Once a dog learns...

I'll tell you how I train

First of all I teach the new ability with food and toys (each one depending on the exercise and the level of precision or speed of the exercise I want). I want the dog to be very motivated, but I also use the food to guide the dog on the first stages (by example, lowering it for a "down"). I give small pieces of food when the dog do exactly what I want, as in clicker training, but I ask more of my dogs and the real reward comes after the exercise is stable, and the real reward is my affection and to play with me (not to give the dog a toy, if you understand the difference).

After the dog knows an exercise in a mechanical way I test the dog putting complications. This make the dog really comprehend what the exercise is, more than repeat and repeat always the same sequence. I force the dog to find his own solutions. By example, I call the dog standing in a bench and giving it my back. If the dog doesn't know what to do because he has always been called  with a person standing in front of him he has just repeated, but doesn't really know what a "Here" is. When the dog is capable to go in front of me and sit as near as he can independent of my position or the obstacles in his way, the I assume the dog is really understanding the exercise. When you do this kind of test from time to time a good dog find his own reward just solving it, in the same way we do puzzles, but you have to be carefully not to do it too much often because some dogs find it very addictive and then don't respond if the exercise is too simple for them. Other way to test the comprehension level is to put a reward opposite the direction of the excercise, by example, to ask a Down putting food over his head or to make a Heel with the dog looking at you but the toy visible on the floor.

When the dog understand the exercise and JUST when he does (there is nothing more frustrating than to be asked to do something you don't understand and here is when many dogs start to hate training), then I start to correct the dog if he is not doing it in the perfect way. I'm the boss, the dog knows his duty and is not allowed not to obey. I DEMAND attention and speed and it not depends on a ball in front of his nose. If I have constructed a good relationship with my dog, he will be eager to please me and the worst punishment will be to stop the work and to be ignored the rest of the day.

I want my dog to Work with me, not to just do tricks for fun and that is when I want, not when they want because there is a hot dog for them.

With trick training and having fun, I don't discontinue treats because I want the dog's drive, anticipation, and excitement level up a bit. I use a voice marker rather than a clicker with trick training because I don't like the over-the-top addictive level clickers bring into the relationship and don't want to simply be viewed as a codependant to a clicker-and-treat junkie. But that's just me.

And here... I really, really agree.

I think we all have different ideas of what's under the hoods of our furry friends. I don't find my dog to be a highly mechanistic object that produces objective results. I behave with her as if she's a thinking, feeling sentient being.

If I angrily shoved a treat in my dog's face every time she respnded to "sit", she'd respond more strongly to the negative energy than to the food reward, and I'd have a really unreliable cue there.

It's really interesting to see your answers. I am asking this question because it's important for me to have an idea where we are going with training... What's our goal? We sure want to have an obedient dog that is able to HEEL, COME - whatever, *one day* without any tools on our part. And, when we do put in years and years in training our dogs, what do we fall back upon when asking our dogs commands? Is it a habit on their part or an expectation of a long due reward? Is it because we've made such a strong association with the reward that it has produced this permanent 'imprint' on their brain? Does it have any relevance to dog's pack mentality? Has a dog learned that following certain commands makes her pack mates happy? And when they are happy, good things happen... possibly even more rewards in a different shape or form? As I see it, the answers seam to be divided between those who think that we - the owners, have something do to with this continuing desire to obey, and those who think it's all the rewards we once gave.

Personally, I seriously doubt that my dog is still expecting treats for peeing outside. If dogs would await for a reward after hundreds and hundreds of times we'd failed to give it to them, then Extinction wouldn't be possible as a method to correct dog's behavior. It just doesn't make sense for a dog to continue pushing the button on a Coke machine that hasn't worked in years; physiologically, it's a waste of energy... [8|]

Praise is usually a predictor of good things coming to the dog (or bad things not coming) which is why dogs respond to it. In and of itself I can't see how it would be reinforcing. But follow up verbal praise enough times with something tangible--a treat, a priviledge, a massage, and the praise comes to mean good things for the dog. If you were to praise the dog and then immediately smack it upside the head every time, you'd wind up with a dog who did not enjoy praise.

