(VICT) Re: Buckley being aggressive......
- From: "Ann Edie" <annedie@xxxxxxxxxxxx>
- To: <vi-clicker-trainers@xxxxxxxxxxxxx>
- Date: Sun, 28 Oct 2007 16:04:54 -0400
Hi, Ridge and Everyone,
I am a great supporter of Jesus Rosales-Ruiz's work, and think his shaping
procedure is very promising. However, I share Sandy's reservations and
concerns in Ridge and Buckley's situation. The CAT technique seems to
require very good control of the environment, not only during the actual
training sessions, but all the time until the desired changes in behavior
have become well established, and perhaps even after that. Since Buckley is
a working dog out in the public, it would be practically impossible to
control every interaction with people between the ages of 1 and 9 years.
Even if there are only two places where Buckley has shown aggressive
reactions to children to date, you never know when this behavior may
generalize to other places or some specific other place or other person.
His agressive response might be caused by the stress of his work, and this
is probably not going to go away simply by shaping a different response to
children in the two locations where it has appeared so far.
But my major concern is how you are going to find children between the ages
of 1 and 9 who can fulfill the role of "scary person" in this approach
safely and with the pinpoint timing required. I would expect most children
either to miss the signs of aggression that Buckley is exhibiting and to
keep on approaching beyond the point of safety, or to retreat while he is
still in full growl or threat mode and therefore reinforce the undesirable
behavior. I think this is a wonderful technique, but I think it must be
undertaken under the supervision of professionals well-versed in clicker
training and shaping.
Aggression of any kind in a service dog is a particularly sticky situation.
The classical conditioning (desensitization) approach takes a major
commitment of time and energy and requires not exposing the dog to the
stimulus in uncontroled circumstances. This can possibly be achieved with a
pet dog which one can keep at home and away from strangers during the months
of reconditioning. But it is impossible to continue using the animal as a
service animal in public during this long period of time. And there is not
a high percentage of success with classical conditioning in the absence of
continued management of the environment, as Sandy has demonstrated with her
Great Dane.
This is why almost universally if a service dog shows signs of aggression
toward humans, the dog must be retired from service work. Since by
definition, a blind handler will not always be able to see the aggression
stimulus coming far enough in advance to commence management measures, I
think this situation is particularly risky for us.
Just out of curiosity, Ridge, has Buckley been trained solely with the use
of positive reinforcement and the clicker, or have corrections been part of
the training toolbox? (It has been one of my interests to see if
correction-free training can reduce the stress experienced by working dogs
and therefore increase the success rate and decrease the burnout rate among
service dogs.)
I hope that you can find the professional training support you need to
address Buckley's issues with children so that you can continue working with
him safely and with confidence for many more years.
Best,
Ann
----- Original Message -----
From: "Meg Irizarry" <arjay@xxxxxxxx>
To: <vi-clicker-trainers@xxxxxxxxxxxxx>
Sent: Sunday, October 28, 2007 12:42 PM
Subject: (VICT) Re: Buckley being aggressive......
> Hi Ridge,
>
> I have been reading about the posts about Buckley and wanted to offer
> my opinion. First, let me say that those who have responded have
> offered some wonderful ideas and suggestions. Sandy is so on the
> money with mentioning management. Until several months ago, I'm not
> sure I would have had anything valuable to add or suggest. However,
> I became aware of a fairly new treatment for aggression within the
> last few months. It's called the CAT approach - Constructional
> Aggression Treatment. Basically, it is a shaping program as opposed
> to a standard program using desensitization. It has shown incredible
> results and compared to desensitization, results are quick.
>
> Here is the basic concept:
>
> First consider that aggression or any behavior is extremely specific
> to the environment. This explains why Buckley doesn't show
> aggression at church. You mention he is only reactive at two
> places. These then, are the places you need to work with him.
>
> You will not be using food at all, since food is not what he
> considers the reinforcer in those situations. I think it's really
> important for us humans to remember that we do not determine
> reinforcers. The dog does. By introducing food into the equation in
> a situation like yours, we tend to muddy up the environmental cues
> and triggers. Not to mention that many dogs are simply too aroused
> to even care about food during episodes of reactivity. It's also one
> less thing to worry about with regards to timing.
>
> So, consider what Buckley wants when he is growling or beginning to
> react. He wants more distance between him and the scary person.
> Having the scary thing retreat will become his reinforcer.
>
> Determine his trigger threshold. This is done the same as in
> desensitization. You'll work outside his threshold to begin
> with. You also want to try and recreate the environment as closely
> as possible to the real life circumstances in which he would normally
> react. Remember - behavior is absolutely specific to the environment!
>
> You then begin by having the scary person walk towards the dog until
> he begins to growl (if this is his first warning). The person should
> approach the dog normally. Sometimes in desensitization, we alter
> our approach by perhaps turning sideways and not looking at the
> dog. With the CAT approach, you do not alter your approach and
> should maintain eye contact with the dog. This doesn't mean trying
> to stare him down, but looking at him normally. When he begins to
> react, the SP (scary person) stops and stands still. The dog will
> normally continue to growl or bark. The SP waits it out. When the
> dog shows any sign of relaxing or stops growling or barking, even for
> a second, the SP retreats, which in essence is issuing the desired
> reinforcer.
>
> You then continue in this manner, shaping his
> behavior. Interestingly, there is also a point in the process where
> the dog "switches over" and decides that SP may not be the SP they
> once thought. You might observe some ambivalence in the dog's
> posture. You then can see the dog actually enjoying petting or at
> that point, treats.
>
> It's a fascinating, new approach to treating aggression, researched
> and developed extensively by Dr. Jesus Rosales-Ruiz and Kellie Snider
> at the University of North Texas.
>
> I hope this helps a little.
>
> Meg
>
>
> arjay@xxxxxxxx
>
>
>
Other related posts:
