt

when i spoke about connections being weakened I was more talking about the original connection (associated with the presence of another dog = fear and aggressive beahviour). So what the training is doing is a kind of re-wiring of circuits. A desensitisation program should re-wire the brain so that (like Aidan said) the dog can feel safe in the presence of another dog; in doing this the circuit that says 'other dog = fear' is not active and can eventually go through "long term depression" (which is the opposite of long term potentiation).

Would it be the same as the critical socialisation period of the puppy where the denditric branching occurs with every new experience during the first 16 weeks? Would sound similar... :rolleyes:

As far as I know it is similar because it is all about neuroplasticity. But the difference is that in a really young animal plasticity is going on throughout all regions in the brain. Once the animal is older plasticity mainly only occurs in the hippocampal formation. I can't quite remember but I think there might be a couple of other areas it can occur, but the hippocampal formation is the major one. the hippocampal formation plays a major role in learning and memory.

I should also ass that when I say 'animal' I don't really kinow if that occurs in every animal - I'm only going off the animals that the researchers have used

anyway, all of that above is still just my musings and should be taken witha grain of salt as well its just a topic I find really interesting!

[raineth]

'Cosmolo' date='6th Dec 2010 - 08:27 AM' post='5005350']

I used to think aggression could be cured and i used to think using physical aversives was almost essential at some point in the process. Experience has changed my views significantly and while i still believe that in some cases application of a correction can be helpful, i know it is a simple behaviour block at a point in time- sometimes i believe this is helpful so that the other desensitisation processes can take place more easily, in other cases it is not needed.

In the desensitising process, unless the dog lights up in an aggressive reaction there is no reason for any type of correction to be applied, so contemplating a correction or aversive measure in the process, the dog must be lighting up or acting aggressively which provides us with the options below:

1. Drop the leash and let the dog go.

2. Drag the dog backwards and heighten defence drive through agitation

3. Correct the dog and take the drive away

I don't agree that either dropping the leash allowing the dog to attack or ramping it up in collar agitation is beneficial in the rehab process. Obviously no one will apply option 1, drag the dog back giving commands the dog ignores which is what most people do is bad IMHO especially if the other dog retreats allowing the aggressive dog to win will bring a large proportion of the desensitsing program undone. If in the situation during desensitisation that the dog does light up in aggression, correction is the only beneficial option to keep the work done intact.

that to me deoesn't really sound like desensitisation :rolleyes:

To me desenstisation is acutally really habituation; which means learning that a certain 'something' is nothing to worry about.

I would think that by using punishments you can't achieve habituation.

I don't doubt that it is taching the dog something, but it is not teaching the dog that the presence of another dog is nothing to worry about.

[Quickasyoucan]

I really enjoyed your post too Kelpie-i

from my understanding, I think rather than the focus being on greater production of certain neurotransmitters; what would be occurring is a strengthening of the connections between certain neurons.

With every repetition of the dog learning to have a different emotional reaction to dogs certain connections between neurons are strengthened.

what can also occur is that the other connections that were resulting in a negative emotional (and behavioural response) can become weaker. As Aidan mentioned though it is not unlearned. But the connection can be weakened.

Some research has found a certain number of repetitions (or really certain neurons always firing before other neurons) will always produce a strong connection (known as long-term potentiation).

I guess this is why you can often see little progress, then suddenly its almost like a switch goes on in their head and you get much better progress

Wow, thanks for that explanation raineth and Aidan...raineth I hope you don't mind if I cut/paste it for future reference ;)

When you say the "connection can be weakened", this would be the dog regressing due to lack of regular exposure...because the brain no longer needs to fire the necessary neurons....is this correct?

Hi Kelpie-i, uou can use it- but really its just my musings on the subject and should be taken with a grain of salt

when i spoke about connections being weakened I was more talking about the original connection (associated with the presence of another dog = fear and aggressive beahviour). So what the training is doing is a kind of re-wiring of circuits. A desensitisation program should re-wire the brain so that (like Aidan said) the dog can feel safe in the presence of another dog; in doing this the circuit that says 'other dog = fear' is not active and can eventually go through "long term depression" (which is the opposite of long term potentiation).

Would it be the same as the critical socialisation period of the puppy where the denditric branching occurs with every new experience during the first 16 weeks? Would sound similar... :rolleyes:

As far as I know it is similar because it is all about neuroplasticity. But the difference is that in a really young animal plasticity is going on throughout all regions in the brain. Once the animal is older plasticity mainly only occurs in the hippocampal formation. I can't quite remember but I think there might be a couple of other areas it can occur, but the hippocampal formation is the major one. the hippocampal formation plays a major role in learning and memory.

I should also ass that when I say 'animal' I don't really kinow if that occurs in every animal - I'm only going off the animals that the researchers have used

anyway, all of that above is still just my musings and should be taken witha grain of salt as well its just a topic I find really interesting!

