Have you ever had an epiphany? Wherein all of a sudden some information you had been turning over and over in your mind fell into place and created an entire new picture? It has happened to me a handful of times in my life, and in each case the result was that I changed some basic beliefs.
Trainers who have switched to positive-reinforcement based training from more aversive-inclusive methods often refer to that process as “crossing over.” I have written about crossing over in many bits and pieces over the years, plus in a couple of longer articles (I’ve linked some at the bottom of this post).
For me, crossing over brought an epiphany. I have always wanted to explain it and to share it far and wide. The other day I had a realization that just might help me explain it better. You could almost say I had an epiphany about my epiphany!
A Changed Outlook
The epiphany I had when crossing over to positive reinforcement-based training was about far more than training. It changed my outlook on life. It was like a bunch of building blocks got knocked down, then magically fell into place into a new and better pattern.
The epiphany was about learning. Not only about how dogs learn, but also about how we all learn. It wasn’t that I was suddenly dedicated to the least invasive methods possible, though I was. It wasn’t some kind of promise never to give a correction again, though I certainly had and have no plans to. It really wasn’t about what kind of trainer I was planning to be.
It was that I suddenly “got” reinforcement. I got that the behaviors I saw my dogs do: desirable and undesirable; cued by me or not; in competition settings, at home, or on a walk: these behaviors were happening because the dogs got something they wanted through doing them. (I would add now that it also could have been because the dogs avoided something undesirable, but that wasn’t part of my epiphany.)
If my dogs were repeating a behavior, I was either reinforcing it, or allowing it to be reinforced.
When my dogs did something I didn’t like, they weren’t challenging my authority. They weren’t giving me the paw. It probably wasn’t “coming out of the blue.” They did it because it worked.
I could suddenly visualize a sort of map of my house with sketches of my dogs (and me!) in different areas, color-coded for the frequency of the behaviors. I realized that behavior is a map of reinforcement. On my imagined map, stronger, habitual behaviors are colored darker than less frequent behaviors.
So actually, the epiphany was about the science.
But my very next realization was about ethics. How could it be fair for me to punish a behavior if it was there because it was getting reinforced, or worse, because I myself had reinforced it? And how could I now overlook the fact that inconsistency on my part (read: intermittent reinforcement) was keeping the unwanted behaviors alive so effectively? This was all on me.
If the dog is doing something because it gets reinforced, doesn’t it make more sense, and isn’t it more fair, to work to remove the reinforcement first, rather than jump into punishment mode? If you jump into punishment without identifying the reinforcers, that means that part of the time the dog gets reinforced, and part of the time she gets punