Motivation: Extrinsic vs Intrinsic

[Alfreda]

In several of our super broad discussions ("Corrections" and "Praise") on this forum, Mr. McCaig and others have expressed concerns about a reliance on external rewards: that it might reduce or undermine the internal motivation(s) of a dog.

 

This interesting question comes up here and on other dog training forums -sometimes there are references to (human) studies, and sometimes it is just a belief that children/dogs should not be "coddled," or over-praised, or that dogs and children need negative consequences and boundaries.

 

Some studies (in human education) for example, do suggest that the approach of "ignore the bad, reward the good" might have some undesirable consequences, and that how and when praise is used can make a difference. If you look into the topic of "Intrinsic vs Extrinsic" motivation one quickly finds a multitude of books and studies- mostly in humans- from fields ranging from pyschology, education, to business and economics.

 

So, I thought I'd bring it up as a topic. So often our discussions get caught up in the bi-polar framework of "correction" vs all R+ training. Yet BOTH of these are extrinsic consequences for behavior. Is there a danger in relying so much on externals? How do we go about nurturing the internal drives and strengths we value? Such as: Problem-solving, persistence, generosity, humility, grit, teamwork, confidence?

 

To get things going: here's a link below to a blog by Denise Fenzi on the subject:

 

file://localhost/Users/rebeccarumsey/Downloads/Internal%20Motivation%20%E2%80%93%20do%20we%20value%20it%3F%20%7C%20Denise%20Fenzi.html

[Alfreda]

http://denisefenzi.com/?s=intrinsic+motivation

 

This link might work better

[kingfisher7151]

I can't comment yet, (I'm at work ) but I'm very eager to watch this topic! Should be an interesting discussion!

[ourwully]

This is one of the most interesting accounts I have read. It all makes perfect sense and is supported by examples. I spend nearly all my time with Juno and I think (?) I praise her on a continual basis but I have to admit I am a walking cookie dispenser. This blog has made me think long and hard about my relationship with Juno. Although Juno is an amazing dog and does many many things well, there are certain times when she just ignores me. At these times I often resort to cookies! Juno is almost two and a half now and I am very happy with her overall but I would not say that I would trust her one hundred percent when there are distractions around. For example, I would never have her off leash in a situation where cars are around. I just wouldn't feel safe for her.

 

This blog certainly makes me wonder if my reliance on cookies for training has had a long term detrimental effect. I am not going to give up cookies, that is for sure, but I think I am going to cut way down on them and make even more effort with praising Juno.

[Root Beer]

Speaking of Fenzi, ourwully, Sue Aisby is actually teaching a class on the Levels through the Academy this coming term. It's 65 dollars to audit, and you get to watch her work with all of the gold level students, who submit their work through video.

 

You might find that interesting, as I know you have worked through some of the Levels on your own, with some success and some brick walls!!

 

I know this is off topic to this discussion, but I've been meaning to mention it. I put it on the forum, not in a pm, because there might be others who are interested, as well.

 

Link, for anyone interested . . .

 

http://fenzidogsportsacademy.com/index.php/courses/7435

 

OK, back on topic . . .

[teresaserrano]

I started getting involved in "serious" dog training 8 years ago, and for those 8 years I've trained in the same place. No food, lots of focus on praise and building a close relationship with the dog. There was the phase where I thought it was right to never use food, then as I read more and more and attended seminars and workshops, I realized there's no wrong tool, it all depends on the dog, and food is a major contributor in teaching a dog. Still, I don't use food in class, although I use it when working with Tess eslewhere. I've found that it is an invaluable resource, not only because it marks more clearly, in some circumstances, what's right, but also because with food one can lure (I've recently discovered the power of luring, it's so much fun and definitely comes handy often).

 

But I'm not sad I've trained there all these years. I learned a lot, and I learned the absolute power of praising your dog, and meaning it. Tess likes working for food, but she clearly likes it even better when I use praise and play as rewards (but sometimes food just works better, in that situation). And not being used to using food when training meant that I thought it was normal that she would be reliable without food (and the fact is that she is reliable without food). Sometimes I feel that food is a bit overused, that people just don't believe the dog can do something if there is no food around, so they never try.

 

When I work with a dog whom I don't know, food is the best tool. But with Tess, we have a relationship, and a close one, so there are a lot of other reinforcers working there, and one of them is no doubt an internal reinforcer. It's not exactly that she wants to do something just because I asked, but I do think she values my aproval, I do think she does a lot of stuff just because she knows I will be happy (with all that implies as advantages for her). When I get up and ask her to give me my slippers, she does and I throw a party and tell her she's just wonderfull. Never once did I give her food for that, and never once did she fail to go get my slippers and act absolutely happy about it.

[Root Beer]

Is there a danger in relying so much on externals?

 

I believe the answer to this question depends on a good many factors.

 

First, what is the handler/trainer's goal? For the one dog I had who was only a pet, and for my retired dogs, I don't care a bit about relying on externals. Especially for an older/retired dog, we are at a stage of the dog's life where life may get harder physically, but I do my best to make it easier when it comes to expectations in training. I continue to do video work with my retired dogs, but I generally stick to venues where I can have food on me, even if it is allowed only with the older dogs.

 

For a younger dog, I don't have any reservation about lavish use of externals, but it is important to smartly pare them back when it becomes appropriate to do so.

 

That's the key for me - I am not on anyone else's schedule. I've been told, "lure three times and then don't lure". Some dogs need a lot more than that. Some need less. Sometimes a particular behavior requires more "luring" or presence of external motivators. For me, the situation, and the needs of the dog, determine my choice in the matter.

 

So, that brings me to the next question - what does the individual dog need in the circumstance at hand?

I never would have gotten Tessa through a tunnel without a TON of treats, and cajoling . . . but once she mastered the skill, she no longer needed external motivators to be willing to go through a tunnel. So, now I don't use them (unless I am working on something in particular, like distance, or discrims)

 

There are times when external motivators aren't needed, or are barely needed. Then I don't use them. It really is about knowing your dog and being able to read the situation.

 

Oh, one last thing - TEMPERAMENT!!!!!

 

If you have a dog with a strong work ethic and a great sense of involvement in learning, you can get away with less external motivation. But if you have a soft dog, or a dog who lacks resilience, then the use of external motivators can be a Godsend that makes training possible at all!!

How do we go about nurturing the internal drives and strengths we value? Such as: Problem-solving, persistence, generosity, humility, grit, teamwork, confidence?

 

This is a great question to consider. I tend to try to bring this out more once a dog has mastered new skills. In the process of weaning back the externals, the internal drives and strengths need to come to the fore, or else it can all fall apart.

 

Tessa is an absolute gem of a dog in this regard. She is the best working partner I have ever had, among my own dogs. I try to nurture her drives and strengths by giving her opportunities to use them (for her competition Agility provides a good bit of that), and by giving her appropriate verbal feedback to let her know that I appreciate what she brings into our mutual teamwork.

 

I think the key really is coming to know the dog and to be able to allow the dog to work through challenges and to grow in whatever work the team is doing together, but without erring on the side of overwhelming the dog.

 

It's a constant learning curve.

[stockdogranch]

When this discussion came up some years ago, someone recommended a book called Punished by Rewards (Kohn, 1999). It is not new, and it is about humans, but it is a most excellent read.

A

[Alfreda]

Great comments! I just wanted to ask about praise- Is it intrinsic or extrinsic?

 

Fenzi seems to see praise as more of an "intrinsic" (at least compared to food); because praise is related to the dog's internal drive to please... I guess I had been thinking of praise as an external form of feedback/reward...? Which is it?

 

We have carrots and sticks that are clearly external: food/toys & leash jerks/shocks. Yet they too are connected to internal drives: hunger, pleasure, play verses the desire to feel safe, comfortable, avoid pain. So what about when we say "good dog" vs "bad dog" "yes" vs "no"?

[Maralynn]

I think praise is both. The external verbal communication and the internal partnership and team work.

