what do you use for treats?

[Eileen Stein]

Derek Scrimageour has observed "These dogs have been bred for hundreds of years to work with humans." We've asked our dogs to do all sorts of difficult, dangerous, unappealing things and - by and large - they've done them.

 

I think of treats as "dog money". Why offer money to a priest?

 

For years, when this subject comes up on the Boards, I've encouraged people to just try training some basic things without treats. Just try it. See how your dog responds. See the great potential these dogs have for working with you, see the pride and enjoyment they have in figuring out what you want and doing it, without dragging in any extraneous rewards. Wouldn't you like to develop and build on this innate capacity our dogs have for wanting to understand us and work with us? Just try it and see how satisfying it is!

 

What I've learned over time is that what I'm saying just doesn't compute for most performance trainers, and probably rightly so. They are preparing dogs for a future where they want to be able to condition a succession of precise and often tiny behaviors, none of which are particularly rewarding for the dog in themselves, or even make sense to the dog. They believe -- and for all I know, they're probably right -- that the best way to train behaviors like these is to click at the exact moment the dog does something right (or closer to right than he has in the past) and then give him a treat of some kind. And if that's how the dog's career training is going to be done, then it makes sense to do ALL the dog's training using this method. After all, click/treat training doesn't come natural to a dog. Your dogs and mine would be bewildered if we suddenly started trying to train in that way. The training method itself has to be taught; the vocabulary and syntax of click/treat has to be built up as the mode of communication between you and your dog. And it's probably going to work better if that's the method you use right from the start, and exclusively. From the c/t trainer's point of view, why mess around trying to train a dog to do anything without treats? Every piece of training you do without treats is a missed opportunity to practice and improve your fluency in the language of click/treat.

 

For what I want to do with my dogs, treat training makes no sense. It leaves undeveloped one of the capabilities I value most in the border collie. (No, not the ability to work livestock -- though there's a parallel there -- but rather the willingness to work solely for the intrinsic pleasure of the work and the partnership.) That's so much the essence of the enjoyment I get from working with these dogs that bypassing it seems like an impoverishment to me. But I can see why it doesn't seem like an impoverishment to them. They are as fascinated by developing their skills in c/t as I am in developing my skills in communicating without c/t. To each his own, I guess.

 

But I do wonder whether the failure to select for this particular capability in our dogs will result in its gradually being lost in the treat-trained population.

[Root Beer]

For what I want to do with my dogs, treat training makes no sense. It leaves undeveloped one of the capabilities I value most in the border collie. (No, not the ability to work livestock -- though there's a parallel there -- but rather the willingness to work solely for the intrinsic pleasure of the work and the partnership.) That's so much the essence of the enjoyment I get from working with these dogs that bypassing it seems like an impoverishment to me. But I can see why it doesn't seem like an impoverishment to them. They are as fascinated by developing their skills in c/t as I am in developing my skills in communicating without c/t. To each his own, I guess.

 

I think that one thing you might not realize is that with clicker training, a working relationship between dog and handler is fostered. I don't mean working in the sense of livestock here, but working in the sense of both dog and handler knowing one another as individuals, respecting one another as individuals, accomplishing things as a team that is joined by a deep bond of trust, affection, respect, and joy.

 

This type of training does foster pleasure of the work and the partnership - that end is simply reached by a different means. The dog's innate capacity for wanting to understand us is developed and built upon. If that were not the case, this type of training would actually not work.

 

The click/treat, or food without the click, is a means of communication, just as spoken language is a means of communication. In fact, it could be described very richly as a language between the dog and handler. The click/treat itself is not the partnership, nor does it substitute for one. It is simply a mode of communication through which the dog learns. And, when done right, the handler learns - even more than the dog in my experience - since in clicker training the dog is just as much a teacher as a learner.

 

I think that is something that an observer from the outside who has not really explored reinforcement based training to a great extent might not be able to see. The richness of clicker training does not come merely from a fascination with developing skills, although that's a huge perk, but with the depth of the bond that is fostered between dog and handler through this type of training.

 

Just as you find training with food reinforcers to make no sense, to me omitting this from my training would make no sense. It would be like trying to write a book with only rocks in the dirt. I could do it, certainly, but I could write the same book faster and better with a computer. The essence of what is being done is the same - I am expressing my thoughts and ideas in written form for others to read - but a computer is a much more appropriate tool with which to do that effectively. (Clarification: I am not comparing your training with rocks in the dirt - the comparison refers only to the training that I do with my own dogs. So, I would find training a dog to back up in heel six steps, do four leg weaves, and then move forward in a straight line with me turned sideways without use of food or clicker to be like writing a book with stones in the dirt. It could be done, but I would probably find it far too cumbersome to bother.)

 

So, the idea that use of food and clickers somehow bypasses pleasure for the work and partnership is to me similar to the idea that speaking Italian with a friend instead of English would somehow bypass true friendship. Speaking Italian with a friend who speaks Italian would serve to deepen the friendship. Training a dog through use of food and clickers opens lines of communication between dog and handler that deepen the dog/handler bond and create a very strong working partnership.

[Alchemist]

Wow, beautifully phrased, Kristine!

 

I just bought a clicker today in preparation for the "Control Unleashed" course we're going to start this week, and spent some time this evening "charging" it for Duncan. I found I was watching him closely to see whether he was "getting it", so your comments about speaking a common language really resonate. He is very anxious to please; this just makes it easier for me to signal to him that he's "gotten it" and that we're on the same page. But of course the sorts of things I'm working on with him at 10 PM in the family room (after spending the day helping out on the kids' robotics team) is as alien as you can imagine from the sorts of things a Border collie is "hardwired" to want to do.

