Jump to content
BC Boards

When should you lie to your Border Collie?


Recommended Posts

Maja, while many misuse food in training dogs (just as many misuse hitting, choke collars, shock collars etc), an understanding of how animals learn helps prevent problems with any training method.

 

As for Punishments for dogs trained via a food reward system, there are many (and necessary) and it depends much on the dog, handler and situation. Good training is not a "One Size Fits All". It is tailored to fit the situation.

 

Proper training using food as a reward SHOULD include weaning the dog away from continuous reinforcement and additionally using other rewards-including your voice.

 

However, dogs do NOT know words and therefore food can help the dog learn the praise words mean good things. Yes, tone of voice can be understood by dogs, but things can be sped up via training an association with a word which means good things-or bad things.

Link to comment
Share on other sites

  • Replies 131
  • Created
  • Last Reply

Top Posters In This Topic

On the matter of sending out a dog to look for non existent sheep, I went to a talk and demo by our local mountain rescue search dog team last night and one of the questions raised was how was the dog rewarded if it had searched an area and found nothing. The answer was that the handler wherever possible would set up a set up "find" for the dog using another member of the team so he could reinforce success. Just an example of always ending on a high note.

 

 

While this has been the common practice, there is another approach to this followed by by SAR trainers here. They want the search itself to be the reward so they have a dog that love searching for the sake of searching. With this in mind, they work on building the desire for the search. Once the dog is fully trained, they won't always set up successful searches. Most of the time they will but they'll throw some victimless ones in, too. This mirrors real life as statisically most real searches won't result in a find. I haven't done this for the most part, but it makes sense to me as a well trained dog seems to love the search as much as the find.
Link to comment
Share on other sites

Regarding sending a dog for nonexistent sheep, I pretty much never do that as a training technique. If I send the dog for stock, I want the dog to believe (that I believe) that stock is there. It's all part of a training continuum for me. If I make a mistake and send a dog where I believe stock to be and they are not there, the dog will return and I'll resend in a different direction. I believe that by approaching it this way, the dog understands that I think there to be stock out there, whether we can see them or not, and it will continue to look for them as thoroughly as possible. If I am wrong and the dog returns to me "empty handed" then I don't think of it as having lied to the dog, because I never send the dog after something I don't believe to be there. From a trialing standpoint, it should mean that the dog will continue to run out and look for sheep until it finally sees them (vs. running out a particular distance and then turning in, something I see all too often from the set out). From a work standpoint, the dog will not give up immediately upon not seeing stock, but will keep looking until it comes to believe they won't be found there. It may seem like splitting hairs, I suppose, but to me it means the difference between a dog being persistent enough to look thoroughly and one who gives up much more quickly.

 

Another side to this is that if I send my dog out to where I know stock to be but can't see them and the dog *doesn't* come back, I can assume that there is a problem and that I need to walk out there and see what's wrong. Twist was always a good dog for this. If, for example, she found the sheep and one was stuck, she wouldn't return; she'd wait for me to show up and do something about it.

 

ETA: I also don't fake throw balls/toys, but I do sometimes use treats for training, especially if training "stupid pet tricks." For example, I think it would have taken much longer to teach Birdie and Pip to sit up if I hadn't started out by luring them with treats. Like Maja, though, I don't generally use treats to teach a youngster to come to me, but then again, I usually call and the entire pack comes running, so a pup naturally learns to come too....

 

J.

Link to comment
Share on other sites

 

I am not against using treats, and I don't think it is lying, but I think it is a pity when people teach the dog to value a treat more than the praise from the hander. Particularly for border collies who were bred to value the handler above all. What is boils down to is that the pup is more likely to come when the reward is the treat than when the reward is the praise of the handler.

 

 

 

I do both. And, if done correctly, in the long run the dog is every bit as likely to do it when the reward is praise - or even when there is no external reward at all - as he or she is to do so for a treat.

 

I'm not saying this to oppose your preference in any way. Just to point out that treats and praise need not be mutually exclusive. This is something that lot of people believe about +R training and, in fact, it is a way that some (not all - some) +R trainers do train. But there is a big movement right now (and I have gone this way myself) to use praise and verbal encouragement much more liberally in +R training. Many of us have learned that praise need not be used as a "non-correction marker", and it has a place in a +R based framework, and we are running with it.