I find it really interesting too, especially since I suffer from anxiety.

I always thing of it as forging new neural pathways and weakening old ones, but I think (both with animals and humans) there is always the danger of falling back onto the old pathway, it never totally goes away.

The only thing I would add about the use of drugs is that part of the reason drugs are used to treat anxiety (in humans) is because research has shown that people with depression/anxiety do not have normal chemical reactions in their brains (being general as I can't remember if the serotonin is taken up too fast or not fast enough and there is a lot more to it than that that I also can't remember right now ).

So, not talking about this particular case, but in general I think there would always be a place for drug therapy with behavioural therapy if it was proven that a dog had an actual chemical imbalance that was causing it to behave the way it was. This talking about a minority of cases not a majority.

[Aidan]

For anyone with an unusually deep interest in the science (or a great deal of persistence in nutting it out), this paper is worth a read:

http://www.cell.com/neuron/abstract/S0896-6273(05)00159-5

(Sorry for being a bit off-topic)

[JulesP]

I agree that drugs should not be discounted. I had a little pup that was the most miserable thing I have ever seen. She looked like a grumpy 20 yo dog that just wanted to die. Was both dog and people aggressive. A couple of weeks on prozac and she was a new dog.

It is also amazing what imbalances in vitamins/minerals can do to a creature.

[Quickasyoucan]

I agree that drugs should not be discounted. I had a little pup that was the most miserable thing I have ever seen. She looked like a grumpy 20 yo dog that just wanted to die. Was both dog and people aggressive. A couple of weeks on prozac and she was a new dog.

It is also amazing what imbalances in vitamins/minerals can do to a creature.

Sounds just like me, except I was never dog aggressive pre-medication :rolleyes:

[mita]

I really enjoyed your post too Kelpie-i :rolleyes:

Me, too. It was one of those posts that I regard as a 'keeper' to squirrel away for future reference. The key issues were spot on & a compelling case made.

Aidan, I don't want to be critical of your point about learning from neuroscience. Because I totally agree it's an area with a lot to offer.

But, quite rightly, researchers in neuroscience make the point that caution is needed in translating their findings into behaviours. Behaviours, animal & human, are mediated by a myriad of other factors, too.

All of which make actual, individual behaviour, not a simple matter to predict.

I liked the way Kelpie-i made the case more on risk-management.

[raineth]

So, not talking about this particular case, but in general I think there would always be a place for drug therapy with behavioural therapy if it was proven that a dog had an actual chemical imbalance that was causing it to behave the way it was. This talking about a minority of cases not a majority.

yeah I thinkk the best outcomes for things like anxiety and depression combines medication and therapy ;)

For anyone with an in the science (or a great deal of persistence in nutting it out), this paper is worth a read:

http://www.cell.com/neuron/abstract/S0896-6273(05)00159-5

(Sorry for being a bit off-topic)

thanks Aidan I'll have a look at that. Although the phrase unusually deep interest makes me feel almost like a deviate :rolleyes:

And yes sorry for being a bit off topic OP

Edited December 7, 2010 by raineth

[JulesP]

I agree that drugs should not be discounted. I had a little pup that was the most miserable thing I have ever seen. She looked like a grumpy 20 yo dog that just wanted to die. Was both dog and people aggressive. A couple of weeks on prozac and she was a new dog.

It is also amazing what imbalances in vitamins/minerals can do to a creature.

Sounds just like me, except I was never dog aggressive pre-medication :rolleyes:

;)

[Aidan]

thanks Aidan I'll have a look at that. Although the phrase unusually deep interest makes me feel almost like a deviate ;) :rolleyes:

Well if the shoe fits...

[Aidan]

But, quite rightly, researchers in neuroscience make the point that caution is needed in translating their findings into behaviours. Behaviours, animal & human, are mediated by a myriad of other factors, too.

All of which make actual, individual behaviour, not a simple matter to predict.

The neuroscience doesn't make it any simpler to predict! :rolleyes: I agree though, my point was only to validate Kelpie-i's perspective from another angle. Behavioural science had evidence for this long before neuroscience, but neuroscience has filled a lot of the gaps. The two disciplines are working very nicely together, the paper I linked to is a good example.

[Kelpie-i]

OFF TOPIC....SORRY!!!!! raineth, mita and Aidan, what sort of science degrees have you guys studied, or are studying. Even before this thread I was looking at going back to school (university actually !!) to study animal ethology to help me better understand the "inside going's on" of dogs. Should have done it moons ago but better late than never I reckon Where do you recommend I start looking and what steps are required? Any help would be greatly appreciated

[Aidan]

OFF TOPIC....SORRY!!!!! raineth, mita and Aidan, what sort of science degrees have you guys studied, or are studying. Even before this thread I was looking at going back to school (university actually !!) to study animal ethology to help me better understand the "inside going's on" of dogs. Should have done it moons ago but better late than never I reckon Where do you recommend I start looking and what steps are required? Any help would be greatly appreciated

I'm studying Behavioural Science with a Psych major. You might be better of with a Bachelor of Science, there is a certain course outline for Psych that is more relevant to human psychology. If you make contact with the faculty of science at a uni nearby they will put you in touch with the right people, this time of year is a good time to do it.