 

I also think that as we build value for the work, it becomes the all encompassing reward. If you have good communication with your dog and train it well for it's job, then doing the job and getting it right is rewarding. A well trained dog works with precision and confidence. And the look on their face when they get it done right is one of happy satisfaction.

[geonni banner]

I also think that as we build value for the work, it becomes the all encompassing reward. If you have good communication with your dog and train it well for it's job, then doing the job and getting it right is rewarding. A well trained dog works with precision and confidence. And the look on their face when they get it done right is one of happy satisfaction.

 

This absolutely.

 

I think the whole "desire to please" thing is way overrated. Dogs like to be praised, but that isn't the same thing. I have also seen dogs who were striving for performance show a brief annoyed response to unmerited praise. Once they grasp what is wanted, an engaged dog will strive for skillful execution for its own sake. A lot of the time one needs to just shut up and let the dog work. Shoving treats at them or nattering inane strings of praise-as-encouragement are fine in a teaching stage, especially for a young pup or an insecure dog that has not had experience with rewarding trainer/learner situations. But a dog, when handled properly will relish a new challenges and delight in working out the puzzle of a new task/ request. At that point, the proffered treat will often become simply an interference or distraction. They will just be unwanted static jamming the airwaves.

 

Once the dog solves the "problem" by the exercise his own diligence and brain, he will bask in praise. When he knows he's done well, he'll be pleased to know that you know it too.

[GentleLake]

You wax poetic, Geonni . . . and perhaps idealistic.

 

This may be true for some dogs, but certainly not all. There are plenty of breeds and types of dog with a much more independent bent who really aren't motivated to work for their handlers in the way you describe.

[geonni banner]

You wax poetic, Geonni . . . and perhaps idealistic.

 

This may be true for some dogs, but certainly not all. There are plenty of breeds and types of dog with a much more independent bent who really aren't motivated to work for their handlers in the way you describe.

Certainly not all dogs, but most... if you raise 'em right. As far as independent - I like 'em that way. I choose 'em for that. And every one I've raised turned out like that. I don't like a sycophant dog.

 

YMMV

[alligande]

It's not just praise what about the premack principle which can be applied in all sorts of circumstances. I used it heavily when teaching recall, come back and you get to go back to doing what you want.

I was recently at an agility seminar with a really great handler, he is not a dog trainer, but he is a Spanish world team member and R was being cranky about weaves and even with the language barrier it was obvious he thought I was nuts when I did not reward him with tug after the poles when he got it right, but continued on, that for my dog is the reward, running and jumping. He asked how will he know he was right, and I said by being able to go on... This is not a reward for every dog but it is for mine, playing the game with me is rewarding, obviously I use all sorts of rewards as well but in this circumstance I know stopping him was not what he wanted.

[Donald McCaig]

Dear Doggers,

 

Since I've been running our dogs from an ATV they've developed sore pads and to have them accept booties I intend to use treats. I thought Ms. Reinzi's (cited) suggestions were sensible though I (generally)use much less praise than she does. That said, I let my two older dogs out at 3:30am and go out myself with my 4 year old who has been known to roam. When he does his business and reappears I generally say "Good dog".

 

So: treats and praise are tools. The rattle paddle, plastic bag on the end of a stick and ecollar are also tools. Corrections (almost all verbal/tonal) are a way of life.

 

I train sheepdogs/companions. I do not train salukis, alaskan sled dogs or Jack Russells. I train dogs genetically predisposed to work with me - they WANT to work with me. That's the case when I'm teaching them to walk behind so I can control them in a motel parking lot or teaching them to shed sheep. I absolutely depend on their desire. I don't put it in them and don't know how I would.

 

My relationship with my dogs does not depend on love nor even kindness I've known brutes whose dogs worked happily and brilliantly. Our relationship depends on respect: they for me, me for them. We both have responsibilities so that our shared world continues to make sense. If they come up on the couch at the end of the day to schmooze, that's their choice.

 

A long time ago, J.M. Wilson answered, "Of course you should talk to your dogs, madam. But talk sense."

 

That gets easier.

 

Donald McCaig

[ourwully]

Root Beer - thankyou for the course info but I am working with a personal trainer for now.

