I’ve been thinking a lot lately about how important it is to remember that “reinforcement” and “punishment” are defined by the receiver. I’m not sure what got me thinking about it. Perhaps it was learning about a study that found cows perceived being yelled at as distressing as being shocked with an electric prod. Or perhaps it was realizing that Willie most likely doesn’t enjoy practicing penning the sheep anymore (the most pressure-filled part of a sheepdog competition) because he doesn’t like the pressure. (Thank you friend Donna for making the suggestion.) Or maybe it was while explaining to friends that getting a free trip to anywhere right now wouldn’t be reinforcement, it would be punishment. This is not the first time I’ve broached this subject.
Here’s what I wrote in 2008 about it: If you told me I could have an all-expense paid trip to Spain or Paris two weeks from now I’d turn it down. Right now all I want to do is to be home on the farm, with Lassie, Willie, and Sushi the cat. I want to watch the sheep eat apples from our wild apple trees, and savor and process the bounty of local, organic produce from the CSA down the road. As much as I enjoyed the trip, more travel right now would be punishment, not positive reinforcement.
What a good reminder that “positive reinforcement” is always, by definition, decided by the receiver, not by the ‘giver.’ This is easy to forget, especially for novice trainers, and so it deserves discussion in any dog training class or private session. For example, the owner may feel good praising and petting their dog if he came when called away from a play session, but it might be the last thing the dog wanted at that moment. (Getting petted while cuddling on the couch is a far cry from getting pets in the middle of romping at the dog park with one’s buddies. Do you want your sweetie to rub your neck when you’re in the middle of a tennis match?) At seminars sometimes I’ll praise and pet a dog for giving up a toy, and ask the audience to grade my “reinforcement.” If I sound happy and exuberant, I’ll get straight “A’s” from the crowd. But then, I’ll ask them to watch the dog’s face and praise as pet as before. This time they’ll notice, that while I was happy-talking up a storm, the dog turned his head away from the petting (I was doing on the top of his head in that pat-pat kind of way that most dogs don’t like) and didn’t look happy at all. I may have sounded good, but I was actually punishing the dog, not reinforcing him. Food for thought.
I think this concept is important for all dog lovers to revisit on a regular basis. The fact is, it is hard to take yourself out of your own reality; we all need frequent reminders that our assumptions about what is reinforcing and and what is punishing are not always accurate. Who knew that cows would find a raised voice as aversive as an electric shock? How many dog owners try to stop a dog from jumping up on them by pushing the dog away with their hands? (And the dog responds by jumping with even more enthusiasm, because moving one’s paw toward another dog is a play signal. Thus, what the owners perceive as punishment acts as a reinforcer to the dog.) I can’t tell how often I see people attempt to reinforce their dogs with something the dog doesn’t want–whether it’s offering a dog a treat when he is too engaged at barking at the window to have any interest in food, or petting a dog on top of the head with no awareness that the dog’s response is to move away.
Although most of us know the more obvious examples of miscommunication between person and dog about what is, and is not, reinforcing or punishing (remember I am using ‘punishment’ to mean ‘anything that decrease the frequency of a response), we all profit by reminding ourselves to drop our assumptions and let the dog’s behavior tell us what is a “P” and what is an “R” to our dogs. I’d love to hear your examples of times when you realized that you and your dog were defining things differently.
MEANWHILE, back on the farm in 2015: Here’s a photo of two lambs on the farm we visited in Georgia. They are here to remind me what lambs look like, because my ewes are so overdue I am at risk of forgetting. Perhaps they have forgotten too? In their defense, they were first exposed to the ram King Charles, who turned out to be very ill, and subsequent died soon after mating with Lady Baa Baa. That mating clearly didn’t take, because she is long overdue if it had. After the untimely death of King Charles, Little Big Man swaggered into the picture, with a lot to learn about courtship. No flowers and dinner out for him, and the ewes were not impressed. Cupcake ran away from him, and Lady Baa Baa’s mother, Lady Godiva, damn near tried to kill him for the first 4 days. I had to keep them separated, with brief dates during the day in which Lady G would lower her head and smash Little Big Man into a post. Eventually, everyone settled down. And hopefully, the girls got bred. Cross your hooves, we have usually lambed by now, or at least started, and I’m starting to pace the floor.
But I digress… I ask you: Is there anything cuter than a newborn lamb?
We do have spring flowers, lots of them. As well as perfect spring weather, nodding daffodils and Phoebee’s, Chipping Sparrows, Chickadees, Cardinals and Bluebirds chorusing in the background. It’s good to have a reminder of why we live through our ridiculously long and brutal winters!
gail lilly says
Please consider re-visiting the alpha dog travesty. A family whose son is one of my clients recently acquired an adorable pup. In preparation for getting the dog they told me about their decision to watch and learn from an Aussie trainer who teaches the whole alpha dog thing from not letting your dog go through a door first to defining food guarding as a sign of disrespect for the human’s alpha ranking.
Naively I attempted to share what I had learned about the debunking of the allegedly wolf based training method but was told that oh no, this guy is great! His dogs are so well behaved etc. etc. While respecting everyone’s right to hold different opinions, the destructive underpinning for this training philosophy, based on dominance, not only diminishes the quality of life for the dog but is teaching the children that certain beings have the right to make all of the decisions for other beings. This happy puppy is now very confused and assumes a belly up position in response to everything. I am not a trainer so my 2 cents is not worth much. The damage done by this well-intentioned method is by no means limited to its intended target: dogs. Our society sadly reinforces the belief that a chosen few have the right to own 99% of the’ treats’. It is small wonder that alpha training instructors are still able to gain advocates who are not interested in research findings. Would you consider reviewing the down side of alpha-dogma??? gail
Gena Lewis says
I loved your book about adopting rescue dogs. Two months ago we adopted a three year old Golden Retriver/pyranese mix(?). She was dropped off in a known dropping part of Houston and spent a few weeks on her own and was attacked by a pack of dogs. She was rescued, had surgery on her ear wounds, and was placed in foster care. We were so happy to get her as our forever dog. She is wonderful, gentle, not destructive, house broken, loves people, does not jump up on you, etc. we have patiently worked with her to bond with us and she has!!!! The only issue we have with her is her aversion to the leash and leaving our backyard to go on a walk. She will just go down and be dead weight. I have friends that say we should force her to go on a leash, but that doesn’t set well with us. We hired a trainer but nothing worked. I can coax her out of the yard with treats, but not far. We are using a harness and that is going ok until the leash goes on! Oh, and she’s a big girl, 75 lbs. do you have any suggestions? Another thing is that she doesn’t play with toys of any kind except a Kong with food in it. We love her just the way she is, but would love for to get more excercize
Stephanie says
I have a new puppy (now 6 months old) during our training I do try to remind myself to pay attention to the puppy and how he responds to me and my rewards or punishment. I often forget and find my self offering the same thing over an over until finally I realize … try something else. The jumping is a perfect example for me. Our puppy jumps and jumps and jumps and I first responded by pushing him down or back, I even try stepping forward towards him or holding his paws (which I learned in a puppy class). I finally tried slowing down, I stand still, stopped using my “paws” and ignored him, I make a calm and quite shhhhh sound. It does not take long for him to sit in front of me and then I sit on the floor and give him the attention is so badly wants. It’s not what I learned in our puppy school but it seems to be working.
I read the other comment from Gail and I would also love to read more about alpha behavior. I have a shar pei and am often told by breeders and rescue groups that I need to alpha roll him every day. I am no expert so I remain polite and don’t challenge the methods they use.
