the role of punishment

And secondly, the reason I'm being "sneery" about it   is that I have been lambasted many times by mudpuppy for using punishment, so I think it's important to know what is meant by the word "punishment", not only in this thread, but in all the threads where I have been so highly criticized for using it. Trust me, if I started a thread about positive reinforcement or operant conditioning and changed it from the scientific definition, I would hear about it. Probably from you. And I would expect to.

I didn't "re-define" punishment, I just wanted to discuss only PUNISHMENT DELIVERED DIRECTLY BY THE HANDLER TO THE DOG.  A subset of the universe of possible punishments. And wanted some practical examples where people felt hander-delivered punishment was the most effective way to produce a nice dog. So far the only ones we've come up with are "stopping a litter of puppies from chewing you up", which is iffy because no one is sure if the verbal is viewed as punishment by puppies, or if it's the re-direction and withdrawl of attention that is actually doing the deed (we need people to talk about DEAF PUPPIES please, anyone?); and careful skilled applications of punishment in the final steps of proofing a behavior for some dogs and some situations (kind of vague, but I concede it might happen).

I do not want my dog to chew my shoes. I do not want to worry about accidentally leaving shoes on the floor and having them eaten.By delivering (OC) punishment when one of my dogs attempts to mouth my shoes, I create a dog who ... does not chew my chews!

Having a million other things to chew will never teach my dog that shoes are on the no-no list. Without punishment, there is no no-no list. Since I have no desire to dictate my dog's actions every waking moment, and I have no desire to have a dog who think anything goes, I find the punishment, as described by OC and not mudpuppy, to be a highly effective part of having a well behaved and pleasant dog.

mudpuppy

I didn't "re-define" punishment

I didn't say you did. I said:

FourIsCompany

for the purposes of this thread, the word had been defined slightly differently.

mudpuppy

And wanted some practical examples where people felt hander-delivered punishment was the most effective way to produce a nice dog. So far the only ones we've come up with are "stopping a litter of puppies from chewing you up"

Trust me, that's not the only ones we could come up with. I could come up with plenty of scenarios where I choose to use punishment, I just don't see the point. You're just going to tell me how I should do it instead... Your way. You'll say I'm unskilled and lying and tell me just how "wrong" I am and how I'm ruining my dogs and shutting them down.

I ask you... why would I open myself up for that? Why would anyone?

Dog_ma

I do not want my dog to chew my shoes. I do not want to worry about accidentally leaving shoes on the floor and having them eaten.By delivering (OC) punishment when one of my dogs attempts to mouth my shoes, I create a dog who ... does not chew my chews!

I'm not criticizing (at all, honestly!) the way you teach it, but just to present another perspective on how I go about teaching something like this. Just for general info to learn about other perspectives, if you will.

When working with puppies (or older dogs really, I suppose age doesn't really matter), using the shoe example, if I saw a puppy going towards shoes, I would simply redirect onto something else (call name - orient - redirect), and reinforce that change in behaviour. Rather than punish for the shoes, what I do is basically say "No no, rather than chew on that, let's chew on this!", and then reinforce for the "this!". I also make it a point to reinforce pups/dogs for choosing their own toys over another. A lot of people don't think of this point, and often take "chewing own things" for granted. But it really does work. If I see a pup chewing on its own toy that it picked out, I will walk over and give praise, cuddles, rubs, or a treat, and walk away again, just to let them know that what they are doing is really really great! Using this type of procedure, the dogs also learn "not to" chew on things that are't theirs, by virtue of being told "this is better!" and remembering to tell them when they have chosen the correct chew toy.

And yes, I do have books low on bookshelves, remotes on the living room table, photo albums on a bottom shelf, mouse paraphenalia (bedding, as an example) on the shelf, nail polish on a low shelf, and often glassware. I don't live in a rubber room with my dogs *G*. I do like ornaments and other living-y things. And they do indeed learn to leave things alone that aren't theirs.

