Once a dog learns...

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

 
Well the motivation will depend on the circumstances of course!  Dogs do what works, and will work for reinforcements of all kinds. Why people think that "treats" are the only reinforcer available is so beyond me.
 
With my guys, their motivation to do as I ask might be for a treat (I never "totally" phase out treats, rather I put them on a VSR), opportunity to go for a walk, or a drive, or out to pee, or for a belly rub, or for attention, for their supper, for the opportunity to chase something, for me to throw the toy, to play tug, to get out of the car, because there is an audience cheering them on, etc.
 
In a lot of cases the motivation is from sheer habit. If you ask your dog to sit 500 times, you can darn well be pretty sure she's going to sit for the 501st cue. However, regardless of training, if you use the best reinforcer or the harshest punisher, if your dog does not receive SOME sort of feedback, that makes it worthwhile to the dog to do, the behaviour will extinguish in time. If you asked your dog to "sit' 50 times per day, and each time simply walked away from the dog afterwards, with no reward or attention for it, and no punishment if the dog didn't do it, I assure you the behaviour would extinguish eventually. Because if there is nothing in it for the dog, if the dog is not somehow finding a "reason" to do it, the dog will eventually stop doing it.
 
I don't ask my guys to do things "just because". When I ask my guys to do something, there is always a reason behind it. One problem that I find a lot of people have in training their dogs, is that they DO ask for thigns "just because" way too often, then get annoyed when the dog eventually starts doing it lazily, or stops doing it at all, when they think "My dog KNOWS this so should do it all the time forevermore".
 
My dogs always receive some sort of reinforcement for doing as I ask. That reinforcement comes in all shapes and sizes though, and that's what most folks don't understand. My dogs don't depend on treats. They don't depend on toys. What they realize is a) when I ask them to do something it is for a good reason, and b) they will get some satisfactory consequence because of it (sometimes paired with the knowledge that "not doing it" results in lack of what they want - P-). The value of the reinforcer varies heavily on the context in which it occurs, and what the dog wants at the time, and lastly my mood (I can sometimes give the highest reinforcer for simple things, and a medium reinforcer for harder things, so that my dogs never know exactly "what" I'm going to give them in return, it's what gives strength to VSR's).
 
I suppose the simple answer to your question is that the motivation is that dogs do what works. Simple, but complex at the same time.

Wow. Reading this thread I am having a lot of [&:] [sm=eek.gif] [sm=asking03.gif] moments.

Kim, here you mentioned:

With my guys, their motivation to do as I ask might be for a treat (I never "totally" phase out treats, rather I put them on a VSR)...

And in one of the other threads you mentioned:

... So this whole idea of doling out rewars left right and center is also a myth that the "other side" (yes yes, it's a JOKE!) seems to have a problem grasping as well, along with the purpose of the clicker being a marker tool, and that both that and treats are phased out over time.

I am confused. I remember reading millions of times:

Gradually phase out the reward. This is important.

The question is actually really simple. Lets work with concrete examples, if that will help.

1. You taught your dog to SIT.

ORIGINAL: JM

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.

2. Exactly! You phase it out COMPLETELY. So lets stick to a simple command SIT. I wouldn't be the only one here who doesn't reward her dog immediately after he SITS? By reward I mean a *tangible immediate reward* such as a treat or tug? (Not just food, remember!)

Question: Why does he still do it?

Instead of SIT, you can use the HEEL command as an example. I read tons of success stories in CM and Clicker Training section were we phased out the immediate +R/-Rs.

I think that we are starting to draw a difference between the immediate rewards and long-term awards. [8|]

When I discuss phasing out the clicker and treats, it is portraying the idea that you don't use the clicker and treats "forever", that you are not, for the life of the dog, just giving them treats left, right, and center everytime you ask them to do something. Because most non-clickers seem to go on the assumption that we are "treat-slinging weenies", that we are just forever doling out treats to our dogs to try to "bribe" them to do what we want.
 
