erstellt am: 03 März 2009 16:31

So, hier ist ein etwas *hüstel* längerer Text... Ich darf ihn einstellen, Alex freut sich aber darauf, auch über die eventuelle Diskussion zu erfahren.

----------------------------------------------------------------
Hier findet Ihr das original-Post auf der Liste

I'm catching up on the last couple of weeks here on the list.  What a great series of posts.  Let me extend my own welcome to all the new members. 
....
 I won't try to jump in on that conversation, but will instead share some of the major take aways from the Clicker Expo.  Since everything is everything else, I'm sure there will be tie-ins with this current thread.


Micro Rhythms and Poisoned Cues
A couple of years back I read Malcolm Gladwell's "Tipping Point".  In it he described a situation where a psychologist was studying a four and a half second video clip of three people having a dinner time conversation. He watched the clip over and over again, convinced that there was something of importance there that he was missing.  He divided the clip up into tiny segments, each only 1/45th of a second long.  He watched these clips for months before he finally saw what he had been missing - micro rhythms.  As the three people talked, their movements, their speech patterns matched up and fell into subtle rhythmical patterns. One person would lean forward as another rocked back.  One would raise an eyebrow, another would respond with a tilt of the head.  The patterns were very real, very distinct, and yet so subtle they were all but invisible on a macro level.

I dog eared this section of the book, (pg. 80-83, The Tipping Point) not just because the concept of micro rhythms fascinated me, but because here was someone who looks at things the way I do!  I love spending months teasing apart some minute detail in a horse's balance looking for the underlying treasures hidden inside.

Why am I writing about this?  Every time I listen to Jesus Rosales-Ruiz's lecture on poisoned cues, I am reminded of the microrhythm story.  I am convinced that there is more there than I am immediately seeing.  And each time I revisit this lecture, I see details and connections I missed before.  The significance of this work becomes even more profound.  I've lost track of the number of times I've heard this talk, at least five or six, and I still consider it the highlight of each Expo.

This year I also was able to see Morton Egtvedt and Cecilie Koste's presentations on "The ABCs of Obedience Training" and "Backchaining".  I saw the post from Inge that they will be giving a clinic in Belgium on backchaining.  For those of you in Europe, their clinic will be well worth attending.   I thoroughly enjoyed their presentations.

Their talks tied in well with the poisoned cue presentation and with my own talk on training mechanics - or as Jesus has suggested I should call it - loopy training. 

The Kay Lawrence Connection
To explain how all these different pieces of the puzzle fit together, let me begin by writing what will hopefully be a brief summary of my visit last December to the University of North Texas.  I spent four days with Jesus and his Orca graduate students.  During the course of our conversations, Jesus shared with me an article on food delivery by Kay Lawrence (http://www.learningaboutdogs.com/html/tossing_matters.html )  and also a youtube video clip she has posted on her web site. The clip shows an elegant bit of training.  Here's the link so you can watch it:  http://www.youtube.com/watch?v=EP7Soem31qg)

The handler is teaching her dog to move energetically to a mat.  The session begins with the handler sitting in a chair and her dog moving freely around her in Kay's large training barn.  The dog knows to approach the handler.  As he moves towards her, she clicks and tosses a treat out behind the dog.  The dog moves off to get his treat and then immediately turns back towards the handler.  C/T as he approaches her.

This process is repeated several times.  Then a mat is put down a couple of feet out in front of the handler's chair.   As the dog trots towards her, eagerly expecting a click and a treat, his front feet pass over the mat.  The handler clicks at this moment and tosses the treat out behind the dog.  He collects his treat and then turns back towards the mat.  His front feet again land on the mat, and he gets clicked and reinforced.

When this pattern has been well established, the handler begins tossing the treat out to the side towards three o'clock, nine o'clock and finally six o'clock.  What she is looking for from the dog is a positive deviation from a direct line back to her so that he lands instead on the mat. The dog clearly understands that he is to stand on the mat and the energy he shows during this exercise is impressive.   It's a fun shaping strategy.  But that's not what was capturing our attention.  What Jesus was pointing out, what was so clear in the video, was the training loop.  The dog performed the behavior, got clicked, collected his treat and immediately returned to the task at hand.  There was no wasted energy, no time lost sniffing around the floor for extra bits of hot dog, no wandering off to other parts of the room.  It was a beautiful cycle.