 

If I send one of my older dogs through a doorway and he or she orients on the other side, almost every time (now), I am going to acknowledge the behavior with sincere praise. Sometimes I will still treat, and sometimes I don't acknowledge it at all, but I usually give at least a sincere "good!" and it is sufficient. They know what is expected and they don't need, nor even expect, a food reward every single time. Right now, Bandit is getting a treat (and praise) every single time. Eventually he will be like the other two, and he won't need that. He's not learning to value the treat more than praise. He is learning what I expect. When he knows that, I will go more to praise without the treat.

 

There are some things that I have started teaching him without food. I finally got to have the experience (after four dogs who absolutely needed food at first) of pointing to a short tunnel to indicate that he should go through and he did. I didn't incorporate food just to incorporate it - if he doesn't need it, I don't use it. But, I'll tell you this - when we eventually start working on entry discrimination, I will absolutely use an external reinforcer to jackpot correct entries. Probably toy play in that case, but maybe food if it is warranted. But in competition, he will get praise for correct entries, or nothing because he will probably be running at full speed.

 

When used correctly, the end result of training with treats should be a dog who can do whatever he or she is trained to do just as fluently when they are not present. If Dean is running across the yard and I call him and he turns on a dime and comes running back, you wouldn't know it's a "treat trained" recall. All you will see is me call, him come, me praise, and us walk off together, no food in the picture at all. The treats are a tool. At some points in training, their value is utilized. In the end, the dog should have knowledge and treat use is not required.

Link to comment
Share on other sites

"Just to point out that treats and praise need not be mutually exclusive. "

 

I never said they were :) .

 

I know how to use treats for dog training with excellent results. I am a whizz with cliker. But I choose not to use treats or toys for training my sheep dogs who are going to work sheep.

Link to comment
Share on other sites

I know how to use treats for dog training with excellent results. I am a whizz with cliker. But I choose not to use treats or toys for training my sheep dogs who are going to work sheep.

 

It seems to me that, in training just about anything BUT stockwork, the handler needs to find a good reward for the dog. I'm not suggesting that the reward has to be food or play or praise... I think each handler has to find the right motivation for their dog.

 

When training on stock the work is the reward and for a BC it's probably the biggest reward on the planet. It just seems difficult to directly compare the approaches of training on stock verses training in agility, SAR, therapy, etc. Even though many of the techniques probably do overlap there's just an inherent difference between allowing the dog to do what is was born and bred to do verses teaching it something that, while still channelling the dog's natural instinct, is probably a bit more foreign to it.

Link to comment
Share on other sites

Camden's Mom,

Yes, I agree, you have find it. But the trainer also cultivates the desirability of certain types rewards. So to some extent the handler decides what it is that the dog loves the most. So I cultivate what I need for livestock work. But, if I had a dog that did not work livestock, I would still raise it without treats.

 

-----

 

One of the comments I had got on Bonnie when she was 5-6 months and had this outrageously excellent recall was that she did it because she and I had a special bond because of the livestock work. Bonnie and I do have special bond, but the problem with this statement was she had not been on sheep yet. The whole puppy-hood is without sheep.

Link to comment
Share on other sites

Camden's Mom,

Yes, I agree, you have find it. But the trainer also cultivates the desirability of certain types rewards. So to some extent the handler decides what it is that the dog loves the most. So I cultivate what I need for livestock work. But, if I had a dog that did not work livestock, I would still raise it without treats.

 

It seems that all of these threads devolve into a sheep > praise > toys > treats discussion. At some point, someone will state that they could train an agility dog, for example, using only praise, and then someone else will ask if that person would work for free. I believe, then, this is the time when I confess that I clicker-trained Rex in agility, and he IS a sheepdog who trials at the open level in USBCHA-style trials. Trust me, he is not looking for treats when we enter the field.

 

In Starting Dogs, the online course offered by Alasdair and Patricia MacRae, Patricia demonstrates how they teach a lie down off stock, and yes, she uses treats. Obviously, once their dogs are in front of sheep, they don't use treats to get a lie down, but using treats to get a behaviour off sheep hasn't hurt their desire to work sheep.

Link to comment
Share on other sites

"I am not against using treats, and I don't think it is lying, but I think it is a pity when people teach the dog to value a treat more than the praise from the hander. Particularly for border collies who were bred to value the handler above all. What is boils down to is that the pup is more likely to come when the reward is the treat than when the reward is the praise of the handler."