When I say that I don't "totally" phase out treats, that means I might give a food reward every 10th, or 30th, or 50th time or even the 100th time I cue my dog to do something. Or I might ask her to perform all of the behaviours she knows, for one food treat. In other words, my dogs WILL work without treats, I have efficiently phased them out of my training (they are no longer continuous), however even the BEST of dogs "needs" reinforcement (or punishment, depending on how you train) every so often in order to keep behaviours strong. That's the basis of why a VSR works. I just happen to use food as "one" of my toolbox of rewards, albeit a lesser-used reward once the behaviour is well-manifested. Once learned I use other rewards instead, as I listed above - things the dog wants.
 
It all comes back to the paycheck idea - if all the people in the world stopped being paid for the jobs they worked at, no matter HOW well-trained you were, and how MUCH you love your job, chances are you wouldnt' continue at that job if you were no longer getting paid for it.
 
I don't know who wrote that last quote of "Gradually phase out the reward. This is very important". I don't know anybody who completely phases out rewards, it's almost impossible. They might jsut learn to use other things as rewards, such as praise, body touching, environmental rewards, walks, etc. If you completely phase out rewards, eventually you're going to extinguish the behaviour. Dogs won't continue to work "just to please us". It's the largest myth humans have created to date. []

You can ask a dog to sit on and off all day long, then ask him to sit to get out the gate the next morning....or 2 mornings afterward....

(insert drum roll)

What did you just do?!!!

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

I think so too!
I asked my husband after reading P. Mcconnell, if he thinks that dogs are "social parasites who have learned to manipulate people...". He said: "Yes. But so are humans..." [] [] Being a part of a social group is nothing more than another need, and it is reinforcing.

When I say "phase out the treats" I mean just what Kim does. But the rewards on a VSR and give a variety of rewards, from petting to priviledges to games to food to everything in between. When I train something my goal is to help the dog understand that doing what I ask is rewarding and basically is no skin off their noses so they'd might as well. It's a win-win. Not only is sitting easy to do but it brings good stuff with it, so when I say "sit" they sit knowing that there's really no good reason not to and lots of good reasons to give it a whirl. When I trained recall, it was all about showing the dogs why recalling when I whistle is a fabulous thing to do with absolutely no down side to it whatsoever. To do that I used a lot of high value treats, yes, but also my relationship with them, their love of attention, and never ever using a recall to mean the end of fun or a punishment (if it does have to be the end of fun I try to give a few more commands in between the recall and leashing up or going in the house so there's not a direct correlation). Now when I blow the whistle it's like a stampede because there's a long track record of recalling meaning something good from me plus a chance to go back to doing whatever they were doing beforehand--a total bonus. Do I still throw a couple treats in my pocket when we go to the dog park (the most distracting place they are ever let off leash)? Yes. To reward particularily excellent recalls in which the dog had to make a pretty tough choice between something already self-rewarding, and recalling.

A variable reward schedule can be highly variable, so that it looks to the average person that comes to visit my house that I'm not rewarding the dogs at all. But they are rewarded, by my attention, by being let out or in, by going for walks, and every now and then when I want to make an impression, or just because I've got one in my pocket, I'll give a food treat.

So like "purely positive", "phasing out the treats" is just a semantic shorthand, but I understand that it could cause confusion.

So, what if, "what works for dogs" is to be in a stable and reliable environment? What if responding to humans in a way that produces "pleased humans" creates a "stable and reliable environment"? What if the dog is motivated by stability more than gratification?

Why don't dogs fight each other day and night to be alpha, to get first crack at food and chicks? Why are the vast majority content to fall in line and be followers? What kind of fulfillment is found in being “just one of the pack”?

What if the dog is motivated to be a functional/safe/harmonious pack member? Humans, especially since the enlightenment, are motivated by fulfillment of the individual. We project this subjectivity onto other creatures as well.

And, so, consider, what if dogs are motivated to behave in ways that promote social cohesion? What if "what works for dogs" has more to do with social organization than individual gratification?

And yet, even millions of people do jobs they are never paid for ... and they do it for their whole lives. Volunteerism is one of the cornerstones of just societies. There you have long term rewards, but rarely witness "instant gratification."