Jesus commented that he sees the same cycles in my work.  When the mechanics are working, when we are splitting behaviors and have defined well the clickable moment, the horses are on tight behavior loops.  We see the same pattern as Kay's dog.  The horse gets his treat and returns immediately to the clickable task.

Treat Delivery
What interested both of us were the points where this cycle breaks down and the reasons for it.  One of the main places is in the treat delivery.  I was particularly interested in this because the mechanics of treat delivery was going to be a major topic for one of my Clicker Expo presentations.  At the Expos I've seen some amazing training, but I've also watched dogs who were struggling with even the most basic exercises.  One of the main reasons the training falls apart lies in the food delivery. 

In one scenario that I see a lot, the dog gets clicked and the handler tosses the treat(s) on the ground.  The dog is not really clear about the whys and wherefores of treats.  He hears the click, but doesn't watch to see where the treat lands.  By the time the owner has pointed it out to him, the treat has lost any connection to the behavior the handler was trying to reinforce. 

In another scenario the handler is not all that practiced in treat delivery, and so she's just as likely to toss out five or six tidbits as a single treat.  The dog finds the first goody and keeps on looking.  His mind has switched off from thinking about working for the click to hunting over the ground for missed goodies. All the time spent scavenging  means the rates of reinforcement have dropped down into the basement.  There's no elegant efficiency of learning.  There's just a food-oriented dog hunting over the carpet for goodies while a frustrated owner looks on. 

With the horses I put a huge emphasis on training basic mechanics.  Think about all the time we spend at clinics fussing the details of the foundation lessons; all the hours and hours of DVD lessons that I've put together; all the posts I've written on basics, basics, and more basics; all the time I personally spend going over basics with my own horses.  Watching these dogs struggling confirms for me why this is so important.  I return again and again to Bob Bailey's statement: "Training is a mechanical skill.  Don't let mechanics get in the way of good training."

An Accidental Result
During my visit with Jesus, his graduate students shared with me their current research projects.  One of these included a study on jackpots.  During the course of the study the method of food delivery was switched from dropping the food by hand into a bowl to delivering it via a PVC pipe to the bowl.  The dog was being clicked for touching a stationary target.  Prior to the change in food delivery, he would hear the click, and head straight to the food bowl.  As soon as he had his treat, he returned to the target.  It was a perfect training loop. 

When the delivery method changed, the dog's behavior fell apart.  He heard his click, turned, but did not go immediately to his food bowl.   He didn't understand that the treat was being delivered through the PVC pipe.  He was looking for the familiar hand movements that were a precursor to the treat landing in the food bowl.  Without those he became confused.  As far as he was concerned the promise implied in the click had been broken.  What had been a consistent, reliable training loop fell apart.  The dog wandered about the training area and finally left entirely.  He was certainly telling his owner what he thought of the experimental set-up!

So this raised some interesting questions about the click.  What is it really?  We've all been taught to think of it as an event marker, a "yes" answer signal.  So we have cues which serve as green lights for a specific behavior and we have the click, the "yes" answer signal, that confirms for the animal that he got it right.  In other words, in this model cues and clicks serve two very different functions. 

I always think of a sentence: The cue leads to behavior leads to a click which leads to a treat. 

The cue is the green light for the behavior.  The click is the "yes" answer that tells the animal he has been successful and a treat is coming.  That certainly looks as though they are serving two different functions, but are they?

Cues, Commands, and Clicks.
We're coming to the pieces that tie an understanding of poisoned cues, training loops, and back chaining together.

We've understood for a long time that cues are green lights.  Commands are something else.  Commands are taught with corrections.  Commands have a "do it or else" component embedded in them.  Cues are different.  Cues are taught with positive reinforcement without the corrections backing them up. 

Here's the significance of this.  You can shape with a positively taught cue because the effect of cues works in two directions.  Hang on.  This is the part that can make your head spin a bit when you first encounter it.  It's easy to see how a cue works running forward.  You say "sit".  The dog sits.  Click and treat.

The cue is working forward triggering the sit behavior.  But it is also having an effect on what came before the cue.

