Puppy Forum and Dog Forums banner
21 - 40 of 187 Posts
He was arguing that we weren't talking about purely positive. And I was saying that nobody ever was arguing for pure positive.
Purely positive doesn't even EXIST. Your dog knows when they did something wrong. You get frustrated, you get tired, you look disappointed, you have effectively corrected your dog on a certain level. You remove a reward, you look stern, you make them try again, you step on a paw when they walk under your feet, you yell in panic because they're about to eat something dangerous - there is no way to be PURELY positive. It can't exist in the real world. It just can't.
 
He was arguing that we weren't talking about purely positive. And I was saying that nobody ever was arguing for pure positive. That's all I meant.
I guess I just meant it wasn't truly an argument as much as it was one person trying to continue the discussion and figure some things out in their own head. I truly think it wasn't meant to "argue", but to understand better what positive reinforcement really entails.
 
I could agree with that. Any negative training that I use has always been for just basic obedience around the home and public. I could definitely see how if I was working in sport training I would want to look into different methods.
I guess I have just come to realize if positive training can get me 20 obstacle performances with different criteria, discriminations, direction changes in 45 seconds, then it can work to teach sit - and honestly? It does.

But yeah, it's been good for me.
 
I guess I just meant it wasn't truly an argument as much as it was one person trying to continue the discussion and figure some things out in their own head. I truly think it wasn't meant to "argue", but to understand better what positive reinforcement really entails.
Maybe argue was a bad word. But I wasn't sure what he was trying to discuss because nobody advocated purely positive anything like it. So I wasn't sure where he was going with it. The only one claiming pure positive training exists is the author of the article who doesn't really understand +R training. Haha ETA That's usually how those articles to. I don't know anybody really who would claim they are positive with their dogs 100% of the time and ignore all bad behavior. Yet I see lots of articles by people claiming that these trainers exist and are completely wrong.
 
The discussion had gotten started about another thread when someone brought up Cesar Milan's methods in the use of training a dog against mouthing, so I think the OP just wanted a little more clarification on why some people are so adamantly against Milan's methods.
 
The discussion had gotten started about another thread when someone brought up Cesar Milan's methods in the use of training a dog against mouthing, so I think the OP just wanted a little more clarification on why some people are so adamantly against Milan's methods.
Yes, I know. I saw the thread. I'm really not trying to be argumentative here? I was jus questioning that one statement in response to my own post, wondering what he was trying to say. I'm on my phone and responses are brief so maybe it read differently than intended.
 
Discussion starter · #27 ·
Nobody on this post has claimed to be purely positive in the sense that there are zero consequences ever, and I doubt you will find anybody who thinks it's possible to train a dog without any consequences for behavior. So I'm not sure what you're arguing? Anybody who think positive training mean ignoring bad behavior and hoping it goes away just doesn't understand positive training. Therefore the author of that article doesn't understand it and really can't make a good argument against it.
That's good to hear! :) I must have misunderstood or extrapolated way too much when someone in this thread indicated they don't ever use the word "no." I suppose that's everyone's prerogative, excluding a perfectly good word of the English language from their lexicon. But I must scratch my head when I read stuff like that. ;)
 
That's good to hear! :) I must have misunderstood or extrapolated way too much when someone in this thread indicated they don't ever use the word "no." I suppose that's everyone's prerogative, leaving a perfectly good word of the English language out of their lexicon. But I must scratch my head when I read stuff like that. ;)
NO doesn't tell your dog anything. No doesn't mean anything to your dog. It doesn't tell your dog what to DO. It might, if you work hard, eventually mean either:
Stop
Or "You're not getting a cookie for that thing you just did"

But it's not going to change behavior. It's not going to make them do something else. It's not going to make them less likely to do it next week. It is, at best, a way to INTERRUPT incorrect behavior.

"Sit" tells your dog what to do. If your dog is jumping up and you say no, you have told them nothing. Tell them to sit and you have told them what to do when greeting people.

That's really all.
 