[Root Beer]

I just bought a clicker today in preparation for the "Control Unleashed" course we're going to start this week, and spent some time this evening "charging" it for Duncan. I found I was watching him closely to see whether he was "getting it", so your comments about speaking a common language really resonate. He is very anxious to please; this just makes it easier for me to signal to him that he's "gotten it" and that we're on the same page. But of course the sorts of things I'm working on with him at 10 PM in the family room (after spending the day helping out on the kids' robotics team) is as alien as you can imagine from the sorts of things a Border collie is "hardwired" to want to do.

 

I can't wait to hear how that class goes for you!

 

One of my most treasured memories of Dean's early training came about when he "got it" and that line of communication was opened.

 

We were playing the game where the dog learns that he gets the treat by choosing (on his own) to leave the treat in the hand alone. This was a dog who was quite hardwired to dive into things - food bowls, counters, etc. The concept of choosing to leave a treat alone by his own choice was about as alien to him as alien can be!

 

Since he could not get at the treat by trying the things that came naturally to him, he backed off for a second to try to figure out how he could get it and . . . click. His eyes immediately flew up to meet mine, and the expression was a plain as the sun on a cloudless summer day - he understood. And not only did he understand, but he was absolutely delighted. In that moment, it was not about the food or the click, but it was all about the fact that he instantly knew exactly what I wanted and his first response was to look to me. A working partnership had begun. His pleasure in this new partnership was clear, and mine was to him, as well.

 

I think that sometimes it is easy to get the impression that because we use behavior science to explain and better understand why this kind of training is effective, and to create strategies to use it more effectively, that the whole process of working with the dog is somehow "scientific" and cold and impersonal. If that were the case, it would indeed be an impoverished approach. That is far from the case! And I think that some get the impression that because we use food (or toys or the environment, etc) to communicate with the dog throughout the training process, that we are merely bribing or distracting or circumventing the dog's ability to think. If that were the case, it would indeed be an impoverished approach. That is also far from the case!

 

At the heart of this approach is the dog's desire to be a working partner. It is all about communication between dog and handler, mutual learning, and accomplishment.

[Eileen Stein]

The click/treat, or food without the click, is a means of communication, just as spoken language is a means of communication. In fact, it could be described very richly as a language between the dog and handler. The click/treat itself is not the partnership, nor does it substitute for one. It is simply a mode of communication through which the dog learns. And, when done right, the handler learns - even more than the dog in my experience - since in clicker training the dog is just as much a teacher as a learner.

 

I think that is something that an observer from the outside who has not really explored reinforcement based training to a great extent might not be able to see.

 

If you reread what I wrote, you'll see that I did say c/t was a mode of communication, a language. Of course I recognize that. And I think that in any training, the dog is as much a teacher as a learner. If you cannot read your dog, you cannot train; in the effort to read your dog, you are learning.

 

The richness of clicker training does not come merely from a fascination with developing skills, although that's a huge perk, but with the depth of the bond that is fostered between dog and handler through this type of training.

 

If what you're saying here is that c/t fosters a deeper bond between dog and handler than other types of training, that is something you will never convince me of. To the extent that you have a working relationship with your dog that embodies a deep bond of trust, affection, respect, and joy (and I believe you do), it arises IMO from the innate strong desire and pleasure of border collies to work as a team with their persons -- something I have always proclaimed and never disputed -- and from the feelings those dogs inspire in us. It does not come from your training method. If it did, Sea World trainers would have a bond of the same depth and intensity with the seals they flip fish to, and I don't believe that they do.

 

Just as you find training with food reinforcers to make no sense, to me omitting this from my training would make no sense. It would be like trying to write a book with only rocks in the dirt. I could do it, certainly, but I could write the same book faster and better with a computer. The essence of what is being done is the same - I am expressing my thoughts and ideas in written form for others to read - but a computer is a much more appropriate tool with which to do that effectively. (Clarification: I am not comparing your training with rocks in the dirt - the comparison refers only to the training that I do with my own dogs. So, I would find training a dog to back up in heel six steps, do four leg weaves, and then move forward in a straight line with me turned sideways without use of food or clicker to be like writing a book with stones in the dirt. It could be done, but I would probably find it far too cumbersome to bother.)

 

If I were training seals at Sea World, I would definitely use c/t. It would be ideal for that. Which is probably why I have no desire whatsoever to train seals at Sea World. And as I said in my post, I accept that it's also probably the most efficient way to teach behaviors like your "back up in heel six steps, do four leg weaves, and then move forward in a straight line." I can understand your feeling that it would be "too cumbersome to bother" training that without the use of c/t -- but I can understand it only if you value the end result over the process. And frankly, it's hard for me to understand why the end result of having a dog back up in heel six steps, do four leg weaves, and then move forward in a straight line HAS any value in itself. To me the only point of training something like that would BE the process. (In that respect, your book analogy is not a good one IMO, because there the finished product is the whole point, not the process of writing it.) And the training process would be deeply unsatisfying to me if I were to use a means which shifted my dog's focus from the intrinsic rewards of the task we're addressing to what he needs to do to get an extraneous treat. Yes, in both cases he is thinking, but to me there's an important difference between thinking "What does she want?" and "What do I need to do to get a treat?"

 

So, the idea that use of food and clickers somehow bypasses pleasure for the work and partnership is to me similar to the idea that speaking Italian with a friend instead of English would somehow bypass true friendship. Speaking Italian with a friend who speaks Italian would serve to deepen the friendship. Training a dog through use of food and clickers opens lines of communication between dog and handler that deepen the dog/handler bond and create a very strong working partnership.

 

Just as I can't accept that c/t opens lines of communication between dog and handler that deepen their bond, I also can't accept that adopting c/t as your training method is like speaking Italian instead of English. C/t training introduces a distinctly different dynamic into the communication between dog and handler in a way that two friends shifting from English to Italian does not.