 

"I do both. And, if done correctly, in the long run the dog is every bit as likely to do it when the reward is praise - or even when there is no external reward at all - as he or she is to do so for a treat."

 

I'm sure this puts me in the "Jurassic School of Dog-Rearing," but I want my dog to do as she's told because it's her job to do so. Not because there's a liver treat or a string of squeaky words in it for her. My dog gets praise, hugging, cookies and silliness in good measure. But not to get her to do what she's told. She gets it because she honors the covenant between dog and owner, which is, I look out for you and give you everything you need, and you give me what I ask of you. In return for her honoring the covenant, I do not ask of her that which is unreasonable, and she understands that. Which is why she will do what I ask, even if it seems strange, irrational, or even a little scary.

 

This has worked for me all my life, with many dogs. My dogs are not led to believe that life is always fun. But that they can trust me that while what I ask is not beyond their capacity to endure. A bath, an exam by a vet, and giving up a chase or relinquishing a delectable bit of carrion are things that must be endured. So are allowing me to extract a foxtail from a nose, or a tick from an ear. Both are endured - not with cringing obedience - but with stoic calm. Because I have the dog's trust, and because we have a deal.

 

Hence, lying is generally not needed, nor is force. Treats are dispensed as just that - treats. It's a hot day and wouldn't an ice cream be a treat?

 

This is my way of doing it. There are others. Some work as well for their adherents as mine do for me and my dog. I feel that my way is the best way, the same as others do. As long as you and your dog are happy, go to it! :)

Link to comment
Share on other sites

"Just to point out that treats and praise need not be mutually exclusive. "

 

I never said they were :) .

 

I know how to use treats for dog training with excellent results. I am a whizz with cliker. But I choose not to use treats or toys for training my sheep dogs who are going to work sheep.

 

I get that. However, in response to your earlier comment ("I think it is a pity when people teach the dog to value a treat more than the praise from the hander."), I just wanted to mention that many +R trainers are not teaching their dogs to value treats over praise. More and more (myself included) are cultivating an appreciation for both in our dogs.

 

I think a lot of people don't realize that this is happening.

Link to comment
Share on other sites

 

I get that. However, in response to your earlier comment ("I think it is a pity when people teach the dog to value a treat more than the praise from the hander."), I just wanted to mention that many +R trainers are not teaching their dogs to value treats over praise. More and more (myself included) are cultivating an appreciation for both in our dogs.

 

I think a lot of people don't realize that this is happening.

 

 

Im curious as to whether you begin with food reward and match the food with the praise and then occasionally replace the food with praise - until the dog gets it that the praise = the food....then back off on the food almost altogether to end up with a dog who's as reliable with praise as with food...(that's a question lol)

Because it strikes me that you'd have to make a tight association at the beginning that you can move forward from, no? My question is because food is a primary reinforcer and praise is somewhere south of it. That's really why food is the +r trainer's method of choice. Technically we could use the other primaries sex or fear as motivators instead but we don't for obvious reasons. I know that there are people starting with appreciation but it's pretty far from the premise.

Link to comment
Share on other sites

Im curious as to whether you begin with food reward and match the food with the praise and then occasionally replace the food with praise - until the dog gets it that the praise = the food....then back off on the food almost altogether to end up with a dog who's as reliable with praise as with food...(that's a question lol)

 

It depends a lot on the dog and the dog's natural response to praise. Some value it from the start. In that case, there might be some behaviors that the dog picks up on easily where praise can be the main reinforcer from the start, and I might never formally "reinforce praise", but just use it. It can also depend on the complexity of the behavior. Orienting at the door is one that I drop food for very quickly. It's simple, dogs tend to get it easily, and it can often be reinforced with praise alone very early on. In the case of a very complex behavior, I might reinforce with food and pair it very specifically with praise for a good while before switching to praise alone.

 

But I have found that the incorporation of praise can be a very natural type of thing. I'm not usually thinking, "I am going to praise and not feed here". But I am making more of an effort to use verbal praise, watching how my dog responds, and going with that flow when I see reinforcement happening. If that makes sense . . . ?

 

Some dogs don't care at all about praise, or are stressed by it (Dean was stressed by it initially). In that case, a strong pairing is important so the dog knows it means something good. I've never loaded praise like a clicker, but pairing hearty praise with a jackpot was a mainstay with him for quite some time. He understands it now and I can use it with him, but not to the same extent that I can with Tessa, who naturally appreciates it. But he's coming along.