"Paychecks" are part of our particualr cultural understanding. (We even put a price on "pleasing" each other.) But many people don't earn paychecks. Some do subsistance farming/ranching/fishing, others work in trade, entrepreneurs defer earnings in order to grow buisness, communists have yet other means for distributing resources. For my money, it's a little myopic to ascribe free market exchange value systems to the aspirations of other species, even if they do live under our roofs.

This is sometimes an excellent reinforcer...and easily overlooked as such.


"plus a chance to go back to doing whatever they were doing beforehand"

Oh yeah, I am a firm believer in this. If a dog's human is always spazzing out on them, you can bet they will try to appease.

ORIGINAL: Ixas_girl

So, what if, "what works for dogs" is to be in a stable and reliable environment? What if responding to humans in a way that produces "pleased humans" creates a "stable and reliable environment"? What if the dog is motivated by stability more than gratification?

Behavior is maintained because it is reinforced.  Behavior is offered because it gets a need met.  The paycheck analogy is applicable to subsistance workers,  the only real difference paychecks are conditioned reinforcers that are more removed from the primary where as substance work reinforcers are less removed from the primaries.   The real problem is many folks are not able to determine what is actually impacting the behavior.

We had an interesting situation yesterday.  DS finally got around to repairing the shelter that we have off the deck for the dogs.  Now, the dogs are NEVER left outside if we are not home, but they often don't want to come in when we ARE.  They enjoy being out and playing together and thats fine.  The shelter is there for when they don't want to come in out of the driving rain or snow, or as a respite from the sun.  DS opened the gate and of course everyone came running out since he failed to say "get back" before opening it.  DH or I can walk in and out of the that without an issue because we simply say "get back" and they do.  With us, they never leave the fenced area without an "ok, lets go play".  They weren't GOING anyplace....they were just kind of milling around the area of the yard where we play fetch with them.....and DS was trying to get them back inside the fence by throwing the ball into the fenced yard.  Instead of going IN, they were running outside the fence as if the ball had been thrown THERE since that's where we typically throw it.  Poor DS was totally freaking that they were loose, and there WAS a lot of traffic on our road.....it was commutting time and they JUST put our road in as the detour for some construction work on the major highway.  I walked out into the yard and called them and in they came.
 
Why?  Because they are used to doing what I ask them to do.  DS is not consistent in the way he deals with them where DH and I are very consistent.  When they came back into the yard I wasn't waiting with a bag full of treats, they simply did what I asked because they have been trained to do what I ask.  I've noticed on our romps through the woods that if DS is along and opening HIS mouth, there is less compliance.  It seems to me that dogs are very much creatures of habit and those habits can develop VERY quickly. 
 
At the facility where Mom was living. DH often takes Tyler for a walk on the bike path out back while I'm scooping ice cream.  ONE time, I went along after the Ice Cream Parlor.....and Tyler was not as complient.  His habit had become walking with DAD on that trail, and having Mom along was confusing to him, so I shut up and let DAD handle things and he was great.  That walk had become his and dads time together.....a HABIT they developed, and I was a distraction in the habit....
 
Just my early morning, not enough coffee yet, musings............

Force of habit definately is a factor here as well. When I had a friend dogsit for us a few weeks ago I had to explain a lot of our routines to her because the dogs are definately creatures of habit. When I showed her where the dog food and bowls were and which one is Marlowe's and which one is Conrad's she asked if it made a difference which bowl is who's. Oh yes. It makes a difference to them! They've each eaten out of the same bowl, Marlowe for a year and Conrad for a number of years. When I release them from their down-stays when feeding them, they each go immediately to THEIR bowl and there is no confusion. If I switched bowls up on them it'd be chaos!  Rounding the dogs up to take them upstairs after dinner (our family room is on the third floor and I like the dogs to either be there with us or in the bedroom in the evenings) is also a habit thing. Their routine for us getting ready to leave the house is also complete habit with very little formal training involved. They just know. Marlowe goes to his crate first and I shut the door, Conrad goes to his crate and I give him his Kong. Then I brush my teeth etc., put Marlowe's Kong on the floor and open his crate, and then I leave. If I try to switch that up on them it gets very weird and they don't really know what to do.
 
