Generalising Behaviours - is it us or them?

[Rivsky]

Do others think dogs are really as bad at transferring learnt behaviours to different environments (ie from home out on the street) as many of the training books state they are?

I'm planning on entering River into our first obedience trial at the end of this year and came to the awful realisation that I'm a bad trainer and haven't taken him anywhere to do any formal training sessions apart from home and obedience training. So today I put his collar that he wears to obedience on, took him to a new park and set up the treats. When I asked him if he was ready he sat next to me and focussed his attention 100% on me and continued doing all the exercises we worked through really well.

To say I was surprised at how well he worked is an understatement and it got me thinking about whether we, the trainers, have more to do with the dogs ability to generalise than anything to do with the dogs themselves.

For example, River knows when we go to obedience he's expected to heel and not leave my side, whereas when we're out walking generally he's allowed to go walk ahead a little as long as he's not pulling. Now I don't necessarily think that this is a failure to generalise on his part but a failure on my part to consistently insist that he heels when we're out walking (simply because that's not I want him to do!!)

Also in today's scenario - I think the way I approached the park with getting out the rewards etc had him actually expecting the training before I'd even asked him to do anything.

So is a dog's ability to generalise really down to us and the fact that we let the criteria we accept slip so much when we're in a more relaxed environment or is it really that the dogs, for whatever reason, cannot transfer a learnt behaviour from one place to the next?

And is it easier for a dog to transfer a behaviour to different environments once they are well practised in it as opposed to something they are either only learning or have only just began to get the hang of?

[GSD]

Yes and no. Sometimes it seems dogs are terrible at generalizing, and other times, I have to laugh at the books and say yeah right. Defineatly the more practiced the better, and once you hit a few places outside of your home, it's more likely they'l transfer that behavior to other new places.

Sometimes its also that everything is so different, they do fall back on what they know, but after they've been there a few times, they feel they would rather go check things out than listen to you.

But I feel humans generalize very well, sometimes too much. We generalize so much that when a behaviorist is studying an animal in a controlled setting, often a "skinner box" they generalize that learning and behavior to everyday life. Temple Grandin had a great quote in her book "animals in translation" about that, i wish I could remember what it was. She also says that when the look outside the box in the natural environment animals exhibit all sorts of behaviors and things that can't be explained by the "box".

[oc_spirit]

When a dog doesn't co-operate in an unfamiliar place or a location that is not a typical training spot I wouldnt really say its the dog not being able to generalize but rather the trainer not working enough on distraction training At the "new" spot there are so many sights/sounds/smells to intrigue the dog that if he hasnt been properly trained, then listening to commands is the last thing on his mind.

But, for example, a dog like OC who has been taught he must respect my commands regardless of where we are (and I did this by taking him to various locations and practicing obedience) is no different at home then he is at the pet store, or when we are in Chicago etc. And Ronan is the same way although he knows far fewer commands LOL I dont train because of where we are I train because of what I want.

I believe another issue is many people's demeanor changes from when they are in private (at home) and when they are out in the public. At home they are relaxed and comfortable so the dog is relaxed and comfortable. But then put them out in the public's eye and they start to get a little uneasy or easily embarassed if the dog doesnt listen immediately and the dogs pick up on these stresses. If the dog is a cue-reader (looks for subtle body signals to cue him on what he is supposed to be doing) then when the trainer is suddenly acting stiff and uneasy the cues get disrupted. They become unclear to the dog so the dog gets confused.

[Sophie45]

Maybe a bit off point, but I remember when I first picked up Sophie from her board and train with our trainer, she heeled PERFECTLY-it was like she was glued to my side. My trainer, who is a perfectionist, said that is how she should always be on a walk.

Weeellll....maybe being the softie that I am, I thought it was a bit overboard.

So when we go for walks, she has to be at my side with a loose leash, but certainly not glued to my side, and has some leeway as long as she doesn't push or pull or try to lead and the leash remains loose. But when we go back to the trainers for a 'refresher session' she immediately goes back to being glued to my side, even though I certainly haven't required it! She just knows when she is with that trainer, that was how she was initially trained, but as soon as we are on our own again, she still does a good heel with a loose leash, but is not glued to my leg!

[Rivsky]

Thanks for your thoughts - it's interesting to see what others think and I tend to agree that we play a bigger role in a dog's ability to generalise than we actually realise. I just find it amazing how they can pick up on the smallest little change in what we do and how it affects them and how more often than not I don't even realise I'm doing something differently all of a sudden until I need to think about why something isn't working!!

[prolibertate]

Because every dog is an individual and what works with one may not work with another, or how one behaves under certain circumstances doesn't mean another will behave the same way (sort of like kids in the same family, eh?) I hesitate to generalize anything much about them.

In my own experience, I've seen dogs who were trained in one area of the home and had to relearn their training in another area within the home; dogs who learned well in the home but got very distracted when elsewhere, and dogs who would do anything their person asked of them regardless of where they were. This last one could be due to the dog, his particular dogs learning abilities, and how he was trained. I know my own puppy would much rather greet every human in the vicinity at puppy class, or even play with the other puppies, than train when we first started class.

One thing I've learned is that one's dog has to learn to look to their human for guidance (for lack of a better term). If one can get your dog to do this reliably, I think one can get them to listen almost as well once distractions are introduced. Also, dogs are masters at reading their humans body language; it's amazing the signals we can send without being aware of it. And using various different tools also helps them understand better, such as using a particular collar/leash when they can do loose leash walking and another for when they need to walk in a heel position. It's much like the principle used with search and rescue dogs; when the vest goes on, they know they're going to work.