Thanks for all the great information!
Marilyn Mele says
What a great reminder for everybody, not just us dog folk! I’m thinking of all the times my dog was unimpressed with my “happy dance” praise. Poor fella was clearly trying to tell me something. A smarter trainer would have looked for the things that brought out the happy dance in my dog, not just in me! I will be ever more aware of this from now on. It reminds me of the recent news about the article in Science magazine describing the responses of dogs and humans when they gaze into each other’s eyes. According to the article, both the dogs and the humans began to secrete higher levels of oxytocin, the hormone connected to bonding. I thought about some of the clients I’ve had over the years whose dogs would definitely not view this kind of intense eye contact as a form of bonding. Surely some dogs would secrete an entirely different hormone with a not so pleasant outcome. Maybe this article should have come with a warning or at least a disclaimer.
Kathy says
I was just thinking about this concept the other day! I have two dogs: one is very treat motivated and he is so wonderful to take out to go to the bathroom. It’s about a two minute process. He pees, races over for a treat, runs out to the field and poops, and then races back for a second treat. Then he runs to the door to get his feet wiped so that he can come inside and. . . get a treat.
The other dog is ruled by his nose and is terribly frustrating to take outside to go to the bathroom–in the evening especially, since he has to be leashed because he’ll catch a scent or hear a noise in the underbrush and be in the next county before I can turn the flashlight on. So we’ve been spending the winter evenings walking around and around the frozen yard as he sniffs every blade of grass and my “go pee” command gets more and more irritated. By the time he casually lifts a leg, I’m so annoyed that I shove a treat at him (which he takes with very little interest) and rush him to the door to get back inside.
So. I’m a little slow on the uptake sometimes, but I finally realized that the sniffing IS the positive reinforcement and the treat is negligible for this dog. So now when we go out, I stand in one spot for about thirty seconds while he sniffs the circumference of the leash and eventually pees and then we go sniff for a full two minutes (I never short him, no matter how cold it is). I go wherever he wants to go and he sniffs and sniffs and is obviously overjoyed. At the end of two minutes, I count down from 30 out loud, and at “zero,” he actually heads for the house on his own. It’s magical. I can’t believe it took me so long to figure it out. I was “punishing” him for peeing quickly by taking him into the house immediately. Now he’s out there for about three minutes and most of that time can be spent sniffing if he just pees quickly enough. He’s a smart dog and he figured that out in no time. I wish I’d been as smart.
Somebody (probably Trisha) once told me, “you’ve got to be smarter than the dog.” Sometimes that’s difficult for me!
Colleen says
What a timely post. I have hit a wall with my dog and the positive reinforcement issue. If he does not like the reward he ignores me, or worse, he takes himself through all the steps of the training himself and turns into a begging session. So I feel like I have some how trained him to do the behavior for the reward. He is smart and stubborn.
After accepting that I can go no farther on my own I found a trainer through the CAAB website and she comes next Saturday to help me. My dog seems perfectly happy with himself at the moment. I am the one under stress.
Trisha says
To Gail: Argh, what a sad story. I will bring that topic up again sometime soon. Seems it is never enough!
To Gena: What a lucky dog you have! I wish I could come over and help! I’d suggest lots of things: Put the leash on/treat/leash off, over and over during the day. Separate the leash going on from going on walks until your pup gets excited when the leash goes on. Try, also, going to the door/gate and asking for the dog to walk in circles within the door or gate. This is hard to describe in text, but basically you are asking your dog to make a tiny circle close to the exit, and then gradually expand the circle so that the dog walks outside of the yard for a few feet, then yards, etc. What can make this work is that the dog isn’t being asked to walk straight toward something scary, but each step forward is curved and is thus going back toward what feels safe. Make sense?
Marilyn: Love your point about eye contact not being reinforcing for all dogs (or people for that matter!)
Trisha says
To Kathy: A perfect example of confusion between P and R! Thanks so much for sharing it. I wish everyone would figure out that “doing one’s business” so often means that the fun is over. You’re super smart to put a clear, objective time on it… I usually just make myself stay out longer, but I like that you put it on the clock!
And Colleen, I ‘m so glad you are getting a trainer in. There’s nothing I can say without seeing your dog, so a one-on-one session is the perfect solution. Meanwhile, take a lesson from your dog and don’t worry!
Lacey says
For a retriever who doesn’t want to retrieve but likes treats: a tennis ball slit so a treat can be inserted makes a very attractive thing for a dog to pursue. If you are lucky you can get the dog to let you open it for her, so bringing it back to get the treat becomes a natural thing.
Nic1 says
Hi Stephanie, it’s really not impolite to challenge methods if you do not understand why you are being asked to do something. You can challenge your trainer politely, especially as you are paying for a service? 🙂
I am no trainer but I had what turned out to be poor advice (not really rooted in science or ethics) from a dog trainer who was advising me to alpha roll my dog ‘when she disobeys you’, amongst other punishment based techniques. Like yourself, I turned to forums to read around this as to me it just felt barbaric to do this to a confused and anxious newly adopted companion dog. I think that if you seeking advice here, then it may not feel quite right for you to alpha roll your dog? Be great if Trisha was to revisit this as a lot of trainers are still very attached and attracted to dominance and punishment based methods, which isn’t surprising given what we know about our own human psychology.
With regard to positive reinforcement, this has got a lot easier with my dog as she has got older and more difficult with me!:)
She is much more food oriented now and still ball and plush toy mad. However, I recently discovered that she is mad for footballs. In fact a bit too mad in that they can’t be used as a reward as she just gets too aroused. I have a dog who definitely can get too aroused if the reinforcement is TOO exciting, so I also have to monitor that too.
Mireille says
Good post! Spot, my Siberian, is an avid hunter that cannot be trusted of leash. When he smells deer in the woods he does not want a treat! He wants to know where they are and chase them. I showed a video to a friend where I hold a handful of Sausage – his fav treat – under his nose and his behaviour can best be described “Yes yes, yummy, but can’t you see I am busy right now?!” (he tries to look over my hand, under my hand, even pushes the hand away).
So I changed my routine: yes, he can watch and sniff but no, he can’t chase. So when he consents to come with me, I throw the treat. That way he gets to chase and pound the treat. The last treat comes from my hand, so I end up with a dog that is once more in focus and “with me”. It is still a work in progress but we are getting there. I have no illusion that this wil solve everything, that same friend was impressed by the speed in which Spot switches from play with her dog (on a 10 mtr leash) to a full sprint when he sees a deer running in the woods.
Ahum.. 10 mtr leash.. for a horse.. but the clip broke (and not my back 😉 ). Was one of those heartstopping moments to see your husky disappearing into the woods ….
I was so very fortunate that he returned within 5 minutes… without the deer… Another intersting thing happende though: Spot came running to mee, happy camper, me very happy relieved boss. He did not com directly but circled around me. I really was’n angry. But his curve brought him within range of my friends dog, a female Riesenschnauzer with which he had been very happily walking together that day and she gave him a very clear “correction” . Obviously she did not agree with him running away on his own? Spot immediately came to me and flopped down beside me, looking a bit chastised. I was rather surprised buy my friend said it does happen between dogs.