 

 

Because it strikes me that you'd have to make a tight association at the beginning that you can move forward from, no? My question is because food is a primary reinforcer and praise is somewhere south of it. That's really why food is the +r trainer's method of choice. Technically we could use the other primaries sex or fear as motivators instead but we don't for obvious reasons. I know that there are people starting with appreciation but it's pretty far from the premise.

 

I would say that +R trainers in general are really starting to branch out from use of food almost exclusively as a reinforcer. Yes, food is still definitely the "go-to" reinforcer. Most dogs desire it, and it is convenient to have on hand and it can be used almost anywhere. But more and more of us are incorporating toy play, environmental reinforcers, personal play (with handler without toys), and praise into our reinforcement repertoire. I would say at this point I am seeing it much more at the fluency building stage of training, but it is starting to come into play more and more at the initial training stage of training, as well. A lot of it would depend on whether or not the dog in question finds praise inherently reinforcing. For those that do, there are definitely options for using it when introducing new behaviors. For those that don't, the handler would opt for something that the dog does find reinforcing at that stage.

 

It is very interesting watching this continue to develop. +R training is in a stage of some really dynamic growth and development right now.

Link to comment
Share on other sites

Hum, I share my lunch with my dogs when working.

 

I remember as a child, no one had to give me a treat to do something I loved.

I think my dogs feel that way.

Link to comment
Share on other sites

My dogs get food for 'tricks' to amuse me, mostly, while feeding them dinner. But I've noticed that their response is as good mostly on walks, where there is nearly never any food.

 

The only exceptions to this are for the recall (because one of them can be bad at it and I'm working on making it more fun for her to come back) and teaching them to relax around something they hate (vet, cars, that sort of thing.)

Link to comment
Share on other sites

"I confess that I clicker-trained Rex in agility, and he IS a sheepdog who trials at the open level in USBCHA-style trials. Trust me, he is not looking for treats when we enter the field."

 

I don't think how anything I wrote would make anybody think I expect a dog to look for treats on the trial field. I think my point was missed way off.

 

Kristine,

Yes, this was my very point in my first post that for the Border Collie the human praise is inherently motivating and it needs to be cultivated. Of course not for each and every border collie. But they have been bred on purpose to be like that.

 

"However, in response to your earlier comment ("I think it is a pity when people teach the dog to value a treat more than the praise from the hander."), I just wanted to mention that many +R trainers are not teaching their dogs to value treats over praise. More and more (myself included) are cultivating an appreciation for both in our dogs."

 

Yes, they do of course. However, if you use food as the primary motivator you are only elevating human praise to the level of food, which then by definition is more motivating. So the final result is more or less the same, but the way one gets there is different. If you first cultivate human praise then if you use treats later you promote the treats/toys, while the praise is the gold standard. Otherwise the treats/toys are the vehicle to elevate human praise.

 

I did say it's a pity many people seem to throw away this special feature of the breed (they are not the only one that have it of course), because it is such a great thing for me. But it was in no way to indicate ineffectiveness of other methods, or to imply that the dogs do not value human praise. I was trying to spell out (in my opinion) a profound difference in how one establishes their relationship with the dog.

 

But in the end they all love us :D and are trained.

 

Tea and Geoni Banner - I agree.

 

Maja

Link to comment
Share on other sites

 

It depends a lot on the dog and the dog's natural response to praise. Some value it from the start. In that case, there might be some behaviors that the dog picks up on easily where praise can be the main reinforcer from the start, and I might never formally "reinforce praise", but just use it. It can also depend on the complexity of the behavior. Orienting at the door is one that I drop food for very quickly. It's simple, dogs tend to get it easily, and it can often be reinforced with praise alone very early on. In the case of a very complex behavior, I might reinforce with food and pair it very specifically with praise for a good while before switching to praise alone.

 

But I have found that the incorporation of praise can be a very natural type of thing. I'm not usually thinking, "I am going to praise and not feed here". But I am making more of an effort to use verbal praise, watching how my dog responds, and going with that flow when I see reinforcement happening. If that makes sense . . . ?

 

Some dogs don't care at all about praise, or are stressed by it (Dean was stressed by it initially). In that case, a strong pairing is important so the dog knows it means something good. I've never loaded praise like a clicker, but pairing hearty praise with a jackpot was a mainstay with him for quite some time. He understands it now and I can use it with him, but not to the same extent that I can with Tessa, who naturally appreciates it. But he's coming along.