 

These discussions are truly fascinating and really highlight the strong convictions that we have towards training and our dogs. It is clear that there are way too many variables to choose one course of action. A quick glance at the current threads will show that some members are having difficulty with basic things like walking on a leash while others seem to get dogs that are good to go right from day one! I am convinced that the nature of the dog is a huge factor but I am also convinced that certain trainers get it right from the start and things just get better as they go on.

 

My experience with Juno has been filled with trial and error. It is really obvious to me now that I have been way too reliant on treats but looking back it seemed the only way that I could get things done. Reactivity on walks was a huge issue early on and we dealt with that quite well by using the Look at That game with treats. Juno also went through a pretty severe fear period and treats helped us get through that as well. Unfortunately, we are now pretty reliant on treats. The challenge now will be to reduce them. With more experience I may have been able to deal with these issues without treats. I am sure if I had the knowledge I have now, we would have done a lot better and reached our goals a lot sooner. All that said and done, we have reached a point where I am really happy with our relationship. I should also say that the challenges, although frustrating at times, have made life interesting and enjoyable over the last two years.

 

Bill

[Alfreda]

My experience with Juno has been filled with trial and error. It is really obvious to me now that I have been way too reliant on treats but looking back it seemed the only way that I could get things done. Reactivity on walks was a huge issue early on and we dealt with that quite well by using the Look at That game with treats. Juno also went through a pretty severe fear period and treats helped us get through that as well. Unfortunately, we are now pretty reliant on treats. The challenge now will be to reduce them. With more experience I may have been able to deal with these issues without treats. I am sure if I had the knowledge I have now, we would have done a lot better and reached our goals a lot sooner. All that said and done, we have reached a point where I am really happy with our relationship. I should also say that the challenges, although frustrating at times, have made life interesting and enjoyable over the last two years.

 

Bill

Bill, I don't think you, or anyone should feel "bad" about using treats at all!! As some of the most experienced folks have been suggesting, we're talking about nuances here to some extent.

 

Treats are a form of reward that's easy and useful. Easy for beginning trainers to deliver. Useful in many situations as Kristine describes. They're also very useful for teaching brand new behaviors or sequences, and for generalizing those skills to more complicated, distracting situations.

 

The downside- is when people don't fade treats out for behaviors once they're learned, behaviors that have become "easy." Continuing to treat those becomes superfluous.

 

External rewards also might be superfluous, or even destructive- *IF the behavior itself is extremely rewarding in and of itself. My dog loves to run with me on the bike- I don't treat for that- he loves it! I did train staying in "heel" and "side" on command when biking, that is more demanding, and curbs his enjoyment for a bit. Calling a dog away from a live squirrel we are competing against a very internally (for many dogs) rewarding alternative. Taking a BC out to work sheep- we're going towards something inherently REWARDING!!!!

 

So, the use of food and/or toy rewards to teach brand new, maybe non-instinctual, unpleasant, or boring (mechanical) tasks is one excellent use of food (in my mind). But once behavior is learned and established, randomize, and fade them out.

 

The use of food to "counter-condition" fear-based behaviors is (to me) in another category. In that case you're not so much just rewarding a superficial behavior- You are working to transform a negative emotional association into a calmer, more positive one. ("classical conditioning").

 

The pros can correct me, but in LAT- the way I see it, is that it works on both levels. The dog looks at the trigger, then, in anticipation of a treat looks away to get the food. On the superficial level, the dog is rewarded for looking away from what it fears and for disengaging its attention. On the deeper level, the dog is building a more positive emotional reaction to seeing the other dog at that tolerable (below threshold) distance.

 

Does that make sense?

 

To me, the interesting challenge is to be aware *when* and under what circumstances will I use external rewards? When will I fade them out? When will I make the most of those activities, behaviors that are inherently rewarding to the dog? And when, if I use too much external reward, might I be in danger of harming the inherent reward (maybe never with some dogs).