(Spot and his bearded lady friend van be seen here http://3.bp.blogspot.com/-malEUB8b_KM/VSLim1_jwKI/AAAAAAAACoA/PaXnWkNtYZ8/s1600/Shannah%2Ben%2BSpot%2B-Kampina%2B%2B-%2B01.jpg Yup, my guy likes lady’s with moustaches…. )
HFR says
This is so interesting to me. I actually was thinking about the complexities of reinforcement vs punishment this weekend. I was watching what is basically a TV show aimed at young kids on dog training. It’s on Saturday mornings. The trainer on the show takes a dog from a shelter and does a “makeover” on him. Grooms him and trains him. It’s all very simplified obviously because it’s for kids. And it’s certainly a great message to send to them. But what was interesting to me was the “combo” of techniques shown.
The trainer used positive techniques like using a clicker and treats to motivate the dog to do what he wanted him to do. But he also used techniques to forcibly get him to comply. For instance, to get him to lie down he would step on the leash very close to his neck so the dog was gently forced into a prone position. Or he would solidly tap the platform with his hand so that the dog would be startled into not taking a treat that he was asking him to “leave”. He would also say things like “Nope, I’m in control, not you.” Now the dog was not a particularly soft dog, so it wasn’t too aversive, but still.
I’ve seen this more and more lately. A trainer that thinks of themselves as a positive trainer, but still uses some techniques that are based in punishment. Clearly it’s worlds better than some other stuff out there, but I wonder if it’s legitimate positive reward training as I’m sure these trainers would advertise themselves to be. Then again, I try to keep myself from becoming what is being called a “pozzi nazi”, a term which I find very offensive. I understand that nothing is black and white, especially in training.
But if the dog does not perceive a technique to be a punishment (or is not really punished by it), even if we see it as one, does that mean it can be classified as a positive technique?
liz says
If there is anything cuter than newborn lambs, I couldn’t tell ya. So adorable.
An area that is a perpetual challenge is the misuse of language. Being human and all, it will always be my tendency to throw words around. What I’ve noticed is that excessive praise, excessive use of names, improper timing, and casual (not in a training setting) use of important training words will confuse and avert my dogs’ attention. Excessive praise seems to dilute, interrupt flow, or just not match the task at hand. Excessive use of names, particularly when trying to encourage a dog to do something challenging, seems to increase pressure as I already have their attention and most likely their best effort. I could go on, but all of these issues fit into the broad category of a misuse of words, and I believe it will be a continuous work in progress!
Rebecca Rice says
Got any advice on incorporating environmental rewards? This seems like such a good idea, especially for loose-leash walking, but I haven’t been able to make it work. The idea, as explained to me, is that if your dog is pulling towards something, say a bush, you ask the dog to walk nicely on a loose leash, and the reward is that you get to the bush and the dog gets to go smell it, which is what it wanted to do in the first place. When I try this with Pixie, she will look at me when we get to the bush with a “why did you bring me over here?” look, and walk away without bothering to even do a token sniff. So I am obviously doing something wrong.
@Gena – have you considered having your dog wear a drag leash in the house? In other words, just attach the leash to her, and have her wear it around, so that she gets used to it? I did that with my dog for a while, since she had started to have some negative leash associations. (Leash = walk, and she did NOT find walks reinforcing! As several trainers have said, she is a classic “environmentally sensitive” dog.) For safety, I slit the handle so that it couldn’t get caught on things. It does seem to change their attitude towards the leash in a relatively fast time.
LisaW says
@ HFR: Or to coin it anther way: But if the dog does not perceive a technique to be a punishment (or is not really punished by it), even if we see it as one, does that mean it can be classified as punishment? If we are considering what is actually positive reinforcement to the dog, than I’d throw in what is actually punishment to the dog, as well.
My two dogs perceive positive reinforcement very differently, and also, negative reinforcement. One is very sensitive to voice changes and postures, the other is more of a do-de-o-do, not much phases her. I try very hard to reward when they respond well to the task at hand and ignore unwanted behavior, but I’m not always successful. And I’m not always as patient as I should be. But, my response to each dog is different — not always clean and well-timed, but different.
One thing that has been running through my mind for a while now is my more reactive dog does seem to process and hold onto the behavior I’m asking for if I tell her I’m not happy with what she’s doing. For example, she will bark, bark, bark at noises outside, the neighbors cleaning their deck or someone across the street making noise, and I tried to redirect her and change her mind with find it’s and move away’s and play. But, I noticed that when I say firmly, “olive, uh-uh” or “no barking,” she responds more to that. It’s not harsh but it is firm. In terms of this conversation, it is negative punishment, but I’m wondering if I was reinforcing the barking with my prior behavior and not giving her clear enough communication on what I actually wanted.
I’ve got to bone up on my “dog.” 🙂
Nic1 says
HFR, great post.
Perhaps the word ‘positive’, as referred to a training philosophy has become a bit confused in what it’s actual definition is in animal learning? Positive reinforcement is only one quarter of the Operant quadrant as defined by Skinner. There is also Positive Punishment so the term positive when applied to a trainer means… what exactly? Positive as in happy/good/upbeat?? It’s a bit misleading to be honest and inaccurate as most training and learning involves some level of stress and I do think that the term seems to be slightly negative (irony!) in some circles?
I don’t think there IS such a thing as a purely ‘positive’ trainer is there? Learning is all about punishment too. However, it’s deciding what is and isn’t acceptable in terms of how we use punishment that’s really important to me. Most humane and progressive trainers would use a combination of positive reinforcement and negative punishment methods as defined by Skinner. However, what you describe with the trainer stepping on the lead does seem to be an example of negative reinforcement – not the recommended technique to get a dog to lie down that’s for sure. But it sounds like the trainer was demonstrating a wide variety of techniques, including clicker training which is great.
Interesting that BAT (what I consider to be quite a humane technique) has been criticised as Negative Reinforcement in the Veterinary Behaviour world which may be accurate in terms of the operant quadrants. But if the dog isn’t distressed, is it a bad technique?
Also, ‘Force Free’ training – some people may interpret that too literally to mean that you never, ever touch the dog? In practice the message is simply not using choking, pain or prong collars to ‘teach’ the dog, which I stand by.
I would try to choose management over using a positive punishment technique every time. But then, how do you know until you are in a situation where your dog has killed a cat? Is it the shelter, euthanasia, extreme micro management or try a shock collar using an experienced professional animal trainer? Incidentally, there is a movement to get shock collars banned in the UK and most members of the APDT won’t use shock collars to my knowledge. So, do you try it yourself?? I don’t have any answers here as I really don’t have the experience.
A great quote on positive punishment though from Nando Brown – ‘Positive punishment is like a detonated nuclear bomb. If the blast doesn’t get you, the fallout will.’
Gosh it’s such a grey area, just like life, isn’t it? But I do think that Sophia Yin’s philosophy was very pragmatic and grounded in real life and not just theory– ‘don’t throw the baby out with the bathwater’.
Sharon Woolman says
I had a similar problem – the ram I called Napolean was apparently firing blanks because the ewes were not bred when I had them checked by ultrasound. In came Nappy II who apparently got the job done as the girls are pregnant. Was hoping to know if they had twins or singles but the vet wasn’t able to tell me that. But this means I will be lambing at the end of June! But the cost of the ultrasound was well worth it as I could have easily spent that much just in gas getting to the Fogt’s Fourth of July weekend trial.
Annika says
I just LOVE the example of the cows perceiving raised voices to be as stressful as electric shocks! I am reading Honey Rock Dawn by Shreve Stockton and I think I am falling in love with cows and their sensitivity and intelligence 🙂
I can’t find a positive reinforcement that’s stronger for my dog than tennis balls. So there is no reward that actually functions as a reward for dropping it and ending a play session with it. On the other hand, I have a similar problem as someone above in that it is too stress-inducing to work really well as a positive reinforcement in other training scenarios.