 

Science tells us that learning theory applies to all, humans included. Yet very few people -- surely only a handful of Skinnerian extremists -- would use c/t with their toddlers and very small children. Why not, if it's such an effective and scientifically valid way of training behaviors? I think that most people would recoil from the idea. Regardless of how effective it might be theoretically, and perhaps even practically, they would find it sad, disrespectful, reductionist, impoverishing. If you can understand such a feeling in that context -- whether or not you agree with it -- then perhaps you can understand why c/t holds no attraction for me in training dogs.

 

ETA: I didn't mean for this to come across as negative toward c/t as it probably does. I do see that you are fully happy with this method, and I don't want to seem to be disparaging that while explaining why it's not for me. My original post was intended to try to explain why, as I saw it, someone whose training was focused on performance events like freestyle might routinely use treats.

[mum24dog]

Yes, in both cases he is thinking, but to me there's an important difference between thinking "What does she want?" and "What do I need to do to get a treat?"

 

How do you know what your dog is thinking? How can you be sure that it isn't simply "What do I need to do for her to get to go and work the sheep?"

 

A reward is whatever the dog finds reinforcing enough to make it repeat the behaviour and the reward of choice is whatever works on the circumstances.

 

You rightly identify the difference between innate behaviour and teaching a behaviour that the dog is not programmed to perform, but that doesn't mean that the process of learning is any different.

 

“External reinforcement” is the way dogs are normally trained: rewards are delivered in the form of a treat or game, even just a pat or a word of praise.

 

“Internal reinforcement” is when the brain gives the body a feeling of pleasure.

 

Each part of the inherited herding/hunting sequence is internally reinforcing. Dogs don’t need a treat as a reward for performing it; they do it out of sheer pleasure. Their brain chemistry changes and they get a buzz of dopamine every time they perform an inherited motor pattern.

 

Manipulation of the human part of the equation so the dog gets a chance to create that buzz is more likely to be nearer the mark than imputing a desire to please us and work as a team - those are human attributes. Of course it does make it easier for the dog if it learns to read us so that it gets more opportunity to perform the behaviour, but it doesn't mean that understanding us and "working as a team" is the object of the exercise for the dog.

 

Seeing the symbiotic relationship between man and dog as a "team" in human terms is rewarding to the human involved.

Mutually rewarding, yes. I guess it depends on your personal definition of a team.

 

Some may call me a cynic, I prefer to be thought a realist.

 

Science tells us that learning theory applies to all, humans included. Yet very few people -- surely only a handful of Skinnerian extremists -- would use c/t with their toddlers and very small children.

 

Not literally, but lots of people use the same principle without being conscious of it.

 

I wish I'd read "Don't Shoot the Dog" by Karen Pryor before I had children; I wouldn't have made so many mistakes along the way. And operant conditioning works in any humam interraction - even with husbands. Now there's a test for you - train a husband without using positive reinforcement.

 

Regardless of how effective it might be theoretically, and perhaps even practically, they would find it sad, disrespectful, cold, reductionist, impoverishing. If you can understand such a feeling in that context -- whether or not you agree with it -- then perhaps you can understand why c/t holds no attraction for me in training dogs.

 

I can understand why you think the way you do because you have presumably never had need to do other than what works for you in your own context. You have admitted that if you had to train for other things you would probably use another method, just as I would be unlikely to use treats and a clicker if I were training my dog on stock.

 

In a smaller way, I rarely need to use them for our BC in agility. Even a game with a ball has become more part of the routine than reinforcement in itself as the chance to run and jump has become self rewarding for him. It hasn't affected our relationship at all; he still gets his reward for cooperating with me (or more often my daughter).

 

Rather than "sad, disrespectful, cold, reductionist, impoverishing", to me the way I view the world, and not just in terms of training my dogs, is simply an attempt to understand what is happening and what I am doing. "That's just the way it is" doesn't cut it for me. Analytical, yes; cold and the other stuff, no.

 

Pam

[Root Beer]

If what you're saying here is that c/t fosters a deeper bond between dog and handler than other types of training, that is something you will never convince me of.

 

That is not what I am saying.

 

Rather, I am responding directly to the idea of reinforcement based training (of which clicker training is one type) as an impoverishment.

 

It seems as though there is an idea that use of external reinforcers somehow blocks the capacity for the dog and handler to form a deep, vital, and highly satisfying working partnership. And this could not be further from the truth.

 

What you see as impoverishment is one of the richest, most satisfying, and wonderful aspects of my own life - the working relationship that I have with my dogs. And I'm not some really odd exception to the rule. My peers who use the same sorts of methods to train for the same sorts of activities with their dogs find the same thing to be true. And the expressions on their dogs faces as they interact - both on and off the course or in and out of the ring - attests to the depth of the partnership that is there. These aren't dogs going to extreme lengths to try to get food! They are dogs who find true joy in working as a team with their handlers.

 

As to other types of training, that really wasn't what I was talking about. My point was not to draw a comparison, but to expand upon on the basic question of "why use treats in training?"

 

To the extent that you have a working relationship with your dog that embodies a deep bond of trust, affection, respect, and joy (and I believe you do), it arises IMO from the innate strong desire and pleasure of border collies to work as a team with their persons -- something I have always proclaimed and never disputed -- and from the feelings those dogs inspire in us. It does not come from your training method. If it did, Sea World trainers would have a bond of the same depth and intensity with the seals they flip fish to, and I don't believe that they do.

 

Well, I said the same thing. Earlier I said, "The dog's innate capacity for wanting to understand us is developed and built upon. If that were not the case, this type of training would actually not work."

 

We don't want dogs to be like performing seals. We want them to be dogs - with all of the depths that their individual "personalities" bring to a working partnership.

 

Your statement above makes me wonder . . . do you think that the fact that c/t is used on seals renders the technique useless on other types of animals? Do you see use of c/t with dogs as producing seal-like qualities in dogs because it is used with seals?

 

I'm not asking to be snarky. I'm really interested in your perspective on those questions. I think it might help me to understand where you are coming from a bit more.

 

If I were training seals at Sea World, I would definitely use c/t. It would be ideal for that. Which is probably why I have no desire whatsoever to train seals at Sea World.

 

I don't, either. I have no desire to train any animal whatsoever other than dogs. I'm just not that interested in other types of animals.

 

And as I said in my post, I accept that it's also probably the most efficient way to teach behaviors like your "back up in heel six steps, do four leg weaves, and then move forward in a straight line." I can understand your feeling that it would be "too cumbersome to bother" training that without the use of c/t -- but I can understand it only if you value the end result over the process. And frankly, it's hard for me to understand why the end result of having a dog back up in heel six steps, do four leg weaves, and then move forward in a straight line HAS any value in itself.

 

I could say the same for many things that people choose to do with their dogs.

 

What value might that have to me? Once the behavior is trained, there is the simple enjoyment of performing those moves with my dog as a team. It might not be your cup of tea, but I find such activity extremely enjoyable. So do my dogs.

 

Beyond that, my dog and I can provide entertainment to people in nursing homes, or to children, or at events that raise money for rescue or service dogs or other types of fundraising.

 

My dog and I can also take such moves into competition and work together toward performing to certain standards. Also not everyone's cup of tea, but definitely something that my dogs and I enjoy doing together.

 

So, yes, those results - the results of the dog learning literally hundreds of distinct moves and behaviors - has great value to me. And it has great value to my dogs because learning those things provides them with opportunities to work together with me as a partnership and team in a lot of very different situations. They truly enjoy that. Especially my Border Collies, who particularly enjoy anything that they perceive as having a special purpose.

 

Yes, in both cases he is thinking, but to me there's an important difference between thinking "What does she want?" and "What do I need to do to get a treat?"

 

The two are not mutually exclusive. Since the treat (or click/treat) is a means of saying "Yes! That's correct", then ultimately we are working in terms of "what does she want", not "how do I get a treat".

 

Yes, there is a point - early in training - where the food is specifically used as a motivator. Over time, however, a point comes where understanding of both the specific behaviors, and the overall concept of working as a team, takes the dog beyond that point.

 

I see this most in the end result. When my dogs perform or run Agility courses, or simply sit and wait someplace because I cue them to do so, there is no food involved! If it were all about the food - and only about the food - the training would not hold up once the food disappears.

 

98% of my time with my dogs is spent without treats on me. They still come when they are called, sit and wait, back away from things, etc. This is because they have learned what it is that I want when I give certain directives. Not because I produce food every time I ask them to do something?

 

See what I mean?

 

Just as I can't accept that c/t opens lines of communication between dog and handler that deepen their bond, I also can't accept that adopting c/t as your training method is like speaking Italian instead of English. C/t training introduces a distinctly different dynamic into the communication between dog and handler in a way that two friends shifting from English to Italian does not.

 

As an analogy, it does limp. Analogies are never perfect. Yes, there is a specific dynamic in training with a clicker. If there weren't, I probably wouldn't bother with it!

 

But it is, at it's core, communication. It tells the dog when he or she has done what I want.

 

Science tells us that learning theory applies to all, humans included. Yet very few people -- surely only a handful of Skinnerian extremists -- would use c/t with their toddlers and very small children. Why not, if it's such an effective and scientifically valid way of training behaviors? I think that most people would recoil from the idea. Regardless of how effective it might be theoretically, and perhaps even practically, they would find it sad, disrespectful, reductionist, impoverishing. If you can understand such a feeling in that context -- whether or not you agree with it -- then perhaps you can understand why c/t holds no attraction for me in training dogs.

 

Learning theory is not just about reinforcement, though. Any person who punishes a child with a spanking or a time out is applying the exact same learning theory. Aren't such things accepted as completely normal by most people? And while most people don't use clickers, don't all parents - to some degree - reward their children for particularly good things that they hope they will repeat? My Dad used to give me $1.00 for every A I brought home on my report card. Does that idea make you recoil?

 

Clicker training is based on learning theory. Learning theory is not synonymous with clicker training. All application of learning theory does not involve clickers.

 

To bring this back to dogs, when you use a correction, you are using the exact same learning theory. Do you recoil equally from the idea of using a correction? You should if the main reason why you object to clicker training is that the mechanics of it can be explained by learning theory.

 

ETA: I didn't mean for this to come across as negative toward c/t as it probably does. I do see that you are fully happy with this method, and I don't want to seem to be disparaging that while explaining why it's not for me. My original post was intended to try to explain why, as I saw it, someone whose training was focused on performance events like freestyle might routinely use treats.

 

And please don't get the impression that I am saying that you should change the way you train.

 

My responses have been intended also to explain my perspective.

 

I guess I just want people to understand that I don't train with treats only for utilitarian reasons. Yes, my goals for my dogs do influence the way I handle many training situations. But ultimately I choose to train the way I do because I find that the working partnership between my dogs and myself is incredibly rich and satisfying.

 

And I think that most who train with reinforcement would agree that impoverishment (of relationship, bond, partnership, etc) is not a part of the picture by any stretch of the imagination.

[Eileen Stein]

How do you know what your dog is thinking? How can you be sure that it isn't simply "What do I need to do for her to get to go and work the sheep?"

 

Well, I have a better chance of knowing than somebody who has not worked sheep with a dog, wouldn't you agree? And I train other things besides stockwork, though I don't train for competitive dog sports. In that other, nonstock training, where I'm asking for a new (non-instinctive) behavior, I frequently see the dog trying different things and looking at me excitedly ("Is this it? Is this it?"), and her manifest delight when I tell her that IS it. No treats involved.

 

Each part of the inherited herding/hunting sequence is internally reinforcing. Dogs don’t need a treat as a reward for performing it; they do it out of sheer pleasure. Their brain chemistry changes and they get a buzz of dopamine every time they perform an inherited motor pattern.