Yes, I know. I saw the thread. I'm really not trying to be argumentative here? I was jus questioning that one statement in response to my own post, wondering what he was trying to say. I'm on my phone and responses are brief so maybe it read differently than intended.
No, I'm not trying to be argumentative either. Just trying to explain what might have been the other point of view. I will let it go though, as it apparently seems as though I am also trying to argue when I'm definitely not.

In fact, I've really been trying to watch my text tone because I feel like there are a lot of sensitive people that are easily offended sometimes on forums. Talking through text gets misconstrued so easily.
 
Discussion starter · #30 ·
The discussion had gotten started about another thread when someone brought up Cesar Milan's methods in the use of training a dog against mouthing, so I think the OP just wanted a little more clarification on why some people are so adamantly against Milan's methods.
Not so much Milan, as the overall approach that argues for human leadership and assertiveness, and yes, negative correction. Let's not get bogged down with Milan (notice I left him out of the initial question in this thread!), as personalities often inflame arguments beyond facts, data, and logic.
 
No, I'm not trying to be argumentative either. Just trying to explain what might have been the other point of view. I will let it go though, as it apparently seems as though I am also trying to argue when I'm definitely not.
Haha. No worries. It's hard to read people over the Internet especially when people are posting from phones.
 
Also "no, or nope, or whatever to mean 'no cookie for you' is probably not bad for 95% of dogs, but given that I own at least one that finds it pretty demotivating and I, again, need happy fast and confidence? Dropping it was a wise choice for me and my dogs.

And it's not like it changes HOW I TRAIN or what the consequences are (ie: Not getting the treat). It just keeps that dog happier.
 
(really good idea to start this thread, IMO)
I think it's a huge fallacy to think you can fake "energy", or teach it. I totally agree that dogs feed off of handler's energy, especially those bred to work closely with people. Adopting a certain posture isn't going to change the chemical signals your body is putting out. If you're stressed, you're still going to smell stressed.

Dogs respond to "energy" in the sense that they respond to body posture and the chemical signals our body put out when we experience different emotions, but these aren't as easily manipulated as dominance theory trainers seem to think they are. You can't fake being calm- your body is going to betray that you aren't. You can try to make yourself calm, yes, but that isn't always going to work. Nor is it entirely necessary to have a well behaved dog.

I totally agree with the things said about both R+ training and dominance theory by others. R+ =/= never correcting a dog and letting the dog do whatever it wants and not being the leader in the relationship. It means not using physical correction. Very few R+ people don't believe in using verbal interrupters/verbal correction, and those that do are usually exception at controlling reinforcement (including functional reinforcement in the environment). I don't think it is good practice to advise people to never use any form of correction, ever. I don't think it is necessary harmful to avoid using physical correction, nor do I think it is necessary harmful to use physical correction on most dogs. For some dogs, yes, but not for all.

At the end of the day, dominance theory is based on several false assumptions:
1) That wolves relate to one another through a linear series of gradually increasing hierarchical relationships (alpha, beta, omega, etc) that are primarily rooted in the assertive displays/behaviors of the higher ranking animals, and include a lot of physical correction/manipulation
2) That dogs can be best understood through the behavior of wolves, since dogs were domesticated from wolves
3) That there are strong inter-species applications of this kind of behavior relationship- that humans can take the things a wolf does to enforce its social status and easily apply them to dogs, and have the dogs understand
4) That dogs relate to us similarly to other dogs, and will understand us relating to them as if we were another dog