He also rates being with us much more highly than food rewards which means that being crated can never be made more pleasant with a treat – sometimes he won’t even take it and even if he does, it just doesn’t do much for him in light of being separated from us or realizing that he will be isolated for a while. I think this is one of those occasions when we read things differently, both because our older dog is highly food motivated and because all the crate training and separation training advice encourages food rewards to make that situation easier.
EmilyS says
I’m one who believes it’s not always “just” about the dog. In the case of aversive, physically harsh training … the defenders say it “works” and they point to happy enthusiastic dogs as justification. I ask: how does it make the PERSON feel to administer such training? Do they really feel NO discomfort about causing their dogs “just a little” pain?
Nic1 says
EmilyS, that’s the complicated human psychology. Some people will feel terrible, some people just won’t. Some people may actually feel vindicated, if the dog is punished for doing something the owner perceives as ‘bad’ and then stops. It works, like you say. And there is the relief for the human. Isn’t that one of the biggest reinforcements of all?
As humans, we have a tendency to do what people in authority tell us to do. Or. If we perceive that person to be an ‘expert’, then we may not feel comfortable challenging them?
I think some people have bought into that with positive punishment based techniques and their dog trainers. It may also be that some people either are not interested in understanding the biology of behaviour or just haven’t had the information. Some people have been very lucky having sound dogs who have survived training with very little fall out.
A big part of the problem is also the alpha myth and dominance meme which seems to promote getting tough with your dog….that just seems so seductive to our primate psychology. So I think it’s actually very complicated, multifarious and is far from easy to resolve.
Naomi says
I have a three year old border collie and there are occasions where food ‘rewards’ for him, never mind petting, are not perceived as positive reinforcements. When running agility, he would much rather be given further direction(s) as to what to run next. I think he would find an 80 item course more fulfilling than a 20 item course with food reward. When we work small practice sets of 3 – 5 obstacles like a drill he will often take a food reward but spit it instantly if he interprets any of my gestures to signal another obstacle. In fact, I think I’ve inadvertently trained him to think he needs to take the treat reward to get what he really wants, to run more agility. Yes, he has tug drive as well, but even that can be a toss up on occasion. He just loves to run the obstacles until his tongue is dragging on the floor. I also have a french bulldog I compete in disc with and she will very readily disregard any other form of reward when playing disc other than more frisbee tosses. Food cannot compete and touching only happens if she recognizes it will be rewarded with a disc. One of my 10 year old Boston Terriers is the same. If she is playing, there is no touching. If you reach for her she’ll back away just out of reach and give an excite bark giving you the hint that all she wants is for you to throw, silly human.
Mireille says
@rebecca; i did use the environment as +r, but the trick is to be precise and quick because Spot will quickly loose interest / shift interest. So he would pull, I would stand still and wait till he himself slackened the leash, then click and move forward. This was my most succesfull loose leash walking exercise. In the beginning I made the mistake in incorporating a food reward : click-treat-forward move but that made him more irritated. It is a very difficult exercise for high drive dogs, it requires selfcontrol so we would do short bursts of it.
Now we have progressed to using a 7 mtr flexi and in training him not to sprint towards the end to see if he can get another half metre of space extra
dieta says
ToGina; I had a similar problem with one of mine. Leaving the lead on all day definitely helped. For Storm home and yard were safe spaces, outside were nasty dogs and people. We sat on the doorstep and watched the world go by and even when we started I turned the moment I saw the slightest bit of stress. And even when we managed reasonable walks it took a couple of months before she would toilet anywhere but the yard. Outside wasn’t safe enough to be caught with the trousers down. Be patient and try not to make too much of a deal of it
Good luck
Eileen says
The first thing that springs to mind is the old water spray bottle, which can be very aversive to some dogs, but I have also seen plenty of border collie types (by that I mean working breeds generally) quickly turn the spray in the face (to stop barking or stop chasing the cat – whatever the owner has been told to use it for) into a game of catch the water spray. Wasn’t it Skinner that said it’s all about the rat (or something similar 🙂 ) It’s all about the dog on the day, because what is rewarding/punishing can change from one day to another and one environment to another. We owe it to our dogs to treat them like the sentient beings they are rather than robots. Love your blogs, and your mixture of compassion and wisdom.
Annie says
My house is very calm, it is just me, my husband and the dogs. I work mostly from home and rarely say a word all day, the dogs don’t bark, we are just quiet people. I had not realized it but it turns out my dogs were used to the quiet and watching my hand signals or responding to soft and short commands. When we were newer in agility we had a very peppy instructor who encouraged us to use lots of encouragement and be very excited for our dogs success. We had been at agility for 3 or 4 months and Copley had been loving it, but for some reason he was less and less excited and more distracted when we were in her class. We figured out what was going on, I was actually upsetting Copley by acting so excited and making noise. He would turn away and I would be told “act more exciting than the thing he is sniffing over there” and he would just look at me and take off the other way. I can only imagine he was thinking something like “she has lost her mind, I’m not going anywhere near that crazy person until she calms down.” Once I stopped acting like I was so excited about everything he did right things really smoothed out for us. For some dogs happy and excited praise seems to do the trick but not Copley!
Trish K says
When I first saw the new blog post picture I had mixed emotions. My first was Awww how adorable, lady baa baa had her lambs and Darn it I lost the contest! Well they are adorable and I’m glad they aren’t yours! 🙂 April 27th , twin males is my guess.
So it was my Bella and a past blog or book of Trisha’s that helped me learn that reaching out over Bella’s head to stroke ( not pat) and tell her what a good girl she was wasn’t a positive experience for her. She would duck and weave and avoid my hand if I came in from the top but was perfectly fine it seemed when I came in from underneath to rub her chest. I’m sure I’ve reached out over towards all my dogs but only Noticed that reaction with Bella. It was at a time when I was reading everything I could find about canine communication and body language. So it was perfect timing. I’m just more aware now of their reactions to my body language which leads me to the topic of dominance as it was defined in some comments above. I don’t think dominance or being the dominant figure in the relationship with our dogs is a bad thing. Is it? I mean if we have an understanding of how dogs perceive our gestures such as hugging or full front greeting and such then it seems to make sense that if we also project a “dominant” posture then our dogs will understand that too. I don’t mean alpha rolling or scaring or harming the dog, more in the way we carry ourselves such as strong straight posture or even gesture eating befor the dog ( I read that somewhere else) and even going through the door first. also by dominating I basically mean that when that pork chop hits the floor there is no doubt in my dogs mind that I will be the one to get it! I mean that in the nicest way :-).
Diane says
Rewards can change with the circumstance. My dog LOVES food – but if he has a set back on the teeter in agility (noise is the culprit), the food lure is now almost punishment. It’s in front of his nose, but he is frightened to even step forward to get it. Since I just run him for fun, I decided on no pressure, and let him run by it without taking it. I would always say “want to try” and if he showed no interest, I just kept going. Then just two times in, he decided he did want to try! No luring necessary. It’s not always easy to try to figure out what they might find rewarding….it this case it was no pressure (then followed by a happily given and happily received food treat once performed!)
Rhonda says
This wasn’t a case of confusion between me and my dog but rather between us and our agility instructor. I had two sheltie in agility classes and while the elder of the two enjoyed it, for her it was really about the time, interaction and being praised by me. At the end of a run she and I, we’d throw ourselves a little party.