 

I'm sorry, but it really gets to me when people who don't train sheepdogs explain to me that what I do is essentially the same as treat training but with the reinforcement being "getting to work the sheep" or "performing an inherited motor pattern." I know it's how you make what we do fit with behavioral theory, but it is simply not accurate. To the extent that being allowed to work the sheep is a reward (which it often is), it is an intrinsic reward of the endeavor, not an extraneous reward. But leaving that aside, because of the nature of sheep, dogs and humans, it is not possible to give and withhold the sheep as precisely as one would a treat. Moreover, we often need to train the dog NOT to "perform the herding/hunting sequence" as his instincts tell him to, and NOT to perform an "inherited motor pattern" (Gah!). In doing so we are preventing him from getting his "buzz of dopamine." Finally, as I said on a previous thread, there are many times when not letting him have the sheep when he's wrong would cause negative consequences. Hence the routine observation that beginners rely too much on the down -- if it were simply a matter of withholding "the reward" when the dog is wrong, then it would always be right to down him. Another example -- a dog who starts out wrong on his outrun can be stopped and called back, and thereby not allowed to have the sheep, but if you do that more than a very few times you risk creating other outrun problems. And how do you deal with clappiness (lying down and staring at the sheep instead of walking up, resulting from too much eye) by withholding the reward? I could give countless other examples. It is just not that simple. There is more to it than that.

 

ETA: I now see that mum24dogs has taken the quoted material, without attribution, from an article on why dogs chase things and how to stop them doing it, but has added the word "herding" to the quote. Chasing is not herding; dogs being trained to work stock are not permitted to chase.

 

Manipulation of the human part of the equation so the dog gets a chance to create that buzz is more likely to be nearer the mark than imputing a desire to please us and work as a team - those are human attributes.

 

If you have never seen in border collies a desire to work as a team with their person, a pleasure in figuring out what their person wants and doing it -- if you think they can't feel such things because only humans can feel that way -- then I fully understand your allegiance to c/t. But that's why threads where c/t is presented as the only training method worthy of consideration disturb me so much -- concern that someone who follows that advice may never get to see and appreciate this outstanding characteristic that has been developed in our breed.

 

 

Science tells us that learning theory applies to all, humans included. Yet very few people -- surely only a handful of Skinnerian extremists -- would use c/t with their toddlers and very small children.

 

Not literally, but lots of people use the same principle without being conscious of it.

 

I'm talking about literally, not vaguer concepts. Click/treat training, the distilled essence of positive reinforcement. If it works so well, why NOT use it on your toddlers?

 

I can understand why you think the way you do because you have presumably never had need to do other than what works for you in your own context. You have admitted that if you had to train for other things you would probably use another method, just as I would be unlikely to use treats and a clicker if I were training my dog on stock.

 

No, I said that if I had to train a seal, or train a dog to do freestyle dance steps, I would probably use c/t. I certainly don't use c/t to train a recall, or a stay, or agility (backyard stuff, not competition), or other general dog things.

 

Rather than "sad, disrespectful, cold, reductionist, impoverishing", to me the way I view the world, and not just in terms of training my dogs, is simply an attempt to understand what is happening and what I am doing. "That's just the way it is" doesn't cut it for me. Analytical, yes; cold and the other stuff, no.

 

Don't really understand what you're saying here, so I'll assume it wasn't meant as an insult. I think I'm as analytical as the next guy, and I probably have as much interest in understanding what is happening and what I am doing as you do. I used the words you quote to describe what feelings people would probably have toward training small children with c/t, not to comment on your world view or to champion a know-nothing attitude.

[Root Beer]

I'm talking about literally, not vaguer concepts. Click/treat training, the distilled essence of positive reinforcement. If it works so well, why NOT use it on your toddlers?

 

LOL!! Do you think I wouldn't? Really? Not with an actual clicker, but some sort of reward based system for teaching skills to a young child? Chances to earn playtime, etc. as incentives in learning.

 

I mean, really - do toddlers typically do everything they are told because they have a strong desire to please their parents? External motivation seems like a no brainer with a toddler.

 

Sadly, I don't think I will ever have the opportunity, but I would have a lot of fun, I think, with that. Wouldn't harm the kid any to learn that there is often some kind of payoff for going the extra mile to do a good job. It's a lesson I wish my own mother had taught me at that age.

[juliepoudrier]

And while most people don't use clickers, don't all parents - to some degree - reward their children for particularly good things that they hope they will repeat? My Dad used to give me $1.00 for every A I brought home on my report card. Does that idea make you recoil?

I sort of chuckled at this comment. I know it was directed at Eileen, but it sure brought back some childhood memories. Wasn't that thread in the Coffee Break (?) section directly related to this concept? Learning for the intrinsic reward of the learning vs. learning in order to gain some external reward? I was one of those kids for whom As were *expected,* no reward in sight. So yeah, it makes me recoil some now just as it did then. As I child I used to be somewhat annoyed that friends would get money for good grades on their report cards, but it didn't stop me from getting good grades anyway, and looking back as an adult I can see the wisdom my parents had in this regard. I *never* got an A just so I could get a dollar (I'm not saying that you did, but I sure knew plenty of kids who would never have applied themselves without the $$ motivator--and depending on your viewpoint I guess the motivator might be irrelevant as long as the grade was an A). I got As because I was proud of my intelligence and wanted to do well. And I wanted a higher education and recognized that one path to doing that was by excelling in school. I guess those could all be considered rewards in a way, but they were either intangible (I feel good about myself) or sometime in the future (a college education and career that followed). I'm sure that some of the kids who recieved money for good grades also went on to excel in post-secondary education and beyond, but I wonder how many of them retained much of what they learned in high school? So what I'm saying is I wonder what the difference in retention is between children who were internally motivated vs. those who were externally motivated (with money). And also their basic outlook on life. Would children raised in a reward system generally always expect a reward in exchange for a job well done? Would children raised like I was automatically try to do a job well for the person satisfaction of doing so? That would make an interesting study! I'm grateful that my parents taught me that learning for the sake of learning is reward enough. What it enabled me to do was go to college and then pursue a course of study that was difficult *precisely because it was challenging to me,* and not necessarily because I would get something (a reward) out of it. And I do wonder if you trained a child with c/t principles if you wouldn't raise someone with a "what's in it for me?" attitude, because, well, human brains work differently from dog brains. You acknowledge that you generally fade rewards with dogs, but I wonder if that would really work with children? I haven't raised any so I don't know. But I do wonder.