If the word "sit" consistently, reliably leads to hot dogs, your dog is going to want to do things that get you to say sit.  If he has learned that jumping up on you gets you to say "sit" and then to present him with goodies, he'll be even more eager to jump up on you than he was before.  On the other hand, if going to his mat gets you to say sit, he'll be eager to head straight for the mat.  You are creating chains of linked behaviors.  In the first example you are inadvertently strengthening a behavior you don't want by linking it to the sit behavior.  And in the second you are correcting the problem, not by directly punishing the jumping behavior, but by training an incompatible behavior.  That's a good shaping strategy.  Add in a cue such as the ringing of the doorbell means go to your mat, and you have an elegant training sequence.  The ringing of the doorbell sends the dog to his mat, where he eagerly sits because he knows that will produce a click and treat.

It's a nice bit of training, but what you really want to understand from this example is we can shape behavior with cues.  Poison those cues by creating uncertainty in the outcome, and the shaping falls apart.  That's what Jesus shows so clearly in his poisoned cue lecture.  His video is some of the clearest I've seen on the use of cues to shape behavior.

und weiter gehts...
-------------------------------------

What is a Click?
So again what is a click?  It's a conditioned reinforcer.  Yes, yes, that's a fine answer, but what does it mean in light of this discussion?  Is a click a "yes" answer signal?  Yes.  But so is a well built cue when it is used in a chain.  In our canine example, "sit" acted like a "yes" answer for the going to the mat behavior.

It also acted as a green light for the "sit" behavior, so the cue acted in two directions affecting the two behaviors it was sandwiched between.

Is that the case with the click? The click is a yes answer for the behavior that immediately precedes it.  Is the click also a green light for a behavior that follows?  Yes.  Absolutely.  The click cues the animal to switch into treat gathering behavior.

What this means for our training is it's all backchaining.  So Morton and Cecilie's presentations become universally interesting.  We don't need to be interested in obedience training to find relevancy in their work.  We all need to understand backchaining: what it is, how it works, how to structure a good backchaining lesson plan. 

And we need to pay particular attention to the treat delivery at the very beginning of the process.  Here's a basic question about the most basic of basic skills: does the animal know how to get his food?

Backchaining and Food Delivery
That's the question we need to begin with.  Does your animal know how to get food?  The answer for the dog with the feeding tube was no.  He understood food delivery when the food was dropped by hand into his bowl, but not when it was delivered down a chute.  The answer for some of the dogs I've watched at the Expos is also a definite no.  For these dogs, when food was tossed on the floor, finding it was very much a hit or miss proposition. 

For the dog on Kay Lawrence's video, the answer to this question was a definite yes.  The click was a signal to him that treats were coming.  As soon as he heard the click, he broke off what he was doing and paid close attention to the handler.  So the click was really a signal to watch the handler's hand motion.  That told him where the treat was going to land. 

The jackpot-experiment dog knew this cue as well.  Click means watch the handler's hand, but with the feeding chute, the hand signal became like a broken traffic light.  If you are used to red, yellow and green lights working in a set order and suddenly a blue light comes on, what are you going to do?  What does blue mean?  Slow down? Speed up? Stop? Make up your own rules?  You don't know.  You may end up driving your car up on the sidewalk in a frozen panic.  That would be akin to the dog quitting the game when his owner changed his food-is-coming cues.

When I first started experimenting with clicker training, I was working with horses who were all used to being hand fed.  None of them needed to be taught the reach into the pocket action meant goodies were coming.  So it was easy to pair the click with that action.  And I found that I didn't need to charge the clicker.  If I chose a simple enough behavior I could go straight to shaping with the clicker.  So it was behavior => click => treat right from the start.  Lumping.  Oh well.  It worked and continues to work, but there are clearly situations where splitting would be better.

If it's all back chaining, I am left to wonder about the horses who struggle with the beginning steps of the process. Are they the ones who aren't used to being hand fed? Is that why behavior => click => treat is not as quickly established?  If these horses aren't used to being hand fed, the links in the chain won't be as strong and the behavior will break down more easily.  In this scenario, it's easy for cues to become poisoned.   We certainly see in some horses the equine version of the "punir" dog in Jesus's talk.  This is the situation where the dog was given a correction if he did not respond promptly enough to the cue. 

The videos Jesus shows during his presentation illustrate powerfully the effect of what looks like a very mild correction.  When the dog is presented with the unpoisoned "ven" cue, his tail is wagging. He's clearly an energetic, happy dog.  When the cue is used to shape behavior, his responses are prompt.  There's a clean, efficient training loop very much in evidence.