The falsehoods of these ideas:
1) CAPTIVE wolves relate to each other in this way. Captive wolf packs are not healthy wolf packs. Wild packs do not work this way. In wild packs, there may be a linear hierarchy, but it is enforced by voluntary appeasement displays of the lower ranking animals, NOT aggressive displays by the higher ranking animals that force appeasement behaviors by the lower ranking ones (such as forced rolls, for example). Given the likely process domestication seems to take, I don't think it is a good idea to assume that captive wolf packs function more similarly to dogs than do wild packs. (see below explanation of how it is thought dogs were likely domesticated for why. Spoiler alert: it seems likely that the early stages of domestication were put into motion through voluntary contact of early wolves with humans, not intentional efforts to take them into captivity and tame them).
2) Dogs are not wolves. Current estimates place the beginning of the domestication of the dog around 14-20,000 years ago. Certainly by 14,000 years ago there were dog-like animals with skeletal markers that distinguished them phenotypically from the wolves of that time. Pretty much, this means their skulls and skeletons displayed the things we associate with dogs, and not with wolves. It is possible domestication started very recently after that (a fox farm study done in the 1950's by Belyaev introduced evidence for the assertion that the physiological and phenotypic changes associated with domestication (such as neotony, curled tails, piebald coats, and "friendly" temperaments) can occur very quickly, within only a few generations, by selecting for a lower flight distance, or the distance at which a person can be to the animal before they start showing signs of aggression). However, that would require a pointed effort of domestication, which considering dogs were the first domesticated animal it seems unlikely early humans actually set out to domesticate them- it wasn't even an idea that was likely to exist. The leading theory right now is that grey wolves at the time probably followed human camps, searching for refuse, which would have indirectly selected for animals with a lower flight distance. Over time, these animals grew more and more unafraid of people and likely were taken into human homes. Eventually, they were a distinct population from other wolf packs and changed enough from other wolves that there wasn't a whole lot of cross breeding among the populations.
- many thousands of years of divergence exist between dogs and wolves. In the early ages of domestication, it seems likely there wasn't any pointed effort, it was just animals finding a new ecological niche they could fit into and natural selective pressures being applied to those animals that differed from the selective pressures being places on wolves that did not associate with humans.
- the grey wolves that roamed the earth when dogs and wolves diverged are common ancestors between the wolves of today and the dogs of today. They are not the wolves of today. What determines the best fit wolf in the modern age is NOT what likely determined the best fit wolf before the divergence between wolves and dogs. Dogs and wolves diverged either at the beginning of the advent of agriculture or before the advent of agriculture. The world has changed drastically since them- the agriculture revolution and global take over of our species have changed what makes an animal best fit in their environment. In fact, for quite some time, one of the main things that made a wolf best fit for its environment was a natural aversion to humans. We have no way of knowing whether that was true before dogs and wolves diverged.
3) There is actually evidence to the contrary- it seems likely there IS NOT strong inter-species applications.
4) The very fact that dog-dog and dog-human aggression are not linked behaviorally is, I think, a testament to the fact that dogs understand we are not another dog and do not actually relate to us in the way they do to other dogs.

Yes, dominance theory training does work. Sometimes. With some dogs. You can build a very well behaved dog following it. Sometimes. Depending on the dog. You can also ruin a perfectly good dog by using it on a dog that isn't suited to corrections for one reason or another, and it isn't going to build engagement. It might not kill engagement if the dog is naturally highly engaged, but it takes a very specific trained to build engagement using dominance based methods where there is none or little to start. IMO, it relies too heavily on shutting down behaviors (in the scientific sense of the term "shut down", not the colloquial sense meaning "stop). I prefer to avoid purposely shutting down anything in my training. You can very easily create a shut down animal that doesn't do anything unless told to do so because it is afraid of the punishment. It's not the handler/dog relationship I want to foster with my own animals or create between owners and their animals. I agree strongly with the assertion above that it works for reasons other than why the people using it thinks it does, though. I agree strongly that using scientifically proven methods- such as operant conditioning- should be the backbone of any good training program, no matter what species of animal you're working with. I think that training that does not purposely follow operant conditioning is bad training, even if there are very small aspects I like about it.

