To follow up on a comment written on yesterday’s blog; what do we do when we are shaping one thing but our dogs get “stuck” offering something she already knows.
My first choice is to set things up in my favor to avoid this from ever happening. Two big things that will help are;
- Your body position. When I shape behaviours sometimes I stand, other times I sit in a chair or sit on the floor, or kneel on the floor — all of can become part of why the dog offers what she does. Knowing what your posture is and what you have shaped in a similar posture will help avoid these erroneous offerings from your dog.
- Manipulate your environment. Reduce the number of possible wrong choices for the dog by manipulating the environment as much as possible prior to starting your session. From yesterday’s example if you know your dog will want to offer targeting with a rear paw then start your session with the dog on a chair or in a box — just shaping head movement to start. Or use a target stick, patch, mat etc to help convey what you want.
The best solution is to have a great plan before you start a shaping session so you can avoid confusing or frustrating the dog in the midst of a session and I want to stress this is what I will do first. Often times if my dogs “obsesses” on a previously reinforced response I stop the session, put the dog away (with a big handful of cookies) and look at my plan — because clearly it was lacking!
Before you read on, re-read what I just wrote. My first choice is always to alter the environment I am working in. Knowing the reason my dog is doing what she is doing is because I did not plan my session well enough to begin. Why continue to frustrate my dog because I didn’t “set the stage” well enough for her to start!
So, lets suppose you have done everything correctly and still your dog gets “stuck” what are your options then?
- End the session and re-group. That should always be a strong option. Re-think, re-plan, re-do is not failure and it often times is the best thing to do.
- Wait the dog out? Just holding your position and waiting for something different I find causes way more frustration in a dog then is necessary. Think about it from the dogs point of view. She is continuing to offer you a response that previously you have reinforced. Perhaps (thinks the dog) you are just trying to build duration or increase the number of repetitions for the response she is offering. Waiting until the response is extinguished may lead to the dog just giving up on your session altogether. I am not saying this won’t work, just it wouldn’t be my first choice.
- Give a NRM to let the dog know you won’t reward that? Personally not my choice as I prefer not to interfere. As soon as you give the dog a verbal cue, you are controlling. However I am not apposed to letting the dog know I am not interested in what she is currently doing in a more subtle way.
- Communicate “try something different” to the dog with a subtle cue from your body language.
Lets say my session started with me shaping a response from a chair. If my dog got “stuck” on a response I could;
a) Just lean back in my chair and avert my eyes away from the dog — immediately leaning forward once my dog stops offering the undesired response. It is a “Game on – Game Off” body language thay my dogs completely understand.
b) Start breaking up my treats, taking my focus completely way from what the dog is doing (but watching peripherally in order to see a change I can reward).
c) Turn my back away from the dog entirely. I am less likely to do this one as it is too severe a punisher for a shaping session. However I will do this if the behaviour the dog is obsessing on is barking at me.
5. Another tack I may take when my dog gets stuck while shaping is to stop the reinforcement the dog is getting by offering the response. Depending upon the dog and the response, interrupting the response may be quick and effective. I won’t say anything to the dog, but I will just stop the pattern. I may do that with a collar grab (which is a game all of my dog’s know and carries tons of reinforcement). Or I may just move my body to prevent what the dog is attempting.
Whatever you do remember you goal is to create a fun, productive, effective and efficient training session. Yes all four criteria are important!
Today I am grateful for the great group of students coming in for this weekend’s Contact Training Workshop. Will be a good one for certain!
This post is so perfectly timed. I think you must have cameras in our homes so you know when we need certain information! Today is a rainy day in Florida, so no chance to play the new contact games I just created (I call it “Board-a-palooza”). So decided to shape some new tricks on the porch instead. I wanted to teach him to sit up (sit pretty), but kept getting Dance instead. Realizing I did not have a good plan, we moved into other things, and now reading your post I have some new ideas how to attack the issue!