But when the poisoned "punir" cue is used instead, the dog's whole demeanor changes. His posture drops.  His tail droops.  He wanders around the room, clearly avoiding the behavior which will prompt the "punir" cue even though there has been no correction associated with it for many trials and the cue leads to a click and a treat.

The video clips show so many things.  They show that cues can indeed be used to shape behavior.  And they show that cues can be poisoned, and a poisoned cue does not work effectively in a shaping process. 

(Just as an aside:  I'm working with Jesus now on a Poisoned Cue DVD so you'll be able to see the video clips I'm referring to.  Barring any more unforeseen technical difficulties I should have it ready sometime in February.  I say any more unforeseen technical difficulties because I have just had a month of computer glitches, but you don't want to hear about that! Instead let's loop the discussion back to a click is a cue.)

A Click is a Cue.
A click is a cue.  What does that mean for us?  Well, we know that we want to get the behavior before we attach a cue to it.  So do we have treat gathering behavior?  With many horses the answer is yes.  They understand the hand reaching into a pocket means treats are coming.  They know how to take treats safely from a person's hand.   That doesn't mean there may not be some mugging issues with these horses.  It simply says that they understand the mechanics of hand feeding.

But what about the horses who have been denied treats? They often struggle through the beginning steps of clicker training.  Horses that have been fed treats are generally not particularly mouthy, especially once the rules around the food have been explained.  But the horses who have not been hand fed seem to take much longer to settle down.  They're the ones who get overly excited by the food and who struggle the most to learn basic food manners.

This makes me think that before we start clicking, we should be asking if the horse has ever been hand fed treats.  If the answer is no, before we introduce the click, we should be hand feeding treats.  That's different from charging the clicker where you click - feed, click - feed until a connection is made.  This is even more basic.  This is feed the way you're going to in a clicker session, repeat, feed the way you're going to in a clicker session, repeat, until the horse is actively engaged in the feeding process.   Once we have the behavior, we can attach a cue to it: the click.

What all this means is the science confirms good training practices, things we are already doing.  If we want to use the "click" cue as a shaping tool, we need to be certain that we don't poison it with uncertainty.  If you click, you treat.  You don't click and then correct the horse for grabby behavior.  In those first clicker lessons clean, efficient, prompt food delivery matters.  In other words mechanics matter.  And if the horse gets grabby, you still find a way to deliver the food in a clean, prompt, efficient manner.  The click means a treat is coming.  It doesn't mean food is coming or you may get whacked in the face.

Once the click is understood as a cue, we can use it as a shaping tool.  Hold the target up, the horse touches it, click, he goes into treat gathering mode.  It's all backchaining.  We teach the last thing first.

Now this kind of explanation may be too much for a novice clicker trainer.  With the great majority of horses it works perfectly well to introduce the click and treat via targeting.  The connections are quickly made, and there's smooth sailing forward.  But I do think that in some cases there is value, not in charging the clicker, but in prepping the food delivery.

The Effect of Treat Delivery on Training Loops.
That's certainly what I see with many of the dogs I watch.  They really don't know how to find their treats which results in an immediate breakdown of behavior.  Food delivery matters because awkward food delivery breaks down the training loop. This was the central theme of my presentation on training mechanics at the Expo. 

We've all seen the phrase:

                                                           behavior => click => reinforcement

This is the core of clicker training.  An operant behavior is determined by it's consequences. 

Well, we really should be thinking of this not as a single phrase but as a training loop.

                                                           behavior => click => reinforcement
                                                                                    =>
                                                           behavior => click => reinforcement
                                                                                    =>
                                                           behavior => click => reinforcement
                                                                                    =>

If food delivery is awkward and clumsy, if the animal can not find his treat promptly and then return to the task, the loop falls apart.  Rates of reinforcement plummet.  The animal gets discouraged and the training environment becomes poisoned for both the animal and the trainer.

And that's another element in Jesus' talk.  The study raised the question: what gets poisoned?  Certainly the "punir" cue was poisoned for the dog, but was that all?  When the "punir" cue was given, the dog was wearing a leash.  Was the leash also poisoned?  Was the room poisoned? 

This question matters to our horses.  What becomes poisoned when a cue leads to uncertainty?  When the  outcome can be either a pleasant, desirable response or an aversive, what is the effect on the horse?  For example, is it just the cue to canter that becomes poisoned?  Or does the arena become poisoned?  Does the sight of a saddle become poisoned?  And what do we do if we discover we have poisoned cues?

