teaching the ‘hold’ without mouthing
January 29th, 2012 § 2 Comments
Over the past 2 weeks, whilst Grey hasn’t been able to go out on walks, I’ve been having 2 clicker sessions a day with the dogs’ regular meals.
I’ve been working on teaching the ‘hold-without-mouthing’ using the dumbbell. Once I have it on the dumbbell, I’ll generalise it to all sorts of articles.
In gundog work, the dogs bring back the game and immediately deliver it – no hanging around me with it in their mouths. They don’t sit to present it, either.
When the dogs are running and holding something, they are unlikely to mouth. I’m not exactly sure why, but I think the physical movement of running keeps the dog preoccupied and keeps their mind off rolling the thing in their mouth around.
Game being (usually) relatively heavy also discourages mouthing, since a dog which mouthed would drop it. Which is all to say that although the dogs have a fairly bullet-proof hold (in terms of never dropping) I haven’t trained no-mouthing – since they don’t mouth in gundog work. So this is something new, for me.
For working trials (and for obedience), dogs have to retrieve things and sit to present them. Then hold them (still) for some many seconds until the judge says the handler can take the article. Mouthing or rolling the article around is penalised.
The crucial thing is the timing: Getting the click when the dog’s mouth is still and not moving. This sounds simple, but mouthing can be very small, so watching very closely is needed. (This is even harder with a hairy beard hiding the mouth.) It is easy for dogs to mouth just at the moment you’re clicking, so you end up inadvertently clicking the mouthing.
So the methods all involve manipulating things in order to get good timing of the click.
Here are the things I’ve been trying:
- Distraction. When the dog is holding the object, show the dog a treat (without clicking). If you’ve taught the dog to continue to hold the object even when a treat is shown (part of the clicker retrieve), the dog will continue holding the object yet become fixated on the treat and stare at it, wanting it. Just as physical movement causes the dog to focus on that movement, and not to mouth, so does showing the dog a treat seem to cause the dog to focus on that, and to take their mind off the thing in their mouth. Then you can click and treat. If just showing the dog the treat doesn’t have this ‘stilling’ effect, you can also move it sloooowly across their field of vision, so they are tracking it with their eyes. Even more distractingly, you can also lure the dog around using the treat, which will almost always have the ‘stilling’ effect, due to the physical movement. The purpose of all this is just to buy yourself some stillness, in order to click.
- Waiting. You can wait for the dog to stop mouthing, even for a split second, and click that. The problem with this approach is: A clicker-wise dog, when the dog doesn’t get the click, is likely to try something else to get it. Like pushing the object at you, or dropping it and trying another behaviour… He or she is unlikely just to continue standing there and staring at you and is unlikely to ‘guess’ not-mouthing – since it is beyond the dog’s awareness, usually. It is very ‘punishing’ for a dog not to get a click and to ‘fail’ like this, and it can be a big blow to confidence. And one of the greatest causes of mouthing is insecurity and anxiety; mouthing usually gets worse when the dog is lacking confidence. So I don’t like this one, despite the fact that it is effectively free-shaping.
- Clicking before mouthing has started. You can only do this if the dog isn’t mouthing as soon as he takes the object. And, even when the dog isn’t mouthing at that moment, it could start just when you click, and you can end up clicking the mouthing. So trying to be quicker than the mouthing, is tricky.
- Gentle tapping of the object. For some reason, this causes the dog to bite down harder on the object to stop it from moving in the mouth. This then gives you an opportunity to click that tighter hold.
- Gentle tuggy of the object. Again, the dog takes a much firmer hold when playing tuggy and getting the dog to really grip it in that way provides an opportunity for a click. However, pulling the object (as in tuggy) and holding the object are two different things – although they do both have the firmer hold in common. Once I’d done a few reps of clicking for a gentle pull, Slate started to try to pull the tuggy away from me when I put my hand on it. Which obviously isn’t what I want, so I didn’t pursue this method further.
- Hold in a stand, in front of me. (I hold dumbbell for dog to take.)
- Hold in a sit, in front of me.
- Hold moving from a stand into a sit. (For some reason, this position-change generated lots more mouthing in both dogs and took a while. I had success with gentle tapping throughout the movement from a stand to a sit.)
- Pick-up from floor, into a sit and present.
I have to say that all this has been incredibly, incredibly tedious and boring – for me, at least. (The dogs are highly motivated and would do this forever, for the treats.) Progress was very slow and I often wondered if we were getting anywhere: No matter what degree of still-mouth we reached at the end of a session, our first hold of the next session seemed to revert right back to mouthing again – no matter how much I reduced the criteria for that first rep.
Part of the difficulty is that this is one of those instances when you’re rewarding the dog for not-doing something (rather than for doing something). And there are a million things the dog is not-doing, at that moment in time. She is not licking her bum. She is not scratching her ear. Not-mouthing the dumbbell is just one of them. So, the reason the learning is so slow, I think, is that it is much harder for the dog to understand what it is being clicked for. I also think that mouthing is usually outside of the dog’s awareness.
Feeling that I was making infinitesimally slow progress, and therefore must be doing something wrong or being incredibly stupid or over-thinking things, or not using some crucial method, I did some googling. I was reassured to find that the things I’m trying are the things others use and that it’s quite normal for progress to be bloody slow and the process boring. So I guess I will just keep on with it.
I especially found this article really good for detailing the various different things you can try. (Although some of them are more about getting duration on the hold, rather than not-mouthing, specifically. I have no problem with duration – the dogs will hold forever – it’s just the mouthing.)
Those photos made me smile, as I am at the point of generalizing my dog’s dumbell obedience retrieve to other objects, with the end goal being birds this spring. Today we used the remote control for the television, a spoon and a camera case.
Also interesting you mention the difficulty of rewarding for doing nothing, as that is what I am facing right now. I’m working on the Stand For Exam exercise for novice obedience for the Canadian Kennel Club. It ends by the handler walking around behind the dog back into heel position (from 6 feet away in front). My dog wants to watch me as I go around, and as I pass directly behind her, the momentum of her swinging her head around from her left to her right causes her right front foot to move about an inch back and out. So I’ve decided instead of rewarding a still foot (as you say, there are a lot of other things she could think I’m rewarding her for), I’m working on her not watching me and looking straight ahead as I pass behind her (and fortunately seem to be making progress).
I think it’s really tough and boring, the ‘reward-for-not-doing-something’! These things are deceptively difficult – they look really unimpressive when trained (since the dog could coincidentally just not be doing whatever it is!), yet they can take loads of training.
Luckily we don’t have to do that stand for examination stuff, even in competition obedience here (I don’t believe).
What about training her to look at something in front of her; perhaps a treat, which she gets after the click? That would be a way to train for doing something (looking at something) rather than not looking around?? Hopefully even when you stop using the treat on the floor, she will have come to associate the cue to stand with the idea of there being a treat in front of her, and would still look forwards?? Think that’s where I’d start with it, although obviously have no idea if it would work! I don’t envy you having to train it!