That is some great information Susan. I have just discovered your blog and will scour it to uncover some more gems in here.
cheers,
Glen
Uf, finally I have time to sit and write something longer than one sentence.
So firstly thank you very much for this post, as it is direct answer to my question 🙂 It’s great to know your opinion 🙂
Secondly I can’t imagine having great plan for shaping session 😮 Yes, I can think about some general plan and yes, I can think about what I was shaping last time, where and what was the context (place at home, my position, objects), but that’s it. All the rest depends on the dog. Of course when I shape a lot with a particular dog we know each other and our ways of thinking and can predict better our responses 🙂 but sometimes dog does something great and totally different that I actually trying to get and I fell totally open to take it, to completely change my general plan in less than a second. For example not so long ago I was shaping with my older dog. My plan was to her jumping into the box and lying down. She jumped to the box very easily and then she offered me standing on her front feet 😮 Can you predict something like this? Plan it? I can’t! We were trying to get this behavior in completely different context and she wasn’t doing it yet! So I was totally surprised and of course I clicked her 🙂
Finally I think that “Communicate “try something different”” is a NRM simply from definition! Which way of saying “try something different” (verbal, non-verbal) is more subtle? I think it is very dog-dependent and way-of-teaching-the-meaning-dependent…
Thank you once more for VERY interesting discusion 🙂
All the best
Maria
I have one dog in particular who gets stuck in a response very easily. Waiting him out has never been a resonable answer because his frustration builds quickly and he becomes vocal.
I’ve tried a couple of the things you suggest, but you’ve also given me some ideas for varying what I’ve done before.
Thanks for the good ideas!
Thank you Susan for answering my question 🙂
Susan,
Since you are in the mood to talk shaping, could you perhaps say anything about not using a clicker? I have a heeler who will throw 10 “trash” behaviors at me the second she sees a clicker and ever since AiDT I only very rarely use a clicker with her for shaping but every time I have tried to explain how you shape by simply feeding immediately for the behavior approximation you wish to reinforce everyone thinks I am crazy.
Re: shaping and dog stuck on offering an unwanted behaviour.
I don’t see how readjusting your body “lauguage” in order to prompt the dog to offer an alternative behaviour could be classed as a NRM.
If dogs learn by “association”.Then the dog may have associated the behaviour it is stuck on with body lauguage the handler did ( and wasn’t aware of )during a previous shaping session.
So by changing your body language / position ,should change what the dog has been offering.
This has worked for me.I think we need to be aware of what we are doing at the time we are clicking.
According to what I have read about dog behaviour.Most of their communication is through body language with other dogs.
That explains why they are more responsive to our body language than to what we say verbally.
So I try to look at it from the dogs perspective and then gauge what understanding is taking place.Then work from there.
Reading yesterdays comment I was thinking my approach would be to use a target to convey that the objective is to go under the chair. But I also thought that perhaps that would be a lure? I saw you mention the use of a target to help convey what you want to the dog in todays comment. I’m a bit confused – would the use of a target be considered a lure, and if not, what is the difference? It’s always been a bit of grey area for me. Hoping you can help clarify.
This NRM blog topic and comment….wonderfully clarifying. and reinforcing whenever I find myself saying Yes! That’s something I’m already doing right! My understanding of a NRM is that, like a click, it is installed and is consistantly unvarying. The subtle movements Susan suggests(averting eyes, leaning back) don’t fit that description. They are just a noticible change, an interruption, I think, not a NRM.
Crystal, you and your dog have some exciting times ahead!! Because of her history with luring she is expecting you to show her what to do .. so when she discovers that she can make the good things happen through her own good choices, there will be no stopping her!! I am sure there will be lots of other suggestions but one simple one from me is to mark and reward something/anything at the start of the session before she even has a chance to lie down. Maybe she will just perk up her ears a bit, mark it then throw the food so that she has to run to get it (no chance to lie down!!), then when she is coming back you may see another perky ears (or other behaviour) that you can mark and reward. Fun, fast and so exciting when you see it dawn on her that she is “making it happen!!”