The simple answer to this last question is change the cue.  That's easy if it is an isolated cue that's been poisoned.  If "punir" is poisoned, just don't use "punir".  Retrain the behavior using another cue.  But what happens if it is our tools or our work environment that have been poisoned?  What do we do if we have a poisoned lead rope? A poisoned saddle?  A toxic relationship?

That's not so easy, but the answer may lie in part in back chaining.  How far back do you need to go to get a simple, clean "yes" answer? A "yes" answer that shows no poisoned cue effects? 

This brings us full circle back to the very beginning of my post.  I've been skimming through the Mirko snapping  thread.  There's quite a bit there and I haven't read all the posts in detail.  Katie has fielded the issue of mechanics so I don't have to step in there.  I'll just mention quickly that the foundation "body neutral" position out of which all the leading develops is taught in the "grown-ups are talking, please don't interrupt lesson" and that's illustrated well in the "Lesson 1: Getting Started with the Clicker" DVD.  When I converted that lesson from the old VHS format to a DVD, I added an hour of new material which updates the foundation lessons.  Included in that is the "Grown-ups are Talking".  This is a key lesson because the slide down of the lead evolves out of the body neutral position of "grown-ups". Rope handling is itself a loop.  It's body neutral => slide into t'ai chi wall => release back to body neutral => slide into t'ai chi wall.

Unpoisoning Poisoned Cues
I sense a tangent developing so I'm going to break off there before I head off into a long mechanical skills discussion.  Let's loop back instead to the concepts that lie beneath mechanics.  Mirko is snapping which makes training not fun.  Mechanics matter.  Details matter, but which details?  What do you need to change?  What part of the chain is broken?  Which cues are not understood?  Which cues are poisoned?  How far back in the chain do we need to go to find consistent "yes" answers?

That's what you need to look at to solve the problem.  I've written about this with Robin.  During the period I was working on the riding book, I poisoned many cues with Robin.  I won't go into the details of the whys and wherefores.  The short answer is he needed me to spend more time with him and time was a very scarce commodity during that period.  So our relationship became fractured, and I was living a poisoned cue scenario. My answer for Robin was "Robin's "yes" game".  All ambiguity was removed.  I put Robin on such a high rate of reinforcement that he could do no wrong.  I loved the result, but at the time, I thought, I can't teach this strategy.  If I share this with others, people will make a total hash out of it, so instead of teaching Robin's "yes" game, I developed the microshaping strategy.   This gives people a way to increase dramatically the rates of reinforcement and to step around many of the emotional-response issues that interfere with successful training.

But to resolve poisoned cues fully it may be that we have to add some other things to this lesson. Where possible, we need to change the cues.  Where possible, we need to change our tools and our environment.  In other words, use a different halter, a different kind of lead.  Train in a different arena. 

A third strategy is look at your back chaining.  Which links are poisoned?  How far back do you need to go to find links that are solid?  Do you need to play a version of Robin's "yes" game to build unambiguous, unpoisoned links? 

If poisoned cues are an issue, you also need to look globally at your training plans, your mechanical skills, and your horse's body mechanics so you don't end up repoisoning your newly built cues.  What happened that resulted in poisoned cues in the first place?  Was it simply a matter of mechanics and novice handling?  We're all a bit clumsy when we are first learning something new.  Horses are amazingly forgiving and this sort of thing usually does not have very long lasting effects.  But mechanics do matter. All the training games we play at clinics show us so clearly what it feels like to be on the horse's end of a lead.  It's startling how unpleasant it can be when the handler is out of balance or tense. And it's equally startling how wonderful the experience can be when things are right.

Other things to look at: were you still being a "horse trainer" while you were also trying to be a clicker trainer?  Did you feel you needed to correct unwanted behavior?  So, yes, you were clicking and treating favorable responses, but were you also punishing unwanted behavior? That's the scenario that leads directly to poisoned cues.  You were placing your horse in a "lady or a tiger" scenario.  No matter how many times he got the right answer; no matter how many times he opened the proverbial door and found a "lady" inside, the possibility always remained that it could have been a "tiger".  And that's the problem with poisoned cues.  Every time you reinforce the favorable outcome, you are also reinforcing the possibility that the bad thing will happen.  They become tied together like two vines strangling one another. 