As they grow up I do teach a "Mine" cue (a "Leave it" cue), but I rarely have to use it for things in the home, it's more used for things outside (such as the yummy dead fox), although I have used it indoors too. Again I choose when to use it, and try not to overuse it so that the dog doesn't learn just to "leave things alone", but so that the dog knows that it has viable alternatives sitting on the floor or in the basket. Our shoes? They happen to be in a boot warmer, for aesthetics sake. But when visitors come and drop their shoes, they are left unattended by the dogs. :-)

Occasionally I have used body blocks - not touching the dog myself, but just standing in the way and preventing access to an item, and I just wait patiently. Usually involves me staring at the wall and quietly reciting some definitions I need to know for the next quiz in Neurobiology. In other words I'm not interfering, I'm just there. *G*Eventually the dog gets bored and gives up. The dog learns two things: 1) I reward the voluntary movement AWAY from the object (just like as I would clicker-teach "Leave it";), as soon as the dogs make that choice, and 2) it's not worth the time and energy to keep trying when mom is just going to stand in the way like a human-shaped rock, so it extinguishes on its own.

So just another way to end up with the same result, and it really doesn't take much effort at all, it's pretty easy.

Kim wrote:

 Using this type of procedure, the dogs also learn "not to" chew on things that aren't theirs, by virtue of being told "this is better!" and remembering to tell them when they have chosen the correct chew toy.

I'm being nit-picky, but what dogs learn is that there are generally better things to chew. This doesn't prevent a dog from thinking 'What the heck! Today, for something different, I'm going to chew that shoe!" If no one has ever communicated "Don't chew shoes" then it remains a very viable option. Shoes are tasty. 

And they do indeed learn to leave things alone that aren't theirs.

But, but ... how do they know what isn't theirs, if never taught "no"? If you hadn't guessed, I take a very liberal and literal use of the word punishment.  In my mind, if you are teaching a dog "Don't do X" (As opposed to simply encouraging Y, which leaves X as a possibility) then you are using P+,  because you are decreasing the probability of X happening. If I had an abundance of raw meaty bones as Y, then shoes as X might never come up. But chewing shoes is rewarding, I don't have raw bones scattered about my house, and I do teach my dogs not to chew things that haven't been presented as theirs. I do this gently, with redirection following the "punishment." And Sasha is an absolute doll about running to her toy basket and getting her toys when she's excited. She knows which toys to grab, and she knows how to ask me for things I haven't given the green light to.

And I have to say - a body block is P+. You aren't a rock. You are actively communicating "stop that." In a very forceful way - more forceful because of your calm!  

 Okay, well I'm quite happy to live in a world where I concentrate on giving my animals something they can chew on rather than teaching them what they can't chew on. Hate to harp on about the rabbits, but really, puppy chewing is nothing next to rabbit chewing (in a rental house) and rabbits aren't as cooperative as puppies or older dogs. I found the best way to tackle this problem in the end was to combine a mild aversive with providing something acceptable to chew on. The mild aversive, such as waving my arms at the offending rabbit/hare, calling out "Oi! Get out of that!" and maybe getting up to shoo them away might have been a mild punishment, but I'm not really sure. As I said, I feel like it's a bit of a grey area because it worked both ways depending on the mood of the rabbit. The way I see it, what I did was teach them that if they want to chew in peace, there's a place they can do this and an object there that they can chew all night if they want to, and they do. So it's more like a "this object is yours" than a "every other object is not yours" thing. Sort of. Because I don't know that rabbits really think in those kinds of terms.

Anyway, I'm quite happy to have the possibility of the animal chewing something I don't want them to chew always being open. With rabbits it always is anyway, but knowing there's a fair possibility I could teach my dog never to chew something ever, I wouldn't bother anyway. Probably because I've got rabbits running around the place discovering new things to chew all the time and I've just become accustomed to dealing with it without punishment and find it pretty easy. That's just a personal lifestyle choice.  