I do think there are some small parts of dominance theory training that can be ultilized, but I also don't think the methods that are used to "create" these parts of the training are necessary to achieve these things. It's more that there are assertions I don't disagree with, even if I disagree with the methods used. I do think that a dog that has an owner who is in a strong leadership position is less likely to develop behavioral problems for environmental reasons. That doesn't mean they are less likely to develop behavioral problems that are caused by genetics. Most behavioral problems exist because of a mix of environmental reasons (the behavior has been accidentally reinforced, the behavior problem arose because the dog was never taught alternative behaviors that were more desirable) and genetic reasons (the dog was predisposed genetically to develop the problem). Sometimes, you can over ride genetic predisposition with good training before the issue rears its head. Sometimes you cannot, and you then have to wait for the behavior to show and them modify that behavior after it has begun to be shown. Sometimes, no amount of training/behavioral modification will over ride genetic predisposition/genetic involvement, and medication may be necessary in the long or short term.

In my mind, there's a difference between behavioral problems being linked to bad leadership by the owner and the problem being caused by the dog trying to take on the leadership role. I do not think that behavioral problems ever arise because the dog is trying to take on the leadership role. Or close enough to never that I feel comfortable saying that as a generalization, even if 1 dog in 2,000 that is true of.

In my mind, training and leadership are pretty heavily intertwined. Most dogs can be trained manners by rewarding good behavior with attention/praise and play and maybe the occasional food reward and withholding or removing attention/play/engagement when they are practicing behavior you do not want to see increase or continue. I can't think of anything I believe must be taught through the use of physical correction. Some form of non-physical correction, sure. You do need to let the dog know when it isn't doing what you want, but you can do that with verbal interrupters, a non-reward marker, and/or verbal corrections. I also think a very exceptional trainer could train a dog by just redirecting the dog before it gets the chance to do the behavior they would then have to correct, but I don't think it's plausible to expect every pet dog trainer ever to be able to do this. I think mostly professional or high level hobby trainers are the ones who are going to be able to actually create a high aceiving dog through this kind of training, and I think that to do so you have to aquire all the knowledge and experience you need prior to getting the dog or very soon after, have a well lain out training plan, and have that be your utmost goal and not speedy training/proofing.

Pretty much, good leadership is always doing a good job of telling the dog what is and isn't allowed, helping the dog make good choices so that it is practicing the things that aren't allowed as infrequently as possible, and working hard to make sure the dog understands fully what you do want before you introduce any kind of pressure (corrections) to training, as well as having as much skill applying pressure/corrections as you do applying reward.
 
I just got (well, 2 months ago) a young dog who ran wild as a farm dog and had zero training. Seriously, he didn't even know "sit". So teaching him house behaviors has been a bit challenging. And, because I'm human and haven't managed to train myself not to yet :p, sometimes I'll jump up and yell "NO, don't eat my shoes!!", and because he can tell I'm mad he'll show his belly and smile and play bow, throw all kinds of appeasement behaviors, but it teaches him NOTHING. As soon as I'm not paying attention, he'll go right for those shoes again, because while he doesn't like it when I'm mad, he doesn't understand that chewing on the shoes is what makes me mad. As far as he knows, I'm just a person who randomly gets mad, while he's just innocently having a nice chew ;). What's working, so far, is dumping the entire Costco bag of rawhides in the toy box and praising heavily when he chews them, and when he has something I don't want him to have, take it away and hand him a rawhide, and praise him when he chews that. But "expressing disapproval" is not a thing that has ever worked for me in dog training.

Now, some people might manage to yell or hit or use other negative methods and get the dog not to chew shoes. Maybe the dog will be shut down enough to not chew anything. But it's kind of haphazard. The dog might understand what you want, or not. Everything in dog training comes down to communicating effectively with your dog. If they don't understand what you want, they can't do it, no matter what you do.

I think that punishment in most contexts (dog training, parenting, management of employees, etc.) most of the time comes down to ego and revenge.
 
