Teaching Commands

[stumped]

Point is we dont have to provide the stimulation to take advantage of it, if the dog feels like sitting , tired legs or not, and you say sit an association is formed, no external reward is needed.

The OP said "without the person providing reward or punishment.."

[sidoney]

It wasn't an action that the dog was carrying out by herself. It was an action where the dog was put into the crate, this was repeated, and the dog started to go into the crate by herself.

My suggestion was that there was a motivation provided for the action.

[Erny]

Point is we dont have to provide the stimulation to take advantage of it, if the dog feels like sitting , tired legs or not, and you say sit an association is formed, no external reward is needed.

The OP said "without the person providing reward or punishment.."

Sorry .... perhaps it is too late at night for me. I do understand what you say there - it's about "association" .... but it's not really "training" per se, is it. IE Doesn't go to follow that, when we want the action we've paired words to, the dog will then go ahead and do it. Unless there's something in it for him, one way or the other ?????

Maybe I've gone off the page ...

Edited October 22, 2007 by Erny

[Erny]

It wasn't an action that the dog was carrying out by herself. It was an action where the dog was put into the crate, this was repeated, and the dog started to go into the crate by herself.

My suggestion was that there was a motivation provided for the action.

Perhaps I'm on your page, Sidoney? I'm imagining the possible motivational shift from "good dog" to dog becoming accustomed and comfortable in the crate. Motivation shifts from *basic verbal reward* to *place of comfort reward*. From the outset, the comfort the crate provided (blankets?) possibly amplified the verbal reward "good dog" (or vice versa). So the reward the dog actually received from being in the crate was quite possibly bigger than what would appear at first glance.

[sidoney]

Erny I see what he's saying ... the dog gets some kind of reward from the behaviour that it does. Since the behaviour is done for some kind of motivation. The reward comes after the action, it just doesn't come from you. No difference to the dog really, it gets the reward. And the classically conditioned cue is associated with the behaviour.

For something that the dog is likely to do infrequently, or quite a bit while you are not there, it may be less efficient, but for something where you can control the situation more and reliably provide a cue with the action, it could be effective.

However that is not the situation described by the OP, with her first question. Since the behaviour is not a self directed one, but the dog is guided into or placed in the crate initially.

[sidoney]

In response to your second post Erny, yes I think we are talking about the same thing. There is a motivation for the action, but it's not obvious, and so it may seem that there is no reward or punishment. And it may not be necessary to work out what the motivation is - if the behaviour is happening and reliable. It is working for this dog and this handler and goodo for that. It may or may not work for a different dog or handler or situation or context, and if it's assumed that there is no reward or punishment, that leaves little room to adjust the approach.

[stumped]

The examples she gives, are not good , as Vickie says later. But the question is about conditioning a response with no outside reward or punishment, and it is usable, easy and has been done for years. No one is putting it up as a training method, it is just another tool that can be utilized , if something is occurring , with no input from yourself, why not add the normal cue as a reinforcement.

[Sarah L]

The examples she gives, are not good , as Vickie says later. But the question is about conditioning a response with no outside reward or punishment, and it is usable, easy and has been done for years. No one is putting it up as a training method, it is just another tool that can be utilized , if something is occurring , with no input from yourself, why not add the normal cue as a reinforcement.

tmc I think what you have said above explains what Vickie has been trying to say perfectly. Many years ago when I first started dog training I was told to help with the reinforcement then when the dog sits on its own for what ever reason you say sit, when the dog lies down on it own for whatever reason you say down ect ect.....

I do believe conditioning a response with no outside reward or punishment is very possible. It just becomes matter of fact for the dog to respond with no great emotion shown by the owner either way in reward or punishment.

[Erny]

What ATM describes is pretty much "shaping" ... in the manner he/she describes though, the reward is that the dog WANTED (for whatever reason) to do it anyway - so reward is self-served.

But this is different to what Vicki describes in her instance of her dog/crate story .... I don't think it is a bad example to describe, but simply that it is not an example of the above.

Edited October 23, 2007 by Erny

[sidoney]

IMO it would be good to clarify a few uses of language here.

"No outside reward or punishment" could be better stated as "no trainer-originated reward or punishment". The dog is getting something from the environment (I include biofeedback in this).