So clicker training is so much more that simply clicking and reinforcing favorable outcomes.  To be maximally effective we need to understand deeply that corrections do not work.  We can not mix corrections and teaching and expect good results.  If this doesn't make sense to you, you very much need to attend the poisoned cue presentation.  Watching the video clips of the dogs will explain this far better than I can.

Here's another question: can the horse do what you are asking?  Are you putting the horse inadvertently into a position that causes physical discomfort?  Is that what is poisoning the cue?  For example, originally was a horse snapping at his lead simply his way of saying: "I can't. It hurts."

Do you need to microshape your requests so the horse can find a comfortable way to perform the task? Microshaping is a very powerful way to circumvent poisoned cues.   It ensures that the horse stays within his capabilities even while he is able to give you progressively more difficult work.

Here's how I think of it.  I have a twist down my right side which makes knee bends problematic.  Ask me to do a deep knee bend and I can't go down very far before my knees start protesting.  But when I was working with an Alexander practitioner, I could do super deep knee bends - provided we took the time to prep them.  If we microshaped good alignment, I was fine.  This experience gave me a great appreciation for what our horses go through. At times we are unknowingly asking for things that cause great discomfort.  It's easy to poison a whole class of requests simply because we have not recognized those first signals from our horses that tell us things are getting hard.

The Bottom Line
Training is a huge puzzle.  There are the questions of body mechanics, and our own mechanical skills, and then there's the whole complex science of operant conditioning to sort through.  But really training is not as daunting as this long, rambling post makes it seem.  Here's what all boils down to:

When you encounter a problem in your training, recognize that the solution will not be found in that outermost layer. 

Instead find a behavior, no matter how simple it may seem, where you can get a clean, consistent "yes" answer.  Focus your attention on that piece of the chain.

Remember that the longer you stay with an exercise, the more good things you will see that it will give you.

As you focus on that simple, unpoisoned "yes" response, you'll see the next step beginning to emerge.   In good shaping, you know you can go after the next detail because it is already occurring.   It will be a tiny step, nothing big.  So it will be easy to ask yourself what skills do I need to make this step maximally successful? 

It's a tiny step, so you aren't adding in huge new skill sets.  You're just adding in one new little piece.  Maybe it's as simple as shifting your balance towards your horse, or sliding your hand up to touch the lead.  It's a little piece for a little step to get back a solid "yes" answer.  Put together enough "yes" answers, and you will have solid, unpoisoned, unbreakable performance.

How will you recognize a system that is not poisoned?  You will have clean, clear, efficient training loops.  So in your training, look for those loops.  Learn to recognize them. Learn to grow them. 

People keep saying I need to come up with my own name for the training I do.  Should we call it loopy training?   Hmm, I don't think so.  Memorable, but perhaps not quite right!  But loops are in there.  Like microrythyms, the more you look for them, the more you'll find them.

Have fun!

Alexandra Kurland
theclickercenter.com

Carmen

04 März 2009 11:08   Loopytraining 1 (Mail von Alex K.)
Vielen Dank für deine Mühe.   Das muss ich mir in einer ruhigen Minute zu Gemüte führen.   

Muriel

04 März 2009 12:24   Loopytraining 1 (Mail von Alex K.)
war keine Mühe, war ja nur kopiert   
aber ich muss es auch noch mal lesen. Und sicher noch ein paarmal, bis man alles versteht, was da drin steckt.   

lindalotze

05 März 2009 8:12   Loopytraining 1 (Mail von Alex K.)
Und wenn man den Text in ein oder zwei Jahren nochmal liest, spuckt er "neue Wahrheiten" aus. Gerade bei Alex'Schrieben geht es mir oft so, dass ich das Gefühl hab, einen Zwiebel zu schälen und immer tiefere Lagen auszugraben - nur der Kern ist noch fern    

Carmen

05 März 2009 11:22   Loopytraining 1 (Mail von Alex K.)
Dann muss ich ihn in ein oder zwei Jahren nochmal lesen. Im Augenblick bin ich mir nicht sicher, was sie damit sagen will...   

Muriel

05 März 2009 12:04   Loopytraining 1 (Mail von Alex K.)
Hast Du Dich denn schon mal näher mit A. Kurland befasst? Wenn nicht, ist der Text m.E. kaum zu verstehen bzw ihre dahinterliegende Intention oder Ausgangsbasis.   

