Sit/down/stand training: do dogs associate the movement or the final position?

Question

I'm training my dog the different positions (sit, stand, and down) with an implied "stay". That means that when I ask him to sit, he's supposed to stay in that position until I "release" him.

For "sit" and "down", so far so good.

But then I realised that these first steps are easy because in the initial position the dog is usually standing.

I was struggling to train a "sit" from a "down", then I realised that might be the problem, I added another cue to "sit from a down" and he quickly got it.

So the question is: are dogs learning the movement or the final position?

Similarly what about a "stand" cue? In my case it means "stand and stay". But if the dog is already standing, he just has to stay. Do I risk confusing him by using a single cue?

What is your experience with this?

Answer 1

This is a really great question... the answer is it depends and both or either.

Dogs are VERY specific, they don't naturally generalize as much as humans do. I don't mean that they do not generalize... it just takes more experience\training to get there. This applies to your question because a specific movement from a stand to a sit is very different then the general "I want you in a sitting position" from whatever current position you are in now. It sounds like this is the difficulty you are running into when trying to teach the sit from a down when your dog already knows a sit from a stand well.

So depending on what you are rewarding you could teach either the position or the motion to get into the position. If you are not consciously thinking about which you are rewarding then you are probably muddying what your expecting which may explain why it is taking longer for your dog to learn some things and faster for others. If sometimes you are rewarding and naming the movement and sometimes you are rewarding and naming the position after they've already completed the movement. If this is the case you are really calling two different things the same name. Does "sit" mean "lower my back-end by bending my knees" or does it mean "hold still while my back-end is down until you release"? Does it also mean your dog should be looking right at you ready for your next command or can they relax and look around? You can make it mean "move in what ever way necessary to get your back-end down and your front end up then hold still in that position and watch me waiting for the next command until I give you a release word" but that is a lot of steps and a really generalized command so you will have to spend some time teaching your dog that it means all those things.

So the best thing to do for a complex behavior like that is to decide what you want for the end product then figure out how to break that down. You will then teach\reward each small part and slowly put them together until you have the final product. And you will need to practice it in lots of different places, with different distractions to help generalize (or proof) the behavior.

For a stand it sounds like you are wanting to teach the general "move into a standing position then stay in that position without moving your feet until I release you". You could start by either asking for the motion where you would start from a sit or down and reward when they moved out of that position into a standing position. Or you could start from a stand and reward for holding still for longer and longer duration then when the dog got that pretty well you could work on "generalizing" it by teaching the dog to move into a stand from a sit. Then also from a down. Then also from a walk. Then also from a full on run. You get the idea... you just have to keep adding different situations to start from until your dog figures out what the general "stand" means.

For commands where I care about the movement and that don't involve a bunch of steps. I like to specifically reward the movement. So if my dog is in a down and I say down again they should try to get lower to the ground... lower their head press their body down... whatever they can to do show me they are trying to "down" even more. For something like "back-up" you would start by rewarding even the tiniest movement backward then continue to reward additional movement backward and be careful not to reward holding still after the backward movement has stopped.

For additional reading on dogs and generalizing:

• Temple Grandin talks about dogs and generalizing in "Animals in Translation" This is an excellent read for anyone who wants to better understand animals
• http://www.nirvanatails.com/training-tips.shtml (more of a how-to... generalizing can also be called or related to "proofing")
• generalizing in new/different environments this one is more about exposing dogs to different situations to help them generalize
• "hey my dog knows how to do this but isn't... what gives?" an example of someone else who also realized that their training needed more generalization