There is some mixing of the concepts of cues and reinforcement. A cue elicits a behaviour. A reinforcement increases the likelihood of the behaviour happening again. A cue can be a reinforcer in a behaviour chain, but in this case it reinforces the behaviour that precedes it, not follows it. By that I mean, as an example, if you are giving a dog its dinner, if you say "sit", and it knows that after the sit it you will tell it to (for example) "eat it", the words "eat it" are both a reinforcer for the sit, and a cue for eating the dinner. Further forward in the chain, if you call the dog to you, ask it to sit, and then tell it to eat its dinner, the word "sit" can be a reinforcer for the action of coming, and a cue to do the sit behaviour.

It's worth being clear about terminology when specialised terminology such as "cue" and "reinforcer" is used, as otherwise it can lead to confusion.

[stumped]

Yep pinnacle, it is easy, probably too easy,so it must be analyzed till it is difficult.

On a separate note, what is the dog in your Avatar?

regards Tony

[Sarah L]

Yep pinnacle, it is easy, probably too easy,so it must be analyzed till it is difficult.

On a separate note, what is the dog in your Avatar?

regards Tony

tmc the dog in my Avatar is my dog Douglas he is a bullmastif x great dane 9 years old now. Rescued at at 9months and has been with me ever since. He is the one dog in a million you get.

I try to keep things as simple as possable especially for the average dog owner who can be easly confused.

Edited October 23, 2007 by pinnacle dts

[sidoney]

Where does one draw the line between "easy" and "too easy"? There needs to be understanding of what is happening, even basic understanding. It's much easier to provide understanding when doing things hands-on and face to face. But in a text based medium such as this, loose use of language can lead to different understandings by writer and the various readers. This is why there is SO MUCH misunderstanding in text based communication.

ETA: There are people who can do things, wonderful and amazing things some of them, but who can not communicate what they are doing and how. Or not verbally. And some of them don't even understand it themselves. They just know if they do THIS, then THAT happens. This can lead to things like method training, without understanding what underlies the method, and inflexibility.

Edited October 23, 2007 by sidoney

[4leggedvariety]

Back tracking a bit, I also taught my dog to go into the crate just by putting him in and closing the door. I suppose is some sort of mild compulsion, i am making the dog go into the crate. There were no food rewards or even a good dog. Very quickly, he goes in by me pointing to the crate.

I crate him just before training sessions and also when I am training my other dogs. I am not sure if he finds the thing self-rewarding, I suppose it is pretty neutral. Is just what I want him to do & he does it. However, it could be a conditioned reinforcer though, 'cause everytime he comes out of the crate, he has his training session with positive reinforcements, so he learns that good things follow his short stay in there.

I agree with Erny, it will be pretty hard to get a dog to do something it does not like without reinforcements.

[sidoney]

That reminds me of a trainer's description of what happened when she trained her dogs to stay on platforms while a chosen dog was being trained. It involved the Premack principle (a more likely behaviour being a reinforcer for a less likely one), and transfer of value. The person who did it found that initially the high value was on coming off the platform to be trained. However coming off the platform to be trained reinforced the dog staying on the platform, and being on the platform increased in value, until being on the platform was of higher value to the dog, and the dog would work in order to run back to the platform. This did not stay stable - the reinforcer of going to the platform increased the value of being trained. She found that the high value flip flopped between training and being on the platform.

BTW having a specific terminology makes it easier to communicate precisely in whatever field that one desires. It assists in keeping people "on the same page", as Erny might say. It's not for everyone, granted, but the basic concepts are easy to achieve, and can reduce confusion, particularly (as I said) when communication is by text. In addition, an understanding of learning theory, as contrasted to method and specific instances, allows one to be more flexible, particularly since learning theory is applicable both cross-situation and cross-species.

I will also add that not understanding at least the basics can lead to errors and inefficient or ineffective training. Some people understand innately without being able to vocalise it. Others don't.

It's worth mentioning because all sorts of people read this forum and I would find it a pity if people went away from this thread thinking that you can train a dog without the dog having some kind of motivation to do things, because ideas and concepts were not described clearly or for some other reason.

More thoughts: I think we are all agreed that you cannot train a specific dog to do a specific behaviour or set of behaviours on a forum. Words cannot be specific enough to accurately communicate specific problems or situations with specific dogs or trainers - words cannot describe the full range of behaviour. This is especially important for problem behaviours.

A forum is a medium for a discussion of ideas and concepts. Ideas can broaden knowledge or give deeper knowledge. Here, the knowledge is of dog behaviour, motivation, training, etc.

The more clearly and specifically the ideas can be expressed and understood, the more useful they are. For people who want to discuss ideas, the clear expression of them enhances the discussion.

The ideas can then lead to understanding of concepts, the trying out of the concepts in the real world, and an expansion of the ability of the person who has participated in the discussion (whether actively or passively).

Edited October 23, 2007 by sidoney

[Vickie]

You know the reason I started this thread was that I think sometimes people make things too hard. While all this knowledge is accurate in a scientific sense is not always used correctly by the average person who will not fully understand it or observe correctly. I can think of a few people who study & teach complex theory whose dogs are generally no better behaved than mine & in some cases worse. For simple & reliable behaviours & for the average dog owner, I wonder if too much theory is a good thing, because it can cause us to over think & consequently overreact. Isn't it possible that consistency with no extremes of positive or negative can be effective as a training method?

I have to keep asking myself...are we, as average dog owners, better or worse for all this theory? We want to believe we are better but sometimes I wonder.

It'd be interesting to go to a sheep trial & ask some of them how they train a recall. My bet is that there will be some extremes at either end, but most of them will have a very simple method with uses neither extreme positive or negatives. And they will be able to tell you in a sentence or 2. How often do you see a sheepdog that won't come when it's called? I don't see many.

I understand that there is value in theory & in science & fully understand how someone who makes a living out of theory will undoubtedly have many issues with what I am saying. It's fine, I still reward my dogs a lot, I try not to be too aversive but mostly I strive for consistency.

[Vickie]

It's worth mentioning because all sorts of people read this forum and I would find it a pity if people went away from this thread thinking that you can train a dog without the dog having some kind of motivation to do things, because ideas and concepts were not described clearly or for some other reason.

I think that is inevitable, it happens every day in many, many threads. The idea of raising this was for discussion. Seems like opinions are divided. That's part of a discussion. I read things that I cringe at all the time b/c they may/will be misinterpreted. I trust that people will come along & provide a balanced view, just as you have done in this thread.

I would also hate people to think that you need to study for a year & read 10 books (I'm exaggerating) to teach an effective recall, but that is the impression that some people could get reading various threads. That's not going to help the average pet owner, they will either a. not do it or b. do it & not understand it. Neither of which will help the dog as it runs on to a busy road.

ETA: could you maybe start a new post for major edits. It's hard to respond to something that keeps changing.

Edited October 23, 2007 by Vickie

[sidoney]

"Learning theory" can be as hard or as easy as you want to make it. For most behaviours, it comes down to some very simple basics and for the average person, the basics is all they need. I could write the basics in about eight sentences.

People who seem to "know theory", especially if they make it seem hard, probably have a deficient understanding.

If they don't seem to be able to train their own dogs, that also suggests that their understanding is deficient.

Isn't it possible that consistency with no extremes of positive or negative can be effective as a training method?

Of course it is. I questioned the possibility of teaching a dog without any positive or negative (motivation) at all. And IMO consistency is more important than just about anything and this may be where the people you are talking about are having problems.

I agree with you that some posts on this forum make what should be simple into something that is very complex! It doesn't need to be! If someone has the basics, they can teach just about any everyday behaviour.

ETA: sorry about editing late! I went out to do some poop scooping and kept thinking about the topic!

Edited October 23, 2007 by sidoney

[sidoney]

The value in a "theory" is that it can be used to understand something that happens in the real world. It's rare that people put theory before doing something. They generally just do it. Animal training that explicitly uses learning theory is an exception.

I'll use an example from the horse world. You get a whole bunch of "horse whisperers" that perform something that can be seen as akin to magic. They have discovered "the language of equus" or something equally silly, and worked up a whole bunch of rituals and rigmaroles around it. They gather themselves disciples that believe the magic and follow the method. But if you observe what they do and interpret it using learning theory, you can break it down and understand what is done in a much more simple way. And then take what is useful without being blinded by the myths that surround it.

(note a new post for a major edit! )

[sidoney]

Err new post again. I still think that accuracy in written language is important. That way people can be talking with some confidence about the same thing.