Carmen

05 März 2009 17:55   Loopytraining 1 (Mail von Alex K.)
Oberflächlich. Ein bisschen auf der HP gestöbert und "Clickertraining für Pferde" gelesen. Mehr bisher nicht.   

lindalotze

06 März 2009 7:56   Loopytraining 1 (Mail von Alex K.)
Carmen, mach Dir keinen Druck. Manches dieser tieferen Lagen erfasst man eher intuitiv und denkt "ja, genau, das ist so in den guten Trainingssessions", als dass man da eine hochintellektuelle Geschichte draus machen müsste. Das kommt von allein, wenn man ein paar Trainingssessions auf dem Buckel hat. 

Beispiel:
"Loopy Training" nutzt man doch allgemein beim Reiten: man galoppiert z. B. nicht nur ein Mal, sondern fünf Mal an (und anfangs möglichst immer gleich, nicht in x Varianten), um den Bewegungsablauf zu festigen.

Und dieses Video, auf das sie verweist http://www.youtube.com/watch?v=EP7Soem31qg, zeigt einfach nur super-gutes Training. Wenn Du versuchst, einen ähnlichen Rhythmus mit deinen Pferden hinzukriegen, machst Du sicher nicht viel falsch.

Auch die Seite von Kay Lawrence www.learningaboutdogs.com/html/tossing_matters.html lohnt sich zu lesen. Ich bemüh mich seither wirklich IMMER um geschicktes und schnelles Füttern. Früher hab ich mir teilweise schon "Gruschdelzeit" erlaubt und mühselig Futter aus tiefen Taschen gefischt. Das ist ja nett für gelegentliche Bestärker bei gut gekonnten Verhalten, aber fürs richtig aktive Lernen ziemlich Panne. Da brauch ich mich nicht wundern, wenn das Pferd nebenbei Zeit hat, -zig andere (weniger erwünschte) Verhalten zu entwickeln.

Die Micro-Rhytms sehe ich am Ehesten bei Pferden untereinander, aber auch bei Pferd-Trainer-Kontakten. Wie ich sie am geschicktesten nutzbar mache, davon hab ich allerdings noch keine genaue Vorstellung? Ihr vielleicht?   

Carmen

06 März 2009 17:09   Loopytraining 1 (Mail von Alex K.)
@lindalotze
Ich mache mir keinen Druck. Ich wollte nur versuchen nachvollziehen, was ein sturmerprobter Clickerer dem obigen Text entnehmen kann. Aber wenn ich nicht sehe, was ihr seht, ist das auch kein Beinbruch. Ich bin sowieso eher der Typ "learning by doing".   

eboja

22 Sep 2009 20:25   Loopytraining 1 (Mail von Alex K.)
Hallo zusammen,

jetzt habe ich mich endlich auch mal durch den Text gearbeitet, nachdem ich immer mehr Fragezeichen über dem Kopf hatte, was denn dieses ominöse "Loopytraining" nun eigentlich ist. Ein toller Text!   

Ich glaube auch, den Großteil verstanden zu haben, vielleicht auch, weil ich Kay Laurence schon länger "kenne" (leider nur in Form von Büchern und DVDs). Sehr interessant war für mich der Gedanke des "poisoned click"s. Ich achte schon darauf, das Futter schnell im Hund (ist halt bisher mein Trainingsobjekt) zu haben, aber ab und zu werfe ich es auch. In einigen Situationen wird das Futter damit zur noch tolleren Belohnung, aber ab und zu habe ich dabei dann auch ein "Stocken" bemerkt, eben weil er wohl erwartet hatte, das Futter in anderer Form (also einfach gereicht) zu bekommen.
Schön, auf das Problem mal mit der "Nase drauf gestossen" zu werden!!!

Was mir aber überhaupt nicht klar geworden ist, ist, was sie mit "microrhythms" meint. Kann mir das jemand in einfachen Worten und/oder anhand eines Beispiels erklären? Ihr Beispiel scheint für mich einfach menschliche Körpersprache zu sein, die ich natürlich auch in Bezug auf ein Tier anwenden könnte. Aber das ist vermutlich nicht, was sie damit sagen will...