Body blocking is another of those things that may or may not be punishment IMHO. Depends on the way you did it. If you block casually and don't make much eye contact, that's a whole different vibe to staring down at your dog as you jerk yourself to be in front of their nose wherever they turn. And body blocking does still offer a choice if you don't do it too intensely. For instance, you can body block Pyry and if he's not that fussed, he'll turn away and forget about it, but if he wants to go there, he'll keep right on dodging to get around you for an eternity, no matter how intensely you might do it. It's important to me that there's still a choice, even if it isn't a very good one. Penny can always choose to ignore me if I use a firm no-nonsene tone with her and there will generally be no consequences for her beyond me being peeved and muttering at her and refusing to be a part of whatever it is she's doing. The only exception being if safety is involved. I don't mess around with safety, obviously. I guess I let my dog get away with a lot of things if she wants it bad enough. Somewhere along the way I just discovered it was kinda funny when she cheerfully ignored my warning tone once she knew I'd never hurt her. I found it was good fun to see her cheeky personality coming out in her disobedience and found meaning in it as well. What can I say, she's an easy dog. I've come to be quite comfortable with accepting limitations in my animals. If I need to punish them to get them to stop doing something, I stop and think first what their behaviour is telling me and how I can address that in a positive way. It started with the animals I had that didn't respond to punishment, but it's seeped into my general philosophy and the only bad side-effect I've noticed is my dog knows I'm not good for any threats these days and isn't easily cowed anymore. I find I quite like her like that and I think it's repaired a lot of the mistrust between us.

Kim_MacMillan

As they grow up I do teach a "Mine" cue (a "Leave it" cue),

Isn't "Leave it" just about the same as using"Eh, Eh".......?

I don't use body blocks or the word NO. I don't believe that "punishing" a dog for chewing a shoe is the most effective way to teach dogs to not-chew-shoes: really, all you've done is teach the dog that ONE shoe in ONE place on the floor when you are present is bad. Dogs don't generalize well; in order to truly teach a dog that chewing shoes is bad, you've have to go throught a rather elaborate proofing exercise, planting shoes of all kinds everywhere and punishing the dog over and over again, and to ensure the behavior holds when you aren't home, you'd probably want to booby-trap some shoes. Weird how people understand they have to teach the dog to generalize and proof things like sit, but they think the dog can somehow learn in one trial about not-doing-things.

I found the most effective way of teaching dogs to not-chew-things is to instead teach them what-they-can-chew: they can chew anything on their chew-mat. They get rewarded for doing so. During the initial training, there is nothing else available, like shoes, to chew on; once the dog appears to grasp the idea, which usually takes less than a week with a young pup, you can start leaving stuff out, and continue to reward the pup for chewing on his mat, and if you see him showing interest in shoe you can gently distract him and redirect to his mat before he starts chewing- he never makes the mistake, no opportunity or reason to punish. Black and white rules, rapid learning, no punishment, no need to be hyper-vigiliant for months, your stuff is untouched even when you're not home.

mudpuppy

in order to truly teach a dog that chewing shoes is bad, you've have to go throught a rather elaborate proofing exercise, planting shoes of all kinds everywhere and punishing the dog over and over again, and to ensure the behavior holds when you aren't home, you'd probably want to booby-trap some shoes.

I disagree. I have never had to do anything elaborate. No proofing, no planting or repeated punishment. No tricks, no games. It all revolves around the foundational relationship a person has with the dog.  

mudpuppy

I found the most effective way of teaching dogs to not-chew-things is to instead teach them what-they-can-chew: they can chew anything on their chew-mat.

What happens when they bring a shoe to their chew mat? I have found it more effective to teach my dogs that they can chew what I give them. Anything else is off limits. They know. That's why I can leave a roast on the counter or low table and they leave it alone because I didn't give it to them. I can set my plate on the coffee table and they won't touch it. Why? because I didn't give it to them.

I'm not saying that this is the ONLY way, and the chew mat is probably very effective. But I prefer my dogs don't chew anything I haven't explicitly approved for chewing. 

mudpuppy

I don't believe that "punishing" a dog for chewing a shoe is the most effective way to teach dogs to not-chew-shoes: really, all you've done is teach the dog that ONE shoe in ONE place on the floor when you are present is bad. Dogs don't generalize well; in order to truly teach a dog that chewing shoes is bad, you've have to go throught a rather elaborate proofing exercise, planting shoes of all kinds everywhere and punishing the dog over and over again, and to ensure the behavior holds when you aren't home, you'd probably want to booby-trap some shoes.