True. If you praise and slap, the dog would hate praise. If you click and slap, the dog would hate a clicker. If every time you'd see me I'd slap you - you'd hate the sight of me. []

I did not read all the posts, so please excuse me if this has already been covered.
 
A dog or any organism continues behavior because it works.  Some need within the organism is met by the consequence that follows that behavior.  In some cases what is at work is a delay of reinforcement.  The dog has learned to tolerate delays between the end of a behavior and delivery of the reinforcement. 
 
In other cases, a secondary reinforcer is adequate to maintain the behavior.  The use of variable schedules of reinforcement are at work in these cases.
 
Reinforcement is always occurring or the behavior would disappear.  The relationship with your dog would readily develop very subtle reinforcers of which the handler/owner could be unaware.
 
To offer a commonly used but pretty powerful analogy,  how many people would continue to work if unpaid.  That would be variable across the board.  However, there would be situations in which something at work was adequate to maintain your attendance even if money was not the object, think of volunteering.  So the reinforcers change and behavior and schedules  of reinforcement result in generalization.  Reinforcement gets shifted from primary to secondary and behavior continues.  It is different for each individual and situation.

ORIGINAL: houndlove

Force of habit and/or a variable reinforcement schedule.

Ding, ding, ding, ding!!!!!  houndlove gets the brass ring! [sm=dance.gif]

It's not "pleasing the master", it's the hope that the master may just remember to produce that reward one of these times...if you never produced another reward ever for that behavior, it would probably extinguish eventually, or get a lot less reliable. (He comes most of the time.  She's about 80% reliable at the "down". Etc.)

ORIGINAL: mrv

I did not read all the posts, so please excuse me if this has already been covered.

A dog or any organism continues behavior because it works.  Some need within the organism is met by the consequence that follows that behavior.  In some cases what is at work is a delay of reinforcement.  The dog has learned to tolerate delays between the end of a behavior and delivery of the reinforcement. 

In other cases, a secondary reinforcer is adequate to maintain the behavior.  The use of variable schedules of reinforcement are at work in these cases.

Reinforcement is always occurring or the behavior would disappear.  The relationship with your dog would readily develop very subtle reinforcers of which the handler/owner could be unaware.

To offer a commonly used but pretty powerful analogy,  how many people would continue to work if unpaid.  That would be variable across the board.  However, there would be situations in which something at work was adequate to maintain your attendance even if money was not the object, think of volunteering.  So the reinforcers change and behavior and schedules  of reinforcement result in generalization.  Reinforcement gets shifted from primary to secondary and behavior continues.  It is different for each individual and situation.

Another salient post, mrv.  I wish that everyone would just memorize this and apply it to their training. 
I will never understand why it is sooooo difficult for people to understand these principles, because once you do, dog training is pretty simple. []

When my dog was a puppy, I gave him a high-value treat whenever he went to the bathroom outside. He rarely if ever gets treats now. I can't remember the last time I gave him a treat.

But he still poops and pees the minute he gets outside, even though it is in his best interest to dawdle.

I don't have a window onto my dog's brain, but I am guessing that I made a sufficiently strong association:

Eliminate and something really good happens

 
This one is a bit different IMO. The act of eliminating is inherently rewarding (just HOW good does that early morning first pee feel?  Be honest.)  Which is why treats alone aren't enough to house train a dog - the other half (or more than half) of the equation is vigilance to ensure the dog never goes in the wrong place.  Also, this is why some dogs can be housebroken WITHOUT treats, just vigilance and a good schedule. 
 
My dogs ALWAYS get a reward for good behaviour.  If its a new behaviour, or an old one in the presence of a new level of distraction, then they are likely to get a high-value food reward or special toy.  If it's an "easy" behaviour, then the reward will be "phased out" to the level of a pat or a Good dog.  Or the reward will be something they want, in the course of NILIF.

I will never understand why it is sooooo difficult for people to understand these principles, because once you do, dog training is pretty simple.