 

Like Eileen, it galls me a bit when folks who don't work dogs try to come up with ways to fit stockdog training into their own training paradigm. (I'd be willing to change my mind if someone could show me *scientific studies,* not theories and suppositions, that a dog doing an innate motor pattern is getting a hit of dopamine for doing so and is thus being rewarded each time it works stock. Actually I guess I'd first want to see the studies on "innate motor patterns" themselves. I'm not really sure what that means. I assume a pointing dog actually pointing is considered an innate motor pattern? What about when a border collie exhibits the exact same pointing behavior? Innate? Dopamine reward? As anyone who has started a dog or three on livestock would know, not all motor patterns exhibited at the start are appropriate, so how do you determine which ones are innate? Or is it by definition just the motor pattern you actually want? How do you explain the youngster that runs straight up the middle and through the stock when the innate motor pattern should be to go around? And if innate motor patterns like that existed, then wouldn't all border collies automatically have them, even if they had been bred for something else, like the show ring, for generations? Isn't that the exact argument the kennel clubs use to justify not breeding for work--that the innate motor patterns, and therefore the working ability, will remain regardless of the breeding choices being made?).

 

And I might argue, like Eileen, that I think c/t is largely unnecessary for breeds of dog--like the border collie--who have been bred to want to work with and please their human partners, no matter what.

 

J.

[kelpiegirl]

Someone wrote here, that you show a border collie something once, and it becomes a habit. I had this happen with Danny. I have video up. ONE TIME, while I was at the kitchen sink, Danny happened to drop the ball and it rolled to my feet. Seconds later, after I returned it, Danny repeated, and repeated. He was even smart enough, that when I gave him the cue (verbal) to make sure it dropped in kitchen, and not on the carpet, he did this, without me showing him. It's kind of amazing.

[ItsADogsLyfe]

My opinion on the c/t thing for toddlers. In raising my 3 kids (now all of adult age, last one just turned 18) I was never in the structured, only breast feed on a time schedule, potty train them young etc. I nursed my babies when they were hungry, probably sometimes when they were not. I never tried to potty "train" them. I showed them the toilet and showed them what it was for and when they were around 2 yrs or so they started using it reliably enough on their own to not require diapers. I liked diapers..so easy when you went somewhere. I never had to frantically find a bathroom to make sure my kid didn't soil their underpants. Every parent I knew who tried to train their child ended up either having their child using the bathroom reliably around the same age as mine, or sometimes even later, I think because of too much pressure put on the kid. I always figured as long as they knew how to use the toilet by the time they went to school they'd be ok. They've all grown up to be very successful accomplished people.

C/t trained dogs seem more mechanical to me. I like to work with my dogs the same as I did with my kids. I tell them, show them then let them know when they are right. Communicating without the technical aspect of a devise that makes a noise to mark the "behavior." All the technical terminology makes my head spin. I can't keep it all straight in my head so I have to go with what comes naturally to me when working with my dogs. If it feels right and I get the right reactions, I go with whatever works at that time. I'm sure the c/t has its place with certain people, but its just not for me with how I want my dogs to respond.

[Root Beer]

Learning for the intrinsic reward of the learning vs. learning in order to gain some external reward? I was one of those kids for whom As were *expected,* no reward in sight. So yeah, it makes me recoil some now just as it did then. As I child I used to be somewhat annoyed that friends would get money for good grades on their report cards, but it didn't stop me from getting good grades anyway, and looking back as an adult I can see the wisdom my parents had in this regard. I *never* got an A just so I could get a dollar (I'm not saying that you did, but I sure knew plenty of kids who would never have applied themselves without the $$ motivator--and depending on your viewpoint I guess the motivator might be irrelevant as long as the grade was an A). I got As because I was proud of my intelligence and wanted to do well. And I wanted a higher education and recognized that one path to doing that was by excelling in school.

 

And that actually makes the exact point that I was making about clicker training.

 

I certainly didn't work my butt off for a tiny little dollar per A. That would not have been worth it at all. Good grades didn't come easy to me, but I worked very hard for them - for the exact same reasons that you did! The fact that my father gave me money did not take anything away from my own internal motivation - and I certainly had plenty of that!

 

Still, the dollar was nice. It really showed me more that my father was happy with my efforts. It did not really serve as a motivator. Could have I have known he was happy just by him saying "good job?" Sure. Did it sweeten things a little to have extra spending money. Yeah!

 

Do my dogs know they have done what I want when I offer praise? Yes. Does it sweeten things a little when I give them food rewards. Definitely.

 

And I might argue, like Eileen, that I think c/t is largely unnecessary for breeds of dog--like the border collie--who have been bred to want to work with and please their human partners, no matter what.

 

Would you say that you hold the position that any externally offered rewards whatsoever (food, play, praise, access to things that the dog wants, etc) are unnecessary for Border Collies to learn?

 

To me I guess that idea is a strange as the idea that corrections are unnecessary for the Border Collie to learn is to you.

[Root Beer]

C/t trained dogs seem more mechanical to me. I like to work with my dogs the same as I did with my kids. I tell them, show them then let them know when they are right. Communicating without the technical aspect of a devise that makes a noise to mark the "behavior."

 

It might surprise you to learn that I do the same thing. Sometimes I use a clicker - especially when I am teaching something that I need to hold up reliably under stimulating and stressful competition conditions. But often I show my dogs things and then let them know they are right.