My other sheltie, however, LIVED for agility. Not only did she love it, she was smart, fast and accurate. The entire class would stop and watch when it was her turn. At the end of a great run she was prancing, looking up at me in obvious satisfaction and I gave her a restrained but heartfelt “good girl.” Our instructor, however, chided me for not adequately praising my dog. So I went to praise her with a little of the enthusiasm I normally did my older girl but this one stepped back and looked at me lIke I’d taken leave of my senses. What I’d realized and my instructor hadn’t was that running an agility course with me, fast and clean, was the reward in and of itself for this dog. Treats and praise had nothing to do with it by that point. Anything I did after her run was irrelevant.
Lyssa K says
This discussion came up at a very funny time! I just got home from a hike and was mulling over that my typical “P” was actually a R for my dog. I have a golden and we are working on loose leash walking using a 15 ft line (she’s familiar with the concept on a shorter lead.) I do not trust a 10 mo old’s recall to let her run loose in the woods, like my other dog is allowed to.
Anyways, as usual, I immediately stopped when there was pressure on the leash. It took her a few seconds longer than usual to give slack in the leash and the R is continue the walk. After the first 20 minutes, I noticed she was reaching the end more often, and when I stopped, she turned and sniffed for a few seconds before giving slack. My P was being taken advantage of as a R.
She already has a “go sniff” for a reward. So I switched things up for the trail hike and gave her frequent “go sniff” (probably every 50 ft) when the leash was loose. When the leash was tight, I called my off-lead dog back to me, to basically use as a lure so she’d give several feet of slack and then allowed to “walk on.” (A handy hiking cue.)
It was warm today and she has not ventured into the water yet. Letting her into the creek I would consider a R. She followed my other dog out (water is the BEST R for him) but as the golden isn’t used to, she ventured in some and then wanted to race all over the place on the bank, which I normally don’t encourage while on lead. For her, racing on the bank was her R every time she ventured into the water on her own (not really a P, but not something she was that excited about.)
This was all experimental for me today, and I was trying to use rewards at my disposal (chasing my other dog to give slack in leash, rewarding with sniffing on cue while leash was loose, racing like a nut on the bank.) The R my golden pup chose were not what I had planned, but tried to work with it.
Terrie says
I attempted to clicker train one of my dogs. He hated the clicker so much that he stopped taking treats of the clicker was out. To this day, when I work with my other dog, the clicker comes out and he takes off for another room. He loves food, but he hates that click more.
The other dog LOVES tennis balls. But she also has one of the worst cases neophilia I have ever seen. Two or three times of offering her a tennis ball and she just goes “I’ve seen this one. What else do you have?” So I now have a lovely collection of tennis balls that get rotated several times a training session.
Margaret McLaughlin says
Words to live by. Last year I had Nina, my young Flat-Coat, in an agility class. Her performance was consistent, but she was slow. Her training has been as close to R+ as I can get it, using mostly kibble, with soft treats or cheese when I needed to up the ante, but with the problem that too many soft treats gave her loose stools.
I said to the instructor (a good one) “Got to figure out a way to get her out of this pretty-dog canter or we’ll never make time in AKC.” She asked, “Will she tug?” No, none of my polite, soft-mouthed retrievers have been willing to tug. “How about a ball?” Well, she had never shown any interest, but I said, “Let’s try it.” It was, literally, night & day. She is a different dog, wildly enthusiastic, hurling herself thru obstacles. weaving like a maniac. Oddly enough, the tennis ball has become secondary to the pleasure of running the course. I still often throw the ball after the weaves (in training), she’ll catch it & run the rest of the course with the ball in her mouth.
Two caveats:
If I need to train precision, the ball is not the best reinforcer, because she gets so excited she can’t find heel position.
If she’s learning something new, & really needs to do the 10-reps-per-minute for optimum learning, chasing the ball takes too long & interrupts the flow.
In either case, the ball is best used at the end of a chain of already-learned behaviors, like a whole heeling pattern. And if the last sit is forged because she’s hoping I’ll throw it, no cigar. Cheese, yes. Ball, no.
An off-topic offering to the world: I learned a new technique yesterday at a guide-dog puppy training session. It is a refinement on ceasing forward motion when the dog pulls–P-. Not only stop, but clamp your leash hand–or both, if you need to–to your thigh. It removes the variable-reinforcer of getting closer to the desired object by using your poor arm as a bungie cord, & so gives the dog the info that pulling will not get him ANY closer. Not. One. Inch.
I love the quote about punishment-based training & a nuclear detonation. Is it part of a longer article?
And I third or forth the request for more on the whole alpha thing. Ran into it yesterday at the drugstore, of all places.
Lyssa K says
Odd R’s. My golden comes up with Rs and makes games out of them. Jumping into the crate in the car is a big R to her, so she immediately tries to jump in, and once in, wants to jump out so she can do it all over again. I will sometimes park on grass so she can do this multiple times. She thinks it’s the best thing ever. (She has a very controlled wait so i can open the crate & a stay so she does not bolt out the crate.) This is just something I let her do for fun, and will throw obedience commands in-between. We don’t even need to drive anywhere.
In terms of touching, when I teach obedience class I go over much of Patricia’s advice regarding petting & hugging and how it’s not always considered a R in all contexts. Many of the owners don’t really believe me. I use the pat on the head example as a “reward” for a class dog, and ask the owners to watch how the dog responds. That starts to help it sink in.
My 13 yr old is not a cuddler and i don’t ask him for that. R are food, sticks thrown in creeks, tennis balls thrown in the opposite direction of deer, anything to do with a clicker (it’s crack for him.)
My golden pup is actually very physically affectionate, which is strange, but her favorite R in a low distraction environment (and sometimes in obedience class) is to basically do a headstand between my feet while I scratch the base of her tail. Out of 12 dogs, I have only had one other do something simliar to this, and that was my Pyr who used to try to walk partway through my legs so I could scratch her “britches” too. I have really gone over what signals I must have been sending out to encourage this, and am clueless.
The golden is not being submissive, I think it’s more of a fun game for her. I try to distract her with other Rs in class, since it’s a bit embarrassing.
Christina Deaver says
I actually made a graphic of this quote of yours recently and posted it on my blog because I love it so much. 🙂
http://comehereboy.com/fun-stuff/9-quotes-inspire-us-train-dogs-positively/
It’s #9 on the list, ’cause it’s the most awesome.
Love those lambs… so adorable.
Rachel says
This has been a hot topic in my house lately, as we recently added a second dog whose personality and learning style are very different from our first dog. Both are GSDs and are half brothers. Sieger, who is 12 months old, was very food motivated from the beginning and later began to enjoy a rousing game of fetch as a reward for a job well done. Within the last couple months he has transitioned to being almost entirely motivated by playing ball or simply hearing “Good boy.” Keep it simple, and he’ll do anything for you.
Little brother Remus joined us when he was 15 weeks old, and had zero interest in food rewards. All his house training was accomplished with enthusiastic verbal praise and belly rubs. I had tried offering him treats for going potty when and where he was supposed to, and if he took them at all it was only to spit them immediately in the grass and walk away. He is quite the communicator, though, and early that first weekend of potty training solicited the kind of reward he really wanted: after spitting another perfectly good treat in the grass, he threw himself backwards against my legs in what would be best described as a puppy trust fall; I instinctively reached a hand down to steady him as he slid bonelessly down my legs, and as soon as my hand touched his belly he sighed the most contented sigh imaginable. So that became the routine: he’d pee, and then I’d rub his belly, and everyone went inside happy. Now almost 6 months, he has decided food rewards are nice for obedience training, but most of the time he’s still my cuddle bug and is entirely satisfied by touch.