I think it gets confusing because it's not emotional.

I mean, we are not all sheepherders or police officers, so we all have pet dogs for largely emotional reasons, not for work. We love them, and they love us. I do.

And I think that Tina is asking about the nature of "obedience" on an emotional level, and there are a few people who are connecting that concept of obedience to love.

I am completely convinced that my dog loves me, whether his head is in the garbage can or he is in the longest down-stay ever. But I don't think he is obedient because he loves me. I think that he is obedient because I trained him well. I established a strong link between What The Dog Wants and Good Behavior.

I like the way this works, because I can enjoy my emotional relationship to the dog without getting all frustrated about his performance... because he makes mistakes and gets worked up and is generally a big puppy with a LOT of energy that we are getting better and better at channeling every day.

Looking at his behavior separate from our emotional relationship helps me to respect that he is a sentient creature with his own needs and wants.

ORIGINAL: Chuffy

The act of eliminating is inherently rewarding (just HOW good does that early morning first pee feel?  Be honest.) 

 
Exactly! It feels great! []

Oh...I see where you are going with this...you can seperate the emotion connected with the dog "pleasing".  It becomes less personal.

 As in, if the dog  doesn't obey a command, the emotion of "he doesn't want to please me" instead becomes, maybe I am not communicating what I want properly.

Is that right?  If so, that is a very good point.  It becomes more about the training...as in the handler and less about the dog.

His performance is not about pleasing you but about your own training. Geesh..redundancy.

I am not articulating this properly am I?   This usually happens to me after an aha! moment

I like the way this works, because I can enjoy my emotional relationship to the dog without getting all frustrated about his performance... because he makes mistakes and gets worked up and is generally a big puppy with a LOT of energy that we are getting better and better at channeling every day.

Looking at his behavior separate from our emotional relationship helps me to respect that he is a sentient creature with his own needs and wants.

The question is "what motivates the dog" in the absence of treats.

JM, yes, I agree with you, training is a means to establish communication between owner and dog. My dog's failures to follow cues are due to my lack of communicating them. One reason I know this, is that she is so eager to "please me," even in the absence of "rewards." When she fails to perform a cue, I look to my failure in communicating it.

But still, the question remains ... what motivates her to comply? Some are suggesting habituated response, others are talking about an internal desire of the dog to "belong" with his social group.

It's obvious when you look at people and their dogs that the dogs aim "to please" their owners, you can see it in the eagerness of their gestures, the look in thier eye.

The question is, when your dog looks at you, and performs those cues you give, does your dog see you as a vending machine or as an inspiring creature with whom he wants to be in good company? Is the dog simply a pleasure seeking, pain avoiding simpleton? Or does a dog have social interests that motivate his behavior?

Why do we have pet dogs instead of pet monkeys, or squirrels, or bobcats? Because dogs are reliable in complying to us. Is this because they are more self serving than monkeys or squirrels or bobcats? Or, could it be that their social needs are more in line with ours, and they are highly adaptive in allowing us to serve thsoe needs?

ORIGINAL: Ixas_girl

does your dog see you as a vending machine or as an inspiring creature with whom he wants to be in good company?

 
Thats what i have always wonder [&:]

A dog's need for social structure *is* often misinterpreted with in the psychoemotional language of human love.

It serves both dogs and humans alike, to understand our own human projections separately from his internal social needs.

We could be super general and say it doesn't matter whether the dog wants treats, toys, or to feel a sense of belonging, he's still just meeting his needs. But that would be a short and boring conversation, as it would apply to all living organisms, equally. Unless of course you get theological and cull a specialness of man among all creatures by some divine state of grace (civilized vs. savage). [8D]

well see...here is the thing that you guys are just not grasping for one reason or another, even though it has been said a gazillion or more times....once a behavior is established, treats are phased out....

I am having a hard time understanding how this can be so hard to understand.

ORIGINAL: espencer

ORIGINAL: Ixas_girl

does your dog see you as a vending machine or as an inspiring creature with whom he wants to be in good company?

Thats what i have always wonder [&:]