 

A clicker is just one way of showing a dog that he is right. It's really not that technical.

 

I'm not saying that you should use one. Just that in practice it is a lot less mechanical than it might seem to someone who has not used it to a great extent.

[Root Beer]

Someone wrote here, that you show a border collie something once, and it becomes a habit. I had this happen with Danny. I have video up. ONE TIME, while I was at the kitchen sink, Danny happened to drop the ball and it rolled to my feet. Seconds later, after I returned it, Danny repeated, and repeated. He was even smart enough, that when I gave him the cue (verbal) to make sure it dropped in kitchen, and not on the carpet, he did this, without me showing him. It's kind of amazing.

 

I'm curious. Do you find this with everything you teach your Border Collie? You show the dog one time, and the dog retains it for life? No reinforcement, no reminders, no corrections, no mistakes - perfect retention, perfect response to your directives every single time?

 

I'm not being sarcastic. I really want to know if this has been your true experience.

[juliepoudrier]

I don't think I proved your point at all. I think I raised some very valid questions about whether paying children to learn is truly effective. But again I would say without evidence to show that one method works better for children than another you can't generalize. Perhaps for you as an individual the dollar was the "icing on the cake." Can you say with certainty that it was the same for all your peers who were also paid for grades? I don't think you can.

 

And no, I won't state an absolute that rewards are completely unnecessary in order for border collies to learn; I would say, however, that they aren't completely *necessary,* and that's a different thing.

 

J.

[Root Beer]

I don't think I proved your point at all. I think I raised some very valid questions about whether paying children to learn is truly effective. But again I would say without evidence to show that one method works better for children than another you can't generalize. Perhaps for you as an individual the dollar was the "icing on the cake." Can you say with certainty that it was the same for all your peers who were also paid for grades? I don't think you can.

 

I still think you did. The point is that use of an external reward does not, by default, create some kind of universal impediment to fostering of internal motivation.

 

The point I was making is that use of food in training does not, by default, cut off any possibility of a working partnership - in all it's fullness - between dog and handler. In fact, in practice, there are countless rich working partnerships between handlers who train using food, and their dogs (including Border Collies). It's not even rare, unusual, or some odd exception.

 

As far as using children as a parallel, honestly, I don't really care a whole lot about what works and what doesn't for raising kids. The chances that I'm actually going to have any are slim. I guess it's interesting in theory, but for me it will most likely remain theory.

 

I know that I graduated high school with a 4.0 GPA, and I continued my education through a Master's Degree. I know, also that I have continued to enjoy learning into adulthood. I really can't speak for anyone else, and I don't really care to.

 

Dogs are another matter. I do train dogs. I train my own and I teach others how to train. So, I"m going to stick with dogs. I shouldn't have gotten into the children comparison because much of what applies to children is not applicable to dogs.

[Ooky]

I love Eileen’s responses to this thread, and this is coming from someone who uses c/t training in many situations and probably always will. But I think she breaks it down beautifully.

 

I started training my pup with c/t, with much success, to do all the typical “tricks” and commands you learn in a puppy class. However, when we got to adolescence and a need to put a lid on his overstimulated behaviors, I felt like c/t wasn’t “strong enough” to get the job done and turned to much more corrective methods, like collar corrections. However, I am not skilled at these, and basically spent a lot of effort on continued correctional “management” that didn’t translate into real success. The “desire” to work with me and for me that is an intrinsic part of BC nature, that both Eileen and Donald speak of so eloquently, was at its *least* apparent at this time – an adolescent pup, prepping for and then recovering from major surgery, and me attempting to use mainly “nagging” correctional methods to control his out-of-control behaviors. Going on advice from Kristine and others, I went back to c/t armed with a better understanding of its ability to help desensitize, teach self control, and communicate on a deeper level than is apparent when you use it in the trick-training, flipping-fish-to-a-seal sort of way, e.g., look at that games, mat work, etc. So then, c/t and a major reduction and change in the corrections I was using got us back on track.

 

However, I don’t think this is a condemnation of non-c/t methods, it just shows that my ability to train w/o it at the time was inadequate. And after seeing several stockdog trainers with Odin, my belief is that very few of them would have opted for these types of correctional methods either, and instead would have chosen a much more effective non-c/t training style that revolved around everything in his life just as much as the CU-style c/t methods (again, the mat work etc) do. In my mind where I went wrong was my own knowledge and skill level with dog training in general, and this idea that training only happened at certain times when it was most troublesome for me that he was out-of-control/overstimulated.

 

After beginning stockwork training, I became open to more correctional-heavy methods again, but I hate using that phrase because that makes it sound like all I was doing was correcting my dog. Eileen explains it so much better – what I started doing in a lot of areas of life, like boundary training, off-leash training, recalls, better manners, dog-dog interactions, basically everything that had to do with working towards him being a top-level trustworthy companion in urban and other environments, I started relying less and less on EXTRINSIC rewards such as treats, and more and more on experimenting with ways to tap into his intrinsic desire to want to please me and work with me. This process can use reinforcement with praise, or a toy, but also it often uses mild verbal corrections intended to impart valuable information – the “no” that I personally feel is absolutely vital to organic conversation of any kind, and pressure-release. While I am the rankest of beginners in stockwork, it is teaching me the incredible fluency and intricate uses of pressure-release, which is NOT just applicable to a “herding” situation, but can be used in so many other areas of dog training.

 

Learning how and when to issue corrections, and how and when to reward (not necessarily with food) can be powerful tools with these dogs for teaching MORE than just stockwork, and can apply to many pet-only type situations. I like that Odin’s recall was not food treat trained, and not because I believe you can’t train a good recall with treats (as I have seen it done). But to me, the process had more value because it focused on really tapping into his desire to want to work with me, and meet my expectations.