Nic1 says
Margaret, the quote from Nando Brown came from an interview with Victoria Stillwell at Crufts. He’s a crossover trainer so has seen it from both sides.
Trish K, Dominating doesn’t describe a behaviour, it describes a relationship, usually between members of the same species. However, it requires the other individual to submit. When it comes to us and our dogs, we control their whole lives so we really are about as dominant as it gets.
Implementing a rule structure with this in mind serves no purpose to the dog as they don’t understand it, but perhaps helps people feel they are teaching their dog who is boss (more about the human).
Trisha has a stack of posts on dominance and it’s misunderstanding and misinterpretation in dog training here:
http://www.patriciamcconnell.com/dominance-theories
With regard to positive reinforcement, I would say that most of the behavioural focused training I do with my dog doesn’t use it much! It’s mostly negative punishment and some negative reinforcement (BAT). I also have done a lot of operant counter conditioning (around cats, other dogs). However, what strikes me is that whatever way you do it, you are essentially trying to do the same thing – create space and get the dog to do something else. The other thing that strikes me is that the most reinforcing reward for my dog outdoors would be to chase cats, wildlife, livestock and flush large birds. She can’t ever really get to do most of that so even sausage, liver and cheese isn’t motivating enough for her, over a certain threshold. The endocrine joy when she has flushed a pheasant and comes running back is natural instinct and total fulfilment it seems.
So, as other people have pointed out, environment is key to understanding your dog too. Because of this, I have had to exploit the relief motivator (negative reinforcement) outdoors. If she has run off, I have hid and observed her. She doesn’t panic when she can’t find me but she is notably a bit distressed – scanning the environment, ears flat. When I appear and call her with my arms stretched wide, she literally melts and runs like lightning to me. I have done this a few times on different walks and she is much more velcro like.
I think it’s a bit like losing your parents in the grocery store when you are a small child. The relief when you find them is immense.
So, I suppose what I am rambling on about is that you may want to always use positive reinforcement with the best will in the world, but you may not be able to find an appropriate motivator, depending on your dog and the environment.
Rebecca Rice says
When talking about dominance-based vs positive reinforcement training, etc., I think a lot of issues come up because of the difference in how words are used in casual conversation versus behavioral science. Keeping in mind that “punishment” in behavior lingo means “decrease the likelihood that the action will recur”, I’m sure that we can all think of things that are punitive without being punishment. Rubbing your dog’s nose in feces, for example, is definitely punitive, but also doesn’t work to actually lessen the possibility of the dog house-soiling, so not technically a punishment. The problem is that in general layman’s terms, we think of punishment as something, generally physical, done to some one when they have done wrong: spanking, slapping, hitting, grounding, etc. They can be associated with threats: “Don’t make me pull over!” But the effectiveness of the action in actually doing what a “punishment” is supposed to do isn’t considered. Punishment is the action, not the result, if that makes sense.
So things get muddled when talking about dog training unless people are very precise about their terms. Or, unwilling to get hung up on them, which is sometimes the easier course to take when talking to people who have never even heard of the “four quadrants of training”. As someone said previously, it can be easier to focus on the human side of the equation: “How do you feel about having to do X (shock, jerk, scold) your dog when you train them? If I could show you a way to accomplish what you want, and have it be more fun for YOU, would you be willing to try that?”
And finally… a question for Trish, that I have been pondering for a while. I was talking to someone who does herding, and he said that there is no way of training herding that is entirely positive-reinforcement based, because you have to be willing to do what it takes to protect the sheep as well. As he put it, because a lot of herding is arrested hunting behavior, if a dog has run in and grabbed onto a sheep, it may very well take beating it to get it to let go before it kills the sheep, depending on the level of drive that the dog has. I have heard the same thing said about snake-breaking dogs: a carefully timed, appropriately-strong negative punishment (shock) is much quicker and more effective at getting a dog to leave dangerous snakes alone than positive reinforcement methods. So…. are there some things that just can’t be trained effectively using positive reinforcement?
HFR says
@Rebecca Rice, what a great question. It’s one of the reasons I stopped doing herding with my dog. I wasn’t judgmental about the techniques because I realized they were for the safety of the sheep, but I just couldn’t come to peace with scaring my dog repeatedly. It may have been fun for her, but not for me. I once asked that question about snake training to an “all positive, all the time” trainer and she pretty much refused to answer. Doesn’t mean she didn’t know, I think she just hadn’t thought about it before. Interested to hear what Trish thinks.
Trisha says
HFR: In my opinion, good sheepdog training does not involve “repeatedly scaring a dog.” It does sometimes, by necessity, involve stopping a dog from chasing or attacking the sheep. Best practice is to do that with body blocks, or in some cases a gruff voice. Right now I have to say HEY to Maggie on rare occasion, because she likes to bust in on the sheep to disperse tension. Because she has enough experience and skills now to be aware of alternatives, I will correct her with a gruff voice if she does it. There is no question a gruff voice is aversive to her, but because she loves to work sheep, and is resilient and stable, it seems to do little but prevent the behavior (which she does less and less.) (I should add that I’ve spent far more time setting her up in tense situations below her threshold and reinforcing her for being calm by increasing the distance between her and the sheep (which makes her feel better) by simply calling her back to me. But if a dog needs to be ‘scared repeatedly,’ I’d argue either the training methods are flawed or the dog is not a candidate to work sheep. Would that make sense HFR?
Trisha says
I’m loving the comments about what is, and what is not, reinforcing for a dog. Especially the comments that remind us that what works for one dog doesn’t work for another. Or what works one week or in one context won’t work in another. It all gets back to remembering to watch your dog, and pay attention to his or her response to your behavior.
I’m not going to comment on the dominance issue, because it so easily takes us off track. I will, however, take that up in another post sometime soon. Promise!
Beth says
Very interesting post. I was especially interested in the data showing cows find shouting more stressful than a cattle prod. When we draw lines in the sand about what sorts of corrections we will and will not give a dog, I often think that we make those decisions based more on what WE would find objectionable than on what a dog finds objectionable. I have it on good faith from Jack that he finds me ignoring him more distressing than some corrections that others might consider more harsh. And I think I’ve mentioned that, when doing shaping-type training, he finds being allowed to run through a whole string of offered behaviors more stressful than my telling him “no” quietly when he is offering a wrong behavior. Then again, I think I would agree with him. If I consider a social situation, I would rather someone ask me to stop some behavior they found objectionable than just ignore me. Who wouldn’t? And I personally would hate trying to learn a new task without someone telling me I was headed on a completely wrong track. Come to think of it, I may prefer a mild shock (assuming I knew it was coming) than being shouted at as well, being very sound-sensitive.
As far as reinforcement, Jack loves food more than nearly anything, but he loves vigorous games of fetch with his special “human-interaction-only toys” even more than food. So much more that when I tell him it’s time to “trade for treats” at the end of the game, he gives a fairly vigorous stress-reducing shake.
He would rather have the treat than just have the game end, but he would much much prefer to keep playing the game. In this case, the treat (and it’s usually a good big one reserved only for after-game trades) is more a consolation prize than a true reward.