Viele Grüße,
Esther   

lindalotze

24 Sep 2009 7:39   Loopytraining 1 (Mail von Alex K.)
Esther, ganz 100%ig kann ich es auch nicht erklären. Ich glaub, das ist dieser gespürte Fluss im Zusammensein, den man mit guten Freunden und netten Pferden halt so hat. Das flutscht einfach, ohne dass man aufpassen muss, den anderen nicht zu unterbrechen oder nicht zusammenzustoßen.

Als Beispiel fiele mir ein: Du sagst einen Satz, Dein Gegenüber nickt ganz leise. - Du weißt jetzt, dass er Dir zuhört. Du greifst zum Glas, Dein Gegenüber reagiert mit einer bestimmten Augenbewegung, blinzelt. - Du weißt, dass alles okay ist mit ihm. Die Reaktion ist Dir vertraut, entweder vom Gegenüber selbst oder von anderen Menschen.

Wenn Du jemanden echt irritieren willst im Gespräch, dann nick mal an den falschen Stellen, blinzle künstlich oder kaum, schau zur Decke oder in Deinen Schoß, neig Dich an unpassenden Stellen etwas vor, leg eine geballte Faust auf den Tisch, sitze extrem breitbeinig usw. - Stell Dich aber darauf ein, dass dieser Mensch anschließend kein sonderlich sympathisches Bild von Dir hat, sondern Dich vielleicht sogar gestört finden wird 
Es ist nicht nur, dass Du eine irritierende Körpersprache zeigst, sondern auch, dass die ganze Interaktion gestört wird. Dein Gegenüber weiß am Ende nicht mehr, was Du als nächstes tun wirst. Man kriegt da regelrecht Angst und emfpindet solche Gesprächspartner rasch als unangenehm, obwohl sie einen weder ärgern noch bedrohen noch sonstwas tun.

Etwas Ähnliches gibt es sicher auch bei Pferden bzw. zwischen Pferd und Mensch. Mit manchen versteht man sich auf Anhieb, obwohl sie gar nichts verstehen (das sind die mit den uns verständlichen microrhythms), und andere findet man doof, obwohl sie alles machen und es keine "echten" Verständigugnsschwierigkeiten gab.

Mehr als obiges Blabla kann ich jetzt leider auch nicht liefern. Ich wüsste aber auch nicht, ob es eine echte Definition für microrhytms gibt?   

eboja

24 Sep 2009 8:49   Loopytraining 1 (Mail von Alex K.)
Hallo Linda,

danke für die Erklärung, ich glaube, ein bißchen klarer ist es mir geworden. Mit Pferd habe ich da bisher nicht ausreichend drauf geachtet, aber beim Hund glaube ich das zu kennen: das Gefühl, das Tier versteht einen ohne Worte (oder auch das "der versteht jedes Wort"-Phänomen, wobei da auch sehr viel Gelerntes dabei ist), aber auch das Gefühl, selbst zu wissen, was der Hund gerade "denkt" und was er als nächstes tun wird (auch da: viel erlernt).

So ein bißchen kenne ich das schon auch vom Pferd, das Gefühl, dass es halt "passt", dass man weiß, ob das Pferd das Putzen gerade angenehm oder unangenehm findet und man weiß, ob man jetzt an dieser anderen Stelle weiter "kratzen" soll, oder nicht.

So ganz habe ich es damit wahrscheinlich auch nicht "gegriffen", aber ich denke, so ungefähr in die Richtung geht es, oder?

Viele Grüße,
Esther   

Muriel

27 Sep 2009 20:03   Loopytraining 1 (Mail von Alex K.)
Oh Linda, vielen Dank für deine Gedanken zu Microrhythms. ich konnte mir nämlich da auch nichts drunter vorstellen.
Hast Du das genannte Buch gelesen?

lg Heike   

lindalotze

28 Sep 2009 6:16   Loopytraining 1 (Mail von Alex K.)
Nö, hab ich nicht. Aber wenn der Typ den Film über hundert Mal ansehen musste, muss es etwas ziemlich Kleines sein...   

Tinka

22 Okt 2009 22:07   Loopytraining 1 (Mail von Alex K.)
Ui,

schwere Kost vor dem schlafen gehen. Mein Englisch ist erstaunlich gut hängen geblieben, sodass ich den Text ohne Wörterbuch lesen konnte. Aber ich hab dennoch das Gefühl so einiges nicht verstanden zu haben. Morgen werd ich mich nochmal dran begeben.