DH also has a harder time getting compliance from the dogs because he's inconsistent with them and doesn't work with them very much. He likes to set up little tests for them to, I don't know, test their loyalty or some such BS, but it always is a miserable failure because he barely does obedience with them at any other time. It's like a substitute teacher walking in to your 8th grade classroom and assigning you calculus.

I understand the analogy with the paycheck, and I am not denying that something is still reinforcing a dog to perform. If it wasn't reinforcing, you'd ask for a HEEL and go left, and the dog would go right, and that would be the end of that.
In my mind, if you say "dogs do what works for them" is like saying: the behavior is reinforcing because it's reinforced. [8|]

ORIGINAL: mrv
The real problem is many folks are not able to determine what is actually impacting the behavior.

Exactly!

A "habit" could be it. But, a habit is a tendency - it can't be a reinforcer... Something still has to reinforce the tendency.

I am starting to differentiate reinforces as possibly being short-term and immediate, and long term ones. I was just reading one article from the Journal of Behavioral Medicine ("Associative Learning, Habit, and Health Behavior"), and there a 'habit' was defined as "a stable behavior pattern overlearned to the point of becoming automatic and marked by decreasing awareness and increasing dependence on secondary rather than primary reinforcement." That's interesting, because as we treat or play with our dogs in getting them motivated to perform a certain behavior, once learned, this reward might well loose its primary significance, giving way to a new reinforcer which is then responsible for *maintaining* the behavior.

"tendency" is not behavior.  Behavior is reinforced by the environmental as well as artificial "people control" events that follow.  You are absolutely correct that behaviors are variable with respect to the delay of reinforcement that is tolerable.  It might be more helpful to look at a paycheck from the stand point of the value of the reinforcer (although CEOs will blow this point to kingdom come).  If you dont need the money from your paycheck immediately but love the work you do, it is less likely it is the money that keeps you there.  (There have been multiple times in my life when my paycheck was mailed home because I did not remember to pick it up).  Now this was not the case in the beginning of my working life, but it is now and has been since I became a school psych. in the 80's.

Many of the activities and behaviors I do now have "inherent" value as conditioned reinforcers.  So it is not the work habits I have that are reinforced but the tasks (behaviors) completed that result in my needs getting met. 

One of the reasons proposed for why this is true for humans is the issue of language and self mediation.  It may be that dogs do something similar.  Look at the service dogs, herding dogs, hunting dogs and other canines with day in day out jobs.  The behavior itself could be something of a conditioned reinforcer.  Now that is all speculation on my part. However the idea of instinct and genetic memory suggest that completion of the behavior results in very specific neurological changes.  Could be we are hard wired to do specific kinds of things if the environmental cues and consequencs are there.

One of the reasons proposed for why this is true for humans is the issue of language and self mediation. It may be that dogs do something similar. Look at the service dogs, herding dogs, hunting dogs and other canines with day in day out jobs. The behavior itself could be something of a conditioned reinforcer.

Here is what Myrna Mielani has to say about working dogs:
...
Additionally such finely tuned, intelligent animals may not respond well to food-training, which also makes sense if you think about it. The most successful working dogs, and therefore the ones most likely to be bred, were those who enjoyed working the most. Consequently, they come hard-wired with a desire to work because it's fun, not because they get some external reward for doing so. In fact, offering some of these animals treats disrupts the natural rhythm of learning. They expect to master a task and move on to another, more challenging one. (Imagine a shepherd giving his dog a treat every time the animal sat!) When the owner keeps rewarding the dog for doing the same tasks and never expands the repertoire, the animals may become bored and even agitated for two reasons. First, by not quickly phasing out the food treat, the owner actually teaches the dog to do the task for food rather than for fun, which makes it much more difficult for the animal to internalize the new lesson. Second, in addition to losing the inherent sense of accomplishment that comes just from doing a job well, the dog never gets the opportunity to realize its full potential. All this results in a most unfortunate paradox: The dog appears dim-witted and unruly simply because it's so smart.