 

 

 

I would say that +R trainers in general are really starting to branch out from use of food almost exclusively as a reinforcer. Yes, food is still definitely the "go-to" reinforcer. Most dogs desire it, and it is convenient to have on hand and it can be used almost anywhere. But more and more of us are incorporating toy play, environmental reinforcers, personal play (with handler without toys), and praise into our reinforcement repertoire. I would say at this point I am seeing it much more at the fluency building stage of training, but it is starting to come into play more and more at the initial training stage of training, as well. A lot of it would depend on whether or not the dog in question finds praise inherently reinforcing. For those that do, there are definitely options for using it when introducing new behaviors. For those that don't, the handler would opt for something that the dog does find reinforcing at that stage.

 

It is very interesting watching this continue to develop. +R training is in a stage of some really dynamic growth and development right now.

 

 

Yes, ok. As a trainer I agree with you that

the pairing of a primary and praise is a

possibility but individual dog dependent for sure.

Link to comment
Share on other sites

Yay! My learning theory is based on science after all! Turns out we backward, outdated folks are the cutting-edge scientific ones! :D

 

Seeing as dogs often have a great internal motivation to eat I'd argue that treats would only enhance the training experience when used correctly.

Read it again, Mara, I think you missed the point (which IS contrary to what behavioral theory would lead one to expect). The motivation to eat is not internal (inherently related) to the training experience/goal, it is instrumental. It's analogous to the pizza parties for high academic performance -- the fact that teenagers may have a great internal motivation to eat pizza doesn't make that an internal motivation to work at learning.
Link to comment
Share on other sites

I'm not really sure the concepts of internal vs. instrumental motivations as outlined here can be applied to dogs.

 

Instrumental motivations as described here seem to be pretty abstract concepts. Even the promise of a pizza party mentioned near the end hardly equates to the concept of an immediate reward when a desired behavior is offered.

 

It's an interesting article when it comes to human motivations and success. But I'm not ready to dismiss the work of behavioral scientists and their understanding of operant conditioning in animals based on this study.

Link to comment
Share on other sites

I'm not really sure the concepts of internal vs. instrumental motivations as outlined here can be applied to dogs.

 

Instrumental motivations as described here seem to be pretty abstract concepts. Even the promise of a pizza party mentioned near the end hardly equates to the concept of an immediate reward when a desired behavior is offered.

 

It's an interesting article when it comes to human motivations and success. But I'm not ready to dismiss the work of behavioral scientists and their understanding of operant conditioning in animals based on this study.

Are you saying that the distinction between internal motivations and instrumental motivations can only be made with regard to subjects who conceptualize abstractly? Or are you saying that a reward given immediately -- even if it's totally extraneous -- thereby becomes an internal motivator rather than an instrumental motivator? Or something else?

 

Have there been any behavioral studies in animals comparing the effectiveness of instrumental motivation with internal motivation, or studying the effect of instrumental motivation on internal motivation? Seems to me you could recast currently accepted learning theory to explain this outcome, but it would lead to a re-thinking of whether utilizing both internal and instrumental motivators "may have the unintended effect of weakening the internal motives." Why is that such an unthinkable thought -- one to be rejected out of hand? Why should we not be open to this possibility?

Link to comment
Share on other sites

My apologies for a short off the wall remark after a glance through the article. Yes, I do believe that dogs also like to "get it right", please their person and have a sense of accomplishment at a job well done - i.e. internal motivation (if I'm understanding correctly). But we train dogs for a variety of different jobs which are at times foreign to them. I propose that a properly timed reward system communicates to them *when* they have it right and that this clear communication increases the internal drive. Personally, I feel that that is the beauty of (shudders) clicker training (which *shudders* incorporates treats). It's a clear communication system that says "you're right" at the instant the dog is right. Is it the only effective way to train? no. Does it cause the internal motivation to wither? Again, I would say no.

Link to comment
Share on other sites

I don't think the internal motivation of a dog is to eat. As Maja noted, the internal motivation of the dog is to please the human. They've been bred to work *with* a shepherd to reach a common goal. It's really not so far fetched to believe that an internal motivator is sufficient. And that doesn't preclude using instrumental motivators when the human deems them to be helpful, but I think the internal motivator should be what we work to encourage/enhance.

 

J.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...