 

As for the pay-for-grades thing, let’s just say I agree completely with Julie. It does turn my stomach a little to think of using such a system with my own child, or using food as a training tool when she is a toddler, which I think is maybe closer to the dog training situation. And I wrote in the coffee break thread about the latest research showing a strong reliance on extrinsic rewards that are easily controlled ISN’T the most effective way to instill motivation in children. (Not that rewards should never be used). And that children who are never allowed to fail and hence be corrected are also often done a disservice. I think parents and teachers, at least good ones, naturally use all four sides of the +/- reward/punishment square, and may even use marker words like "Yes!" But it would be weird to see someone repeatedly saying Yes! to a toddler, then giving a cheerio, right? And that is the main point here in relation to dogs - not never using rewards, but specifically food treats metered out in the success=treat sort of way.

 

But dogs aren't children, and like I said I do use treats for some things. For a lot of stuff, I think c/t is great!! Especially for detailed tasks that aren’t inherently valued by the dog, i.e. freestyle moves (which I enjoy attempting to train, BTW), silly commands like “kisses” and “ears up”, or even more serious stuff like desensitizing a dog when it has become irrationally fearful. I still pull out the clicker or at least my treats for such problems. But I also think we (me and Odin) gain something – something that is so hard to describe -- from tapping into the intrinsic desire to work together with corrections and incorporating treatless foundational training even when not on stock.

[Root Beer]

As for the pay-for-grades thing, let’s just say I agree completely with Julie.

 

Wow! I didn't know I was going to strike such a nerve with some of you! To me it was so not a big deal. Just a nice thing that my Dad did to express approval in a special way.

 

I didn't realize it was so politically incorrect, but I'm also glad my Dad didn't know that when I was in High School! :D It is something I recall many years later with gratitude.

[Ooky]

I think it actually IS the politically correct way to do things now (pay for grades etc). I'm pretty sure the LAST thing my father was is PC - I even got *gasp* spanked! But strangely I don't think it damaged me in the least.

 

What do you think of the toddler-food analogy?

[ItsADogsLyfe]

I'm not saying that you should use one. Just that in practice it is a lot less mechanical than it might seem to someone who has not used it to a great extent.

 

I didn't mean the use of one was mechanical, I meant the dogs that I have seen trained using that method seem more mechanical and less natural to me.

 

Dogs are another matter. I do train dogs. I train my own and I teach others how to train. So, I"m going to stick with dogs. I shouldn't have gotten into the children comparison because much of what applies to children is not applicable to dogs.

 

I find this funny because even though I've raised 3 children and numerous dogs, I always seem to use a dog analogy when talking to my son about raising his two boys. I can relate everything in life to dogs.

Really I find no matter what method is used the most important thing is consistency.

[Maralynn]

I'm curious. Do you find this with everything you teach your Border Collie? You show the dog one time, and the dog retains it for life? No reinforcement, no reminders, no corrections, no mistakes - perfect retention, perfect response to your directives every single time?

 

I'm not being sarcastic. I really want to know if this has been your true experience.

 

It's been this way with Missy. I taught her to crawl, spin, roll over, jump up, in about 3 minutes each with both vocal and hand signal. I told her I didn't want her in the basement once and I think she's been down there maybe 4 times in almost 8 years. I started asking her to get here dish before feeding her a couple months ago, she knew exactly what I meant from the beginning. I tried using a clicker with her once, but it confused her, so I stuck to just telling her what I wanted.

 

The other two are a bit different, but I'm seeing signs that Kenzi might be very similar once she matures a bit.

[Root Beer]

What do you think of the toddler-food analogy?

 

In theory, if I had a toddler to raise, I would not do anything that looked like what you said here . . .

 

But it would be weird to see someone repeatedly saying Yes! to a toddler, then giving a cheerio, right? And that is the main point here in relation to dogs - not never using rewards, but specifically food treats metered out in the success=treat sort of way.

 

I'm talking about things like earning play time by learning to do age-appropriate chores. I think that's a much better parallel. Maybe earning a chance to watch a video or a chance to play with a very special toy (since some toys would be put up anyway due to the presence of dogs in the house!).

 

And, not for everything, but for the things that need, in particular, to be learned.

 

I was just in the living room watching some of the Olympics and I decided that I wanted Dean to curl up with me on the futon and snuggle. I said, "Dean Dog!" and he immediately jumped up and settled in. I did not need food, a clicker, toys, or anything external to get him to do that. I actually probably never used any external rewards in the first place to teach that.

 

So it's not like I'm repeatedly clicking and giving treats all day long for everything my dogs do! The clicker and treats are used to teach behaviors that they don't know, or do not know to do on cue in various situations, or that I need to stand up to the stress and stimulation of competition.

[Root Beer]

I didn't mean the use of one was mechanical, I meant the dogs that I have seen trained using that method seem more mechanical and less natural to me.

 

Would you do me a huge favor?

 

Would you watch this video and tell me in what way this dog is mechanical? What would indicate to you that she could be more natural? I'm really interested in knowing exactly what you see lacking here . . .

 

 

I'm not trying to be argumentative - just trying to understand.

 

And I'm not asking whether or not you find Agility itself to be unnatural!! What I'm asking about is this how the dog is mechanical and less natural than other dogs. What about her demeanor or body language or whatever indicates the lack of natural-ism?

[Blackdawgs]

As for the pay-for-grades thing, let’s just say I agree completely with Julie. It does turn my stomach a little to think of using such a system with my own child, or using food as a training tool when she is a toddler, which I think is maybe closer to the dog training situation. And I wrote in the coffee break thread about the latest research showing a strong reliance on extrinsic rewards that are easily controlled ISN’T the most effective way to instill motivation in children.

 

The irony of this discussion is that the pay-for-WORK thing is alive and well for us human adults. Our weekly (or biweekly) monetary reward for 40 hours of work is a powerful motivator, is it not? Or maybe I should work for free because my ancestors installed a powerful work ethic and desire to please my employer in me.