How do you get the dog to start offering behaviors? My dog offers a single behavior if I have treats and give no commands to her, and that is lay down. She will lay down, staring at me. The biggest change she will make is to lay her head down.
She does know a number of commands and tricks, but they have all been taught through luring, either with my hand or getting her to follow my gaze (she will target to something if I stare at it, etc).
When my dog gets stuck by offering down while shaping, I do one of two things:
– look away, watching with peripheral vision for any signs that he’s getting up. I click for that and toss the treat so he REALLY needs to move to get it
– click for any kind of movement (head movement, ear twitch), click that and toss the treat away from him. I used this one in the beginning.
Back when he didn’t yet know that it’s not a good idea to lie down again I used to also click as he was comming back to me and do a few more treat tosses this way to reinforce him for movement.
But he has been clicker trained from the start. With a dog who is used to luring I would probably do a few sessions just clicking any movement, without a plan of shaping something in particular. I think this is the least frustrating way to teach dogs that unprompted actions
can get them treats as well.
All these comments prompted me to wonder “what would Bob Bailey say?” I bet he wouldn’t waste time getting stuck on what he didn’t want.
Very fascinating …
I went back to “Bob Bailey gives a shout out” posted on 03/10/09. I remember him saying “When I wanted behavior I targeted, lured, modified the environment, essentially anything to get the behavior going”…
Looking forward to learning more…
Trudie
Very interesting about non-verbal NRM’s. My two dogs read my body language very well. We play It’s Yer Choice a lot and they have the game down pat. Occasionally I will “test” them when I am preparing supper by accidentally flipping some food on the floor. They don’t move a toenail.
On the other hand if I REALLY accidentally drop a crumb of food they are on it immediately. Since I obviously can’t progress this behaviour until I can control my unconscious body language I sit my dogs further away now.
I noticed today when training that I DO use verbal NRM’s more than I thought I did and even though my oopsies are said in a light tone there was no valid reason for using it.
Timing is everything and this I needed to read today. I went to a training session with Susan last year. I did a shaping session with Trudy who is not a quitter, my leg definitely had the marks of that. I was shaping a toe touch and got umpteen leg whaps instead. One way to learn how to cue and not cue correctly!!
This morning I was working on the down signal and then the sit. Trudy immediately drops down, great drop, awesome. Then just as quick gets up (its like lets do another down – reward me please, I will do this a million times for you). However I want her to hold the down so I can teach her the sit signal. I have changed nothing about me and expect my dog to learn that we are doing something new, hmmmm, smart dog, shame about the handler.
Note to self, crouch down next time I do the down signal with her and reward duration for staying in the down. I will gradually grow (stand up) and then I will teach the sit signal!!
Thanks again Susan!
I’ve thought of a couple of examples of our dogs impressive intelligence, things we take for granted.
How does your dog act when you put on your shoes? Does he act differently when it’s a pair of “leaving all day” work shoes vs “time to go out to the agility field” tennis shoes? Do you think he doesn’t understand the “cue?”
Does your dog bark when the average big-rig or heavy truck goes by? But how about the UPS truck? My dog barks her head off every time the UPS truck or the Fed-ex truck brakes squeal outside, but ignores the trash truck, neighbors small moving van and other big trucks. She knows that the UPS truck is a “cue” for a knock on the door, which means excited barking and jumping around.
If you receive bad news in an email or on the phone, does your dog come snuggling up beside you, or crouching around your feet? Even if you didn’t say a word?
If they read such subtle cues, maybe we aren’t capable of avoiding the NRM altogether? What does that mean – should we school our bodies even more carefully, or acknowledge the non-verbal NRM and try to use it more actively?
I would love to see a research study done on this subject, what a fascinating bunch of ideas.
Question:
You wrote:
“# Give a NRM to let the dog know you won’t reward that? Personally not my choice as I prefer not to interfere. As soon as you give the dog a verbal cue, you are controlling. However I am not apposed to letting the dog know I am not interested in what she is currently doing in a more subtle way.