Wow, what an interesting view on this......I used a verbal to get my dogs over the shoe chewing stage.......you are talking planting shoes all over and booby trapping shoes.......that sounds like overkill when one can just use a verbal .....all my dogs are loose in the house when I am gone....the youngest being 16 months old.....although, I have to admit Nanook was a little tougher than the others......but, he is 100% shoe proof now......

corvus

 Okay, well I'm quite happy to live in a world where I concentrate on giving my animals something they can chew on rather than teaching them what they can't chew on. Hate to harp on about the rabbits, but really, puppy chewing is nothing next to rabbit chewing (in a rental house) and rabbits aren't as cooperative as puppies or older dogs. I found the best way to tackle this problem in the end was to combine a mild aversive with providing something acceptable to chew on. The mild aversive, such as waving my arms at the offending rabbit/hare, calling out "Oi! Get out of that!" and maybe getting up to shoo them away might have been a mild punishment, but I'm not really sure. As I said, I feel like it's a bit of a grey area because it worked both ways depending on the mood of the rabbit. The way I see it, what I did was teach them that if they want to chew in peace, there's a place they can do this and an object there that they can chew all night if they want to, and they do. So it's more like a "this object is yours" than a "every other object is not yours" thing. Sort of. Because I don't know that rabbits really think in those kinds of terms

I'm kind of struggling here, to understand why the emotional baggage around the word punishment is so strong despite the clear distinction between conventional everyday punishment and OC punishment. A mild aversive that produces less of a specific behavior - IS PUNISHMENT! I'm not creating the terms - dudes in white lab coats did. Punishment can be very mild. In my opinion, it should always be the mildest necessary.

My focus with my dogs as puppies is absolutely on what they CAN chew. My punishment for chewing wrong items is very mild, and not meant to stop all errant chewing immediately. I am going for mild punishment and a million repetitions. I don't expect a puppy to generalize, nor do I freak out if a grown dog gets a freak chewing storm. But not unlike mudpuppy's mat, my dogs learn that what they do chew are things that have been presented as dog toys. There is always a presentation of a new toy, and that signals "Go to town!" My dogs always have plenty of their own toys available. An embarrassing amount of toys, with different textures for different moods.

I'm also wryly amused that my insistence on using the technical definition of punishment has got people thinking I'm overbearing and conventionally punishing. I start from a philosophical orientation of positive, and add in OC punishers bit by bit, as the need arises. I usually err on the side of too little punishment. Yes, I can see mudpuppy rolling her eyes but perfect compliance is not something I need or want from dogs. It is the same principle I use with my daughter: if it is not dangerous or destructive, I'm willing to negotiate and encourage. Which often means not getting MY way, or not getting it quickly. I see puppies and children as long term projects. (No, my dogs and child aren't things). I'd rather act towards the big picture than get hung up on perfection or instant gratification.

I think I will try to bow out of this conversation - no promises! - because it is clear we won't all come to a meeting of the mind on the term punishment. No point in talking past each other. Wish me luck!  

Dog_ma

I'm also wryly amused that my insistence on using the technical definition of punishment has got people thinking I'm overbearing and conventionally punishing.

I am convinced that that is the great assumptive misunderstanding on this board! It's terribly sad that we're all DOING about the same thing, give or take a few percentage points, it's just that some are calling it something different and assuming that others are blasting their dogs with over-corrective and conventional punishment.

Dog_ma

No point in talking past each other. Wish me luck!  

Well-said! Good luck! LOL  

snownose

Isn't "Leave it" just about the same as using"Eh, Eh".......?

You might think so, and it might even result in the same behaviour - leaving something alone - but the motivation behind what I do and what you do - the reason the dog is performing the behaviour, is likely quite different. Or maybe not. It depends on how you teach "Eh eh".

But the way I teach it is that the dog thinks that "leaving things alone" is a good thing - that better things will happen if they do, rather than worrying about punishment if they don't. It's a subtle difference, but a clear one.

Dog_ma

If no one has ever communicated "Don't chew shoes" then it remains a very viable option. Shoes are tasty. 

But it doesn't, honestly. In terms of reinforced responses - if something, such as a dog's toy, is always reinforced, and attempts to chew at shoes are always redirected and never reinforced, then a dog will learn to leave shoes well enough alone. If you haven't tried it, it might not be clear, but for those who use it it's very clear. You don't have to punish to teach dogs to leave things alone. Dogs learn habits well, and once a habit of "these are things that I have been told are great to play with", they will play with those toys. I guess you could say the communiction of "Don't chew shoes" is communicated through saying "Here, chew this instead" by redirecting and reinforcing for that.