Beth says
I correct myself: the cows found the shouting as distressing as the prod, not more distressing.
Mireille says
I see that part of my previous pos has disappeared 😉 So I’ll write again since it was so funny… Well sort of. I had a video of Spot in full alert mode, I think he saw a deer or something like it. Spot then tends to forget everything around him, including the fact that there is a human attached to the other end of the leash (ouch). I am holding a hand full of delicious sausage under his nose and he trying to look under or over my hand and even pushing my hand away. A fried saw this video and commented he looked for all the world like some-one “Yes yes, I know but cant you see I’m busy right now can you please go an bother someone else” (Sorry, I deletede the video from my iPhone… ) So at that moment sausage was not a reward. Giving him time too look and sniff to his hearts content, then asking him to move away with we and rewarding that is doing the trick. The reward is not a treat handed but a treat tossed. It’s our game of “hunt the treat”. When Spot smells or sees deer (or rabbit, or hare or… ) he is in hunting mode, so I reckoned that instead of trying to get him out of this modus, we use the energy and turn it into a game between us. He loves it. I laugh at him pouncing at the treat so instead of frustration at both sides we play a funny game and end up laughing. Off course it’s not 100%, but he is now getting much better at “spotting” and then redirecting his attention to me. I toss several treats but the last one he gets by sitting and I hand it to him, thereby moving his mental state of beging from “full pursuit mode” into “full attention mode”
This also touches at another theme: what is an aversive. Spot is a strong boy, 26 kilo explosive muscle, when I walk him and his brother I need some aid in controlling them. Otherwise I run the risk that my 62 kilo get dragged into the street when de neighbour’s cat suddenly crosses the street right in front of us. I used a ulead = front ring harness but Spot arches his back in a strange way when I use that which is not good for him. I used a gentle walker after Spot had chiropractic treatment and was forbidden to pull for a while – very difficult for him – but than I came to a dog trainer who said that is was a very mean harness to use and then introduced a halti…. Spot hated the halti. So much that whenever he saw the thing he retreated to his crate. It was effective in controlling him but his behaviour towards other dog deteriorated. One day he saw a deer running, I was controlling him and Shadow trying to walk on and suddenly I felt the pressure disappear and he was walking beside me. I looked at him and suddenly noticed that he had managed to get the halti from his nose. Suddenly it “clicked” : the halti was frustrating him so much, he started fighting it and thereby his arousal peaked. I got out the gentle walker again, and guess what: he comes to me to have it put on and – as part of a mulitfactorial approach – I have a much more relaxed dog on the leash. He and Shad both prefer the gentle walker over the head halter. And I got another dog trainer who respects my choice because she looks at me and my dogs, not at “the books”.
Monika & Sam says
Oh my heavens, those lambies are beyond adorable! thanks so much for sharing their sweet faces. I tend to think of reinforcement and punishment a lot like the ‘glass is either half full or half empty.’ Perception is everything and while one thinks it is one thing, the one on the receiving end can view it entirely opposite. As with most things in life, I guess it’s all about the eyes of the beholder. 🙂
HFR says
Trisha, yes, that does make sense, totally. This is how I was shown to teach my dog herding: A group of sheep in front of me in a barn, my dog on a long rope and me with a rake in my hand. As my dog frantically circled the sheep and got too close or was trying to nip or hurt the sheep, I was to slam the rake on the ground between her and the sheep to teach her to stay away. So she was being scared repeatedly, by the rake. Granted, I only took a handful of lessons, because I just could not get used to frightening my dog over and over again. I’m sure as she learned more, the rake would not be needed as much. Like I said, she still loved the activity, but it was not what I would consider a way I wanted to spend my free time (unlike all the other activities I do with my dogs). But still, to me, there is nothing more beautiful than watching an experienced dog work sheep.
Nic1 says
HFR, that doesn’t sound like fun and could maybe induce a fear of rakes in your dog too? So why would you continue? No positive reinforcement for you either….
Interesting reading about the clicker…….
Do some sound sensitive dogs not enjoy the sound of it? Or is that some dogs get frustrated with the process of shaping? Is that more to do with the methodology or technique, not because positive reinforcement isn’t appropriate or the reinforcer isn’t motivating for the dog perhaps?
My own dog doesn’t enjoy the process of shaping much at all, but she quite likes it when I get the clicker out. I just can’t spend a lot of time with it and have to keep sessions really short. Otherwise, she starts doing these funny sneezes and has on occasion growled in frustration when I haven’t rewarded her when she is trying out behaviours to get me to click. I just don’t think I am that good at shaping and then she gets frustrated when she doesn’t understand what I want her to do. Fair enough!
Vicki in Michigan says
I have a BA in psychology. One day, our behavior modification prof said one of the most useful things I heard in college: “You don’t get to choose what someone else thinks is reinforcing.”
So SO true — for EVERYONE. Not just dogs. Husbands. Mothers. Fathers-in-law. Children……………….
Just because we think something OUGHT to be reinforcing does not mean it is……..
Trisha says
To Nic1: HFR didn’t continue if I understand correctly. She stopped working with that trainer because it was too aversive for both her and her dog. (Have I got that right?) And I have to say: A rake? That could hurt a dog badly, I’m surprised, to say the least, that someone would use that. I’ve seen people throw things like a plastic water bottle, but a rake? Eeeeps!
Nic1 says
Trisha – I totally got that HFR didn’t continue and was trying to reason that it made total sense that she didn’t! Maybe I didn’t explain very well so sorry to confuse!
Trisha says
Nic!: Ah, I re read what you wrote and I mis interpreted it! So sorry! I’ve been out of town a week, barely slept and am pretty woozy. That’s my excuse anyway. Sure is good to be home!
Mireille says
I just couldn’t help wondering whether you could train a dog to modify intrinsically motivated behaviour “all positive”. I wondered because Trishia speaks about the herding and I was thinking about training my sleddogs to run & pull. They love to do that, don’t have to teach them, it’s just go-go-g0 from the start. But…. it is nice if they go in the direction you want them to go…. Their mindset when they run is “hot pursuit” and they are eadily distracted by other things running. So I have to teach them that they cannot run after the hare in the fields. I teach them by applying the brakes (they do not like to stop) and pointing them in the right direction. Sometimes by dragging them back on the path ;-). And then letting them run once more. Which is the ultimate reward at that point in time I can give them at that moment ( going after the hare and catching it is slightly more rewarding 😉 but not an option.). There is always a point in time where I have to tell them “Guys, I know you want this, but it ain’t happening”. I don’t hit them, I don’t scare them, yes I do occasionally yell/ use loud gruff voice and I use force in the sense that I pull on the leash / collar / harness. We are now at the point where a firm voice is enough – also because I am reading their body language better. (side note: when I started with these dogs initially I felt almost blind? I had been training with my previous dogs for >10 years and I just “knew” what they were going to do. I think I recognised sublte changes in body language better, I am now getting almost as good at it with teh two current dogs, took me 3 years… )
You can ruin a sleddog by being to harsh/forcefull by the way or applying too much pressure. An innocent example: one day, in the beginning of my dog-scootering days, I was “nagging” too much. Don’t pee there, don’t drink from that puddle, pass that dog and at one point the dogs just flopped down on the path and refused to move. A very humbling experience because it leave you totally helpless…. At first I was concerned, than a bit angry and then I started laughing, hugged them and sat down with them for a while untill they were ready to go ;-). After that we got to a better understanding.
Nic1 says
Gosh Trisha, even when you are exhausted you are so kind and gracious….