# Communicate “try something different” to the dog with a subtle cue from your body language”
I feel like these are the same thing. The only difference is one is using your voice and the other is using body language. To me it seems that your body language would be less subtle than a verbal. My dogs certainly pick up on my body language faster than verbals. To me it seems that you are using your body language as a NRM. Using your body to let them know you aren’t interested in that behavior is interfering the same as a verbal. Maybe this is a technicality but I feel like they are the same thing just one is verbal and one is not. I try not to use either one of these but on occasion will. More often I use body language as my dogs respond much better.
Melissa, I had the same feeling. I suspect that the non-verbal is a technical “NRM” meaning there is no punishment, just information is passed along, and the verbal NRM is associated with more “punishment.” I mean punishment in it’s technical “negative reinforcement” sense, obviously I don’t think anyone here is physically punishing their dogs. But any word like “ah-ah” or “no” or “wrong” that is associated with not being allowed to go on, not getting a reward, or worse, frustration and impatience on our part, will be mildly punishing, and potentially discouraging.
It sounds like Susan has consciously or unconsciously conditioned a very mild NRM with no punishment involved – almost a “go on” or “try something else” command, which happens to be body language. After spending a lot of time recently reading about how exceptionally good dogs are at reading our conscious and unconscious cues, I’m curious if we even have the ability to avoid NRM? If our body language is an open book to our dogs, does it matter whether it’s a “No” or an unconscious body posture?
I love this discussion, thank you Susan!
GREAT one! I am working with a newly adopted rescue pup, she’s 10 months, and her personality is so different from my other dogs. I start with a shaping plan and she offers me something so fabulous I hate to miss it! So, I’ve only had her for about 6 weeks now, and I’ve been marking all the different behaviors, just in the hope that she will offer them again! Some things I have gotten on cue, like head down and roll, and “bounce back” and crawl (cause she offered it! and we’re into freestyle) and when she does crazy things like try to stand on her head practically, how can I not click it? She is a little octopus, and so if I go into it thinking I am going to shape a roll over, but she offers something else that is awesome, I can’t help myself! I grab it, or at least reinforce it, while the gettin’s good! And I am *thinking* (am I wrong?) that at this stage, where she is being so creative and funny with all her behaviors, it maybe is okay to just give all those different behaviors some little bit of history of being reinforced, because I hate to have her try backing up, or a headstand, and not mark/reinforce it, and the next time have her not try that.
So, the question is, is it okay to mark and reinforce general creativity? I don’t think that interferes with her understanding of progressive increments, but it is like a different game. Any thoughts on this? Maybe I should wear a different hat when playing the creative game?
Thinking through 1. the mechanics of how I will click and deliver treats, 2. what specific behaviors I will c/t, 3. what possible behaviors the dog might offer and 4. what I will do when the dog offers behaviors for which I am not looking were some of the most important things I took away from my utterly amazing experience at Advances in Dog Training in Cincinnati last year!
This is a very useful and timely reminder of those rules fr me, thank you!!
Thanks Susan for your blog! The discussion from around the world is extremely valuable. It allows all of us to reflect on our training. I appreciate the clarity of your training philosophy. You always have a clear picture in your mind of your goals and what are acceptable, valuable ways to achieve those goals.
Thank you so much for these posts – great information and very timely for us.
I have just started shaping with my foster dog. I rescued her from death row 3 weeks ago tomorrow. I have only known her a short time but it has already been a fun and informative time for us. We started with crate games and games from the recallers ecourse and she is doing super on responding to her new name, collar grab (slowly slowly), its your choice and nose touches. This week we also tried some free shaping. She is a smart cookie and has picked up the idea of shaping really quickly.
It has been an interesting experience though as we started with nothing – no words or commands – not even a name. I have been paying a great deal of attention to my verbal “patter” and in light of your posts will be careful not to undo our hard work by counter conditioning her name or introducing a NRM – she does not need any negative words in her life that is for sure.
Thanks again