I'll be nearing 40 by the time I finish, one of my mates is doing the same course in his 50s, although full-time so won't be moving at my plodding pace.

[abed]

1. Drop the leash and let the dog go.

2. Drag the dog backwards and heighten defence drive through agitation

3. Correct the dog and take the drive away

There are more than 3 options.

There probably is too, but every second the dog is forging at the end of the leash in defence driven aggression, the aggression level and determination to attack the other dog is rapidly rising. It's the reason that applying a weak correction or pain induced aversive with prong and Ecollars at that point can escalate the situation and agitate the drive into a higher level and on occassions will provoke redirected aggression where the dog can bite the handler. The only thing that will lower the drive naturally in the aggressive dog is the other dog running away.

The classical symptom of dog aggression is the dog lighting up in defense at the end of the leash and ignors known commands and the handler looses control of the dog other than physical restraint, and how that sitiuation is addressed has a major bearing IMHO upon the strength of the rehabilitation platform. The dog reacts to try and lower it's stress level and the worse thing is the dog to learn that an aggressive reaction reduces the stress and becomes a learned behaviour and becomes part of the dogs disposition. This happens often in working breeds where a dog owner likes the idea of their 3 month old Rotty pup or GSD barking at strangers thinking they have the makings of a good protection dog. The puppy barks through insecurity and the owners reward the behaviour and the person or dog the puppy barks at retreats. The puppy learns this behaviour and by the time it's 9 or 10 months old, the owner has developed a fear biter by that time is uncontrollable.

[abed]

'Cosmolo' date='6th Dec 2010 - 08:27 AM' post='5005350']

I used to think aggression could be cured and i used to think using physical aversives was almost essential at some point in the process. Experience has changed my views significantly and while i still believe that in some cases application of a correction can be helpful, i know it is a simple behaviour block at a point in time- sometimes i believe this is helpful so that the other desensitisation processes can take place more easily, in other cases it is not needed.

In the desensitising process, unless the dog lights up in an aggressive reaction there is no reason for any type of correction to be applied, so contemplating a correction or aversive measure in the process, the dog must be lighting up or acting aggressively which provides us with the options below:

1. Drop the leash and let the dog go.

2. Drag the dog backwards and heighten defence drive through agitation

3. Correct the dog and take the drive away

I don't agree that either dropping the leash allowing the dog to attack or ramping it up in collar agitation is beneficial in the rehab process. Obviously no one will apply option 1, drag the dog back giving commands the dog ignores which is what most people do is bad IMHO especially if the other dog retreats allowing the aggressive dog to win will bring a large proportion of the desensitsing program undone. If in the situation during desensitisation that the dog does light up in aggression, correction is the only beneficial option to keep the work done intact.

that to me deoesn't really sound like desensitisation

To me desenstisation is acutally really habituation; which means learning that a certain 'something' is nothing to worry about.

I would think that by using punishments you can't achieve habituation.

I don't doubt that it is taching the dog something, but it is not teaching the dog that the presence of another dog is nothing to worry about.

That's a good point and ideally true, but given the road to rehabilitation is a long one, unless you can provide absolute controlled conditions at all times to train the dog reducing the reactive distances to raise the aggression threshhold, at some stage in reality I suppose is the term for it, the dog will light up in aggression, maybe an off leash dog running into the training area, a person with a dog popping up around a corner on a casual walk whatever, but it will happen, and dealing with that situation when it does I think is an important part of the process.

[corvus]

OFF TOPIC....SORRY!!!!! raineth, mita and Aidan, what sort of science degrees have you guys studied, or are studying. Even before this thread I was looking at going back to school (university actually !!) to study animal ethology to help me better understand the "inside going's on" of dogs. Should have done it moons ago but better late than never I reckon Where do you recommend I start looking and what steps are required? Any help would be greatly appreciated

I'm studying Behavioural Science with a Psych major. You might be better of with a Bachelor of Science, there is a certain course outline for Psych that is more relevant to human psychology.

If you go through a biology school I think it pays to check out who is on the teaching staff and what their research interests are, because it affects what you're taught. I somewhat haphazardly walked into the most behavioural ecology/bird researcher rich department of biological science in the country without realising that if I hadn't, I would not have had all the opportunities I did in undergrad to work on two of my all-consuming passions: animal behaviour and birds! What they teach you in behavioural ecology is an evolutionary basis of behaviour. I still find it a valuable approach even though I've crossed over into ethology now, but I struggle with the psychology side of things at times.