It’s your blog and I should have been clearer really so no apology required. But thank you for allowing us the opportunity to learn so much.
🙂
Hope all is well back on the farm and you get some serious shut eye.
It’s fascinating reading all these comments (as usual) and got me thinking that the economics of behaviour in dogs is really complex because they can’t talk to us and for us to be effective we really have to understand what they are ‘saying’.
If we wanted to measure positive reinforcement and a dog’s perception of it, how could we do that objectively?
Dopamine upregulation?
Bruce says
Fascinating post and thought-provoking comments, thanks to all!
Using the dog’s behavior to tell us what is “P” and what is “R” is so useful. Red Dog does not view even gentle strokes on her head as a reward – she sees it as play initiation, or she jerks her head away, yawns, etc. I am acclimating her to head touching, but I do not use it as a reward. In contrast, our Pug loves to have her whole face scrunched, and “purrs” with pleasure when her ears are rubbed.
When my mother passed away we inherited her sweet but nervous older Samoyed. Simply getting the Sammy from my mother’s apartment to the car was a challenge. I forgot to bring any treats, so when the Sammy stopped the only way I could lure her further was to sit on the ground facing sideways and making gently encouraging noises. I had to do this quite a few times (thank goodness it wasn’t raining!) until the Sammy eventually concluded “Oh, we are going to the dog park”, and trotted along happily. Fortunately, I had parked near the dog park.
Luring the Sammy into an unfamiliar car presented a similar challenge. After about ten minutes I gently picked her up and put her in the front seat, and she accepted the car ride with reasonable contentment. The Sammy is strongly food-motivated, so she quickly learned to use the dog door with food rewards. She gets along swimmingly with Red Dog and the Pug, although she has not yet accepted Red Dog’s many entreaties to play (lots of ear-licking).
We are starting basic training with the Sammy using food rewards. The biggest challenge will be her love of playing “you can’t catch me” when out in the yard (which will be scary if she ever gets loose). It will be a nice challenge to find a bigger reward than her love of independence!
Rachel says
Nic1- I’m guessing that if we really wanted to quantify the positivity or negativity of a method as perceived by a dog, we’d either have to measure dopamine changes or use something along the lines of an EEG or (if we wanted to be really fancy) functional MRI scanning. I’m not sure how tolerant dogs would be to any of that, though- I think there would have to be some significant conditioning to the equipment first before meaningful data could be obtained.
Nic1 says
Hi Rachel, very difficult in practice as you say. I was thinking about fMRI as dogs have been conditioned to it using R+ and CC.
What I am taking home is that although it is very useful to understand the operant quadrants, what we define as being acceptable and unacceptable as reinforcement can be extremely fuzzy. Just thinking of the water spray – dogs who like it (and seem to get a huge thrill) and dogs who find it aversive. You could probably get a lot of information from body language too.
With regard to BAT, the dog is rewarded for displaying threat reducing signals/calming signals, by being allowed to move away from it’s trigger. I wonder what is going on with cortisol and/or dopamine there? I think it’s a powerful and effective method and have used it a heck of a lot.
I read a study recently that cortisol levels were raised in clicker trained dogs in novel environments. I’ll see if I can dig it out. Clicker training is mostly associated with the reward centre of the brain. Possibly the novel environment is the source of stress?
Nic1 says
Just to follow on from the clicker/shaping observations:
http://www.sciencedirect.com/science/article/pii/S0168159110000754#
‘Rapid shaping of behaviour associated with high urinary cortisol in domestic dogs’ Blackwell et al.
‘This study has shown interesting associations between physiological and behavioural measures of stress and ability to learn a simple classical association, and two operant
associations in dogs entering a rescue kennel.
The precise features of kennelling that induce stress have not been separated, but they are likely to include spatial restriction, an unfamiliar physical environment, the proximity of unfamiliar people and dogs, and separation from familiar people and dogs. In addition to the effects of social and spatial restriction, cortisol in dogs is known to increase at veterinary examination, on exposure to aversive visual and auditory stimuli and this increase to be reduced by human interaction, provided this is given by persons experienced with stressed dogs .
While it might be expected that stress will generically impede ability to learn, the results of this study suggest that this is not necessarily the case, and that the relationship between stress responsivity and learning is not a simple one. Indeed, dogs showing high levels of cortisol appeared to learn particular tasks more readily.
However, dogs that appear very attentive and learn quickly, and therefore appear to have
good welfare, may actually be in a state of greater HPA axis activity than those which learn more slowly. The extent to which cortisol levels relate to emotional state, and hence welfare, requires further investigation, and studies involving retesting of previously trained tasks would be needed to investigate the effect of stress on long-term memory in this species.’
Lacey says
@Nic1:
Thanks for the link to the abstract. Very interesting!
Nic1 says
I have access to the paper if anyone would like further information?
What I take away from it is that although the operant quadrants are essential to understand learning, when you have an environment that’s stressing the animal, can you really afford to be chuffed about using R+?
Seaworld comes to mind….
Mireille says
I am not surprised that stress does not impede learning. Stress is not always negative, it is also a way to grow. There is such a thing as “healthy stress”. Think performance athletes fi. It’s chronic stress and stress that is “inescapable” that produces the ill effects. Studies in humans show that a rewarding but busy job is much less stressing than being out of work with no way of bettering your situation.
Nic1 says
Any amount of learning has some stress and it is well acknowledged that learning IS about getting out of your comfort zone, otherwise how else do you develop? The important things is not to venture into the ‘brown’ zone too often (it’s called that for a reason right?). And the challenge is recognising our dogs brown zone when we are working with them. So, even when we use techniques built around positive reinforcement and are not necessarily aversive, some dogs still may find the process more stressful than others.
It interests me because of my dog’s stress attack during shaping using the clicker.
‘arrgggh! I don’t know what you want me to do!’ she seems to be saying.
Judi says
With respect to rakes in training herding, they are leaf rakes with flexible plastic tines, not garden rakes with rigid iron tines. They give a larger visual signal than a stock stick and can be used to separate dog and stock if necessary. Ideally they’re used mostly to give a visual block to the dog (wouldn’t it be easier to go the direction that the rake isn’t blocking?) although I’ve seen them used to block a dog rushing in to grip (many young dogs in my breed are prone to gripping indiscriminately in their early stock experiences).
Yes, you can hurt a dog with a leaf rake, but you can hurt one with a stock stick, too.
Nic1 says
http://eileenanddogs.com/2015/05/05/myths-about-positive-reinforcement-based-training/
http://eileenanddogs.com/common-misconceptions-positive-reinforcement/
A couple of nice blog posts on positive reinforcement. Pragmatic, reasoned and explains a little about why some people struggle with the concept.
‘The “cultural fog” around learning and behavior that Dr. Susan Friedman refers to makes us leery of reinforcement, and can cause us to equate it with mere indulgence or even moral corruption’.
Jenny H says
I have a very nervous German shepherd Bitch, and have done a lot with her to get her to be able to cope in public. (And talking of THAT, she needs regular social experiences or she goes back to being a quivering wreck).
Following regular and oft repeated advice, I used to take extra special treats with me when we went out together. Until I notices a definite reluctance to come with me when I had those treats with me. Trying various things I realised that she simply does NOT like food treats when out and about. I no longer carry any food with me when we go for walks. I DO wear a treat pouch at club training, but Sal prefers praise and being allowed to ‘hold hands’ with me.
She get her extra-special treats now as SOON as we get home again.