Heck, I have even taught my guys to handle various household items with their mouths - remotes, brushes, combs, make-up bottles, etc, in teaching "retrieving to hand" exercises, and even then they leave them alone unless they were asked for. :-) You would think that by reinforcing those things would make them think it was okay to play with, but it doesn't, since it's the specificity of context.

I'm sure you did not have to punish your dog for every item in the house "except" its toys, rather you likely punished for a few, but also reinforced a lot more for its appropriate items. That would make for an unfortunate puppyhood, as a house holds a lot of items *G*.

Dog_ma

But, but ... how do they know what isn't theirs, if never taught "no"?

Well, I don't know what dogs think of as "theirs" vs "mine". I don't think dogs really think much like that, possession to them is quite temporary, only as long as they have it in their possession is it "theirs". But what I do know is that it works well, and the dogs tend to leave things be that are not theirs, through consistent redirection to appropriate chewing items. The key though is consistency and good timing, if the dog has already been chewing on a pencil for five minutes, it's not as effective as if you saw the pup doing it right as it went to grab the pencil, as the dog has then become reinforced for the pleasure in chewing the pencil.

Dog_ma

If you hadn't guessed, I take a very liberal and literal use of the word punishment. 

I had guessed, yes, based upon your next statement:

Dog_ma

In my mind, if you are teaching a dog "Don't do X" (As opposed to simply encouraging Y, which leaves X as a possibility) then you are using P+,  because you are decreasing the probability of X happening.

That is not how P+ is defined though. The behaviour of "chewing on the pencil" hasn't been punished, rather it goes extinct through lack of reinforcement. If the dog is never reinforced for touching it, eventually it won't be worth touching, when there are lots of "fun" touchables on the floor. I do "prevent" the dogs from doing "X", but not by saying "Don't do X", but rather by saying "Y is much more beneficial". You added in that it still leaves X as a possibility, but in my experience it never ends up being the case that way. If the dog isn't reinforced for doing X, it won't keep trying to do X, as the cost/benefit analysis doesn't pay off when the perfectly acceptable Y is lying right beside it. Why keep trying to do something that never really pays off?

I can teach a formal heel by giving rewards simply when the dog is where I want it. By walking around the yard aimlessly and giving a reward when the dog gets where I want, I can teach a strong heel. That doesn't mean that I have punished "being away from me". It just means that I have made "being next to me" very rewarding.

A dog that tries to paw for a leaf on the other side of the fence, but cannot reach it, will eventually stop trying to reach the leaf. The dog wasn't "punished" in any way, the attempts at doing it simply didn't pay off with reinforcement (the leaf), so the dog gave up trying.

Do you see the distinction yet? I have tried to supply a few other examples of the same meaning to see if it helps.

Dog_ma

But chewing shoes is rewarding,

It's only rewarding if the dog knows that chewing shoes are rewarding. For chewing shoes to be rewarding, the dog would have to have had experience chewing shoes. My goal is not to give the dog that rewarding experience in the first place (management at first, then redirection), and to make something else much more rewarding.

Dog_ma

And I have to say - a body block is P+. You aren't a rock. You are actively communicating "stop that." In a very forceful way - more forceful because of your calm!  

No it's not. I'm not applying anything. I'm just there. I'm not communicating "stop that". I'm communicating "Try all you want buddy, but I ain't moving". They are two very different thing! If I was communicating "Stop that", I would be doing something that stopped it immediately. If I were to say "stop that" I wouldn't just stand there. But I'm not stopping it immediately, I'm just making it non-rewarding for the dog, and not worth the dog's time. When I say I do a body block, I'm just there. I'm not pushing into the dog (but I'm not moving backwards either, just holding ground), I'm not even interacting with the dog.  In order for it to be a positive punishment I would have to "apply" something (such as "no", pushing into the dog, staring at the dog, or popping a leash on the dog). If one were really stretching to try to prove that someone used P+ you could say that my "body" was the applied stimulus, but the fact is the dog stops trying because reward can't be obtained, not because I've actively done anything aversive to make the dog think it should avoid the object. It would be no different if every time the dog approached the object and a magical wall appeared as the dog approached it, where the dog could see it but could not reach it. Same as the fence analogy above. The fence certainly isn't punishing the dog, it's simply preventing access. They are two very different things. It stops behaviour, but through extinction, not punishment.