Discussion starter · #35 · (Edited)
I think that punishment in most contexts (dog training, parenting, management of employees, etc.) most of the time comes down to ego and revenge.
Sorry, but if we really believed this, we wouldn't have a legal code nor courts and law enforcement. Retribution and restitution are part and parcel of any ordered social system. No social system works in the complete absence of punishment. That we should aim to make that a last resort is not the argument. But positing absence of it or even diminishing it into irrelevance is not dealing with the real natural world. And yes, we can get into arguments about the effectiveness of our current punitive system--that it isn't restorative enough--and we might agree we need to improve on things, add a little more positive reinforcement, perhaps. But that would not remove the logical necessity of punishment in each and every instance. Not even in a large number of them.
 
One could argue that there are a lot of countries that have legal systems that focus on restitution and rehabilitation, not retribution (revenge) and it works for them. Even better than the legal systems that are more vengeful. But that's OT :).

Even if punishment is occasionally necessary in a legal context, it doesn't change the fact that most of the time, those who punish are doing it out of anger, frustration, desire for revenge, and ignorance of alternative methods.

There are some who use punishment fairly and thoughtfully, and as a last resort. But, uh, that's pretty rare.
 
Discussion starter · #37 · (Edited)
One could argue that there are a lot of countries that have legal systems that focus on restitution and rehabilitation, not retribution (revenge) and it works for them. Even better than the legal systems that are more vengeful. But that's OT :).

Even if punishment is occasionally necessary in a legal context, it doesn't change the fact that most of the time, those who punish are doing it out of anger, frustration, desire for revenge, and ignorance of alternative methods.
Well, we're going to have to disagree on that last point (punishment is demonstrably not "occasionally necessary" in a legal context--a walk through our jail system should show that). Without statistics and hard evidence to back it up, it is just an unsubstantiated assertion. That restorative systems work better can, however, be shown through data. But there, too, one must sift to separate the positive examples from cases where no amount of rehabilitation will suffice. At some point you are going to need jail cells that lock at night. And that's punishment--of the necessary kind.

ETA: I think it is also necessary that we differentiate between those to "R" words, revenge vs. restitution. One cannot understand the legal system and its need for punishment without appreciating the latter.
 
Lol, not going to argue about legal systems.

Do you prefer: "in dog training, punishment (P+) is most often (but not always) motivated by ego and desire for revenge (and ignorance of alternatives)"?
(This does not mean that person is a terrible person or abusive. It just means their behavior is being motivated by ego and/or revenge. Nothing more nothing less. Many dog training books are quite open about their motive being "can't let a dog get the better of you"---that's ego)

Because I will argue for that one :D.
 
I'm really encouraged to read your response, coming as it does with so much varied experience. As a point of reference, have you raised all your dogs from puppies, or have you had an opportunity to employ your 100% positive approach with an older dog you've taken on, say, as a rescue, or out of a bad situation? I'm thinking of cases where the dog comes with "baggage" brought on by bad training and/or circumstances. If you haven't dealt with such dogs, can you imagine some scenarios where a dog with bad ingrained habits (maybe even bad enough to be out to "get you") requires more than 100% positive reinforcement?
I'm not CptJack and I think she's answered well and clearly but to add--

I haven't actually trained a dog that DIDN'T come with some sort of baggage. Ranging from mild like not being taught leash walking at full grown large dog, to medium like random stray with no training, to major like being crated 22 hours a day and given no training at all but lots of trauma. Several dog-aggressive dogs.

100% positive reinforcement doesn't happen in any situation. Dog training, horse training, kids, employees, whatever.

I use mainly positive reinforcement (rewards) and negative punishment (timeout/remove a toy type thing) and yes, some positive punishment (ex. correction with a prong collar). 95% of issues fall into the first two categories AND the more the dog has "trauma" in the past? The more I avoided positive punishment. As in, a prong collar can serve well for a dog that is just over excited with major prey drive but does not serve well for a dog with fear aggression. The rougher the dog's past, the more rewards and TRUST matter.

The thing is, none of that is because of dominance theory. It is strictly a practical approach to balancing the safety of the dog and I while on walks with the lowest possible form of correction. I start with harness walking, if that doesn't work, I move to a martingale or a no-pull harness, then if the safety issue is still a problem, I use corrections from the prong. Or, if the dog finds a no-pull harness stressful, the dog might prefer the prong as it changes their gait less.

I come from the horse training world, I don't recommend trying "dominance" tactics on a 1500 lbs animal :)

And now I'm going to bring something over from the thread and context that started this-

I've been thinking a lot of the concept of dominance itself and the applications that dominance theory trainers put it to. It would be stupid to argue that dogs have no concept of a dominance hierarchy at all. They do. It isn't always a linear hierarchy of A, then B, then C, etc, but there is substantial evidence for individual hierarchies existing between pairs and trios of dogs, though those are often in flux and may be more situational than fixed. That said, I think that the assumption that there are strong inter-species applications is a mistake. That is what dominance theory is based on entirely. The argument for the type of correction techniques they use in dominance theory training is that "dogs correct each other all the time, just like this." It is my belief that some individual dogs, breeds, and types are going to be more open to the concept of a human correcting them like a dog- often, these are breeds where in past decades, they were selected for something to do with training- leading to dogs able to "take a correction" without huge risk of fallout (unintended consequences such as redirected aggression, aggression to handler, severe escalation, etc). Only in the last decade and a half/two decades has positive reinforcement come to play the role it does in dog training. Before that, training methods were based heavily in physical manipulation and correction of behaviors. Being biddable also helps some. Some breeds are VERY poorly suited to this kind of training, and are more likely than not going to have fallout. I'm thinking particularly of northern breeds (Malamutes, Huskies) and a lot of the Asian and Asian-import breeds (Akitas, Shiba Inus, Shar Peis, Chow Chows). These are the breeds/dogs/types with higher than average defense drive, or a low pain threshold, mainly- dogs that are either going to meet pressure with equal pressure or shut down from the pressure.
Dominance theory doesn't have much to do with dog breeds in the sense that adherents to it think it applies across the board. It was based on flawed studies on captive wolves and the author of those studies has retracted them.

Humans aren't dogs, dogs know that. Like you note, hierarchy between dogs is fluid and not the same thing; dogs in homes are different than feral dogs are different than wolves. Just because a dog is open to the correction doesn't mean it is the best choice to training basically.

Unintended consequences are one of the big risks, I mean, why choose something that can turn really bad when you can choose something that might at worst turn mediocre?

Some of the trainers that have made their recent success with dominance type training are actually using some of the most biddable and emotionally soft type dogs around. Ever wondered why their demonstrations of training are with pit bulls types? It isn't because the bullies are physically tough or "need" a strong handler, but maybe consider it is because they are very human oriented and willing to put up with a lot of crap in order to continue working with their humans....
 
An interesting side discussion developed in the Mouthing and Biting thread which I'd like to follow up here.

Here's the thought starter: What are the benefits and demerits of positive (reward-based) vs. leadership (dominance) training? Which do you follow and why? Are there valid points to each we should incorporate in our training, and if so, which are they?
I'll bite. I would describe my training philosophy as "force free, positive reinforcement based training" if I had to put it concisely. The benefits of reinforcing behaviors you like are: it works, the dog loves it, there is little to no risk of fallout if it's done improperly. The tradeoff is results may take longer compared to using corrections.

I don't even think dominance training is an accurate term, but assuming it means those who use positive punishment, the benefit is when done correctly the results should be attained very quickly. The tradeoff is if done improperly it can increase the chances of fear, anxiety and aggression, it does not work for every dog, the dog does not enjoy it, and if it becomes the foundation of your training then your training program is based on fear.

I don't think the latter should be used often at all. But it works for a lot of people. I understand why people do it. I don't think people who punish are horrible people or even horrible trainers in many cases. But corrections are often applied improperly, and the potential consequences are very real.

Here are points from GENERAL dog training I think should apply in every case:
-An understanding of the dog's motivators
-Looking out for subtle canine body language and not just "is my dog doing what I want"
-Proper timing
-An efficient schedule of reinforcement OR punishment
-Realistic expectations
 
21 - 40 of 187 Posts