 

Personally, I enjoy some aspects of my work: working with kids: YES!, paperwork: NO. If I didn't need the salary, which part to you think I'd do for free --just for the joy and challenge of it? OTOH, I can also recall times when, as a kid in the midst of coloring say, a beautiful universe full of mermaids and star-trees, a teacher who gave me evaluative praise - made me feel contaminated. She had inserted herself into my creative process in a way that interrupted (for a moment) my deep enjoyment of it. :-)

[rushdoggie]

You wax poetic, Geonni . . . and perhaps idealistic.

 

This may be true for some dogs, but certainly not all. There are plenty of breeds and types of dog with a much more independent bent who really aren't motivated to work for their handlers in the way you describe.

 

Yep.

 

I have 3 dogs, they are all different.

 

1. BC boy: really only cares about getting it right as a win ("I figured it out!" ) and making me praise him. He would turn his nose up at most cookies for a touch from me or a chance to try again. If we play a game where I use a clicker to shape a behavior, 3/4 times he either swallows cookie whole or lets it fall from his mouth, his reward is the noise that told him he figured it out. I have owned 6 BCs and 5 of them were similar to this.

 

2. Pap-zilla boy: likes the idea of work, enjoys activities with me, likes praise and touching. Gets stressed easily if he can't figure it out, not as quick sometimes to work it out as BC (naturally). Is rewarded most by movement and speed then by cookies. Would choose cookies over praise any day. I weaned him off cookies for challenging precision obed behaviors by teaching spins, weaving between my legs and heeling in circles (he has to go fast).

 

3. Old Tiny girl Pap: would leave me for anyone who has the food. Prefers to not be touched except in limited circumstances initiated by her, hates hands on her head or face. Praise is nice, but not worth much. Wants to know at all times, "whats in it for me." Has a very reliable recall in 95% of circumstances but at age 12 sometimes still drags a long string because those circumstances do not include not bothering cats or other people who she knows have food.

 

All 3 are generally well behaved calm house dogs, all 3 trained similarly except the BC who has done stockwork.

[ourwully]

The downside- is when people don't fade treats out for behaviors once they're learned, behaviors that have become "easy." Continuing to treat those becomes superfluous.

 

This has been my downfall. For example, I taught Juno to turn left and right on command. She got this behaviour long ago but I still give her treats. I will start to randomize the rewards aiming towards eliminating them completely for known behaviours. It will be interesting to see if she continues to comply.

 

Bill

[Root Beer]

Great comments! I just wanted to ask about praise- Is it intrinsic or extrinsic?

 

Fenzi seems to see praise as more of an "intrinsic" (at least compared to food); because praise is related to the dog's internal drive to please... I guess I had been thinking of praise as an external form of feedback/reward...? Which is it?

 

In my experience, this depends on the dog.

 

As I mentioned, Tessa naturally responds to praise and it seriously motivates her. That comes from within.

 

Dean isn't like that. If I chatter too much, even praise, he gets nervous. He does like applause, which I guess amounts to the same thing . . . but then, it isn't.

 

I could load up praise with treats and work to make it a secondary reinforcer. But I've never felt it worthwhile.

 

Some dogs are naturally motivated by praise, but others simply aren't.

[Root Beer]

The downside- is when people don't fade treats out for behaviors once they're learned, behaviors that have become "easy." Continuing to treat those becomes superfluous.

 

Yes, it does.

 

However . . . I want to say something about this that is really different from anything I've ever heard anyone else say.

 

When I trained my first dog, Speedy, I had no idea whatsoever how to fade treats.

 

Add in to that the fact that he had "issues" that made it so he actually needed the direct presence of food (only thing I could use since toys overstimulated him) much more than the average dog does.

 

But I competed with him, so the point came where I needed to start reducing the use of treats with him to some degree.

 

OK, here's what I want to say:

 

Even though I had - absolutely superfluously - lavished treats on him (one treat for every correct response, and most incorrect responses!) for years, when the time came when I needed to pare down the use of treats, I was able to do so.

 

How can that be?

 

Because when the time came to reduce reinforcers, I approached it as teaching him something new.

 

I wasn't thinking in terms of "getting rid of the food". I think a lot of new trainers get into trouble when they think in those terms because they end up trying to rush, or they raise criteria way too high. Instead, I was thinking in terms of "new behavior".

 

Personally, I err on the side of using too much food (or toys) for too long because I know that I can - very effectively - reduce the use of such reinforcers when the appropriate time comes to do so.

 

And if their use is superfluous for a time . . . so what? As long as we get the desired results in the end, it's all good.

 

Granted, having experience with this now, I handle it smarter. But there are still times when I need to make a judgment call in this regard. When in doubt, I provide the extra motivation.

[Donald McCaig]

Dear Doggers,

Bill wrote: "A quick glance at the current threads will show that some members are having difficulty with basic things like walking on a leash while others seem to get dogs that are good to go right from day one!"

 

A lot of that has practically nothing to do with training theory and or practice. If your body language makes sense to the dog it doesn't matter what silly ideas you maintain.

 

I can't tell whose dog is going to go well when handler and dog walk to the post but I can tell who will do very badly. And I'm not even a dog.

 

Donald McCaig

[ourwully]

I can't tell whose dog is going to go well when handler and dog walk to the post but I can tell who will do very badly. And I'm not even a dog. - When I was working I spent a lot of time observing teachers and ,Donald, what you have said is absolutely true when applied to teachers, and probably people in all types of endeavors, as well. My job was to help teachers improve, but on the occasions when I could predict poor performance with a single glance, it was a difficut task indeed. Body language was the tip off but under the surface was usually a minefield.

I wonder if this explains why so many people who obviously love and care for their dog so much have difficulty with their dogs. The dogs just know they can get away with anything!

Because when the time came to reduce reinforcers, I approached it as teaching him something new.

I wasn't thinking in terms of "getting rid of the food". I think a lot of new trainers get into trouble when they think in those terms because they end up trying to rush, or they raise criteria way too high. Instead, I was thinking in terms of "new behavior". - Root Beer, Could you give me a practical example of how this worked?

thanks

Bill

[denice]

I do think there is a large difference in breeds what works for them and what it is we are 'training' or asking of them. There are also always individual personalities - theirs and ours - that come into play.

 

I do believe what needs to be center no matter what is Respect, Trust - for us and from us and Consistency. If we have those three things I think you can do about anything.

 

If you raise your 8 wk 12 lb pup to be a 50 lb dog you be more cognizant of treating him and educating him so life will be good at 50lbs rather than suddenly changing the rules. If as an 8 wk old you trim nails weekly just by asking him to be still and keep trimming if he squirms and wiggles or complains making him hold relatively still. Letting him know those moments he is doing well and not so well by verbal information and tone then at 6 mo and 4 yrs and 15 nail trimming is just matter of fact. They might not like it but no big deal. If you teach them to stay close and not pull when young most do well with simple reminders as they grow and become more independent.

 

Sure treats can used to reward behaviors and are great at instant gratification, especially for pups whos attention span is limited. But along with treats should be praise then you can easily phase treats out quickly leaving reward as praise.

 

I think there is a huge difference in the dogs mindset if they are working for approval/ to please you or if working for food and self gratification. If for food they can tell quickly if you have any and some will simply decide it is not worth it. IF you have a respectful relationship they are more than likely to Choose to do as asked because they enjoy you being pleased or do not want you to be displeased. Yes I think they make choices as to what to do in certain situations.

 

Stock work takes things to another level. It is about instinct, skill, and gaining confidence but still goes back to respect and trust.

[Cass C]

In my opinion there are certain behaviors like recall that I will always treat and reward like it was brand new, because I don't want my dog thinking twice when there comes a time I need to use it without treats I want the dog to think it would still result in a treat party no matter what. I know some people think a dog should just recall because you said so, but I feel that by always rewarding it the recall will be so heavily reinforced that the dog won't hesitate for even a second whether or not to come no matter what is going on. However things like sit I don't always reward, but occasionally I will if I'm behavior chaining. That is mainly to enforce each step is a sperate behavior.

 

I have seen dogs that will only preform if food is present. My grandma has a toy poodle that won't do anything unless food is being shown to him. It is the most frustrating thing ever. In my opinion that type of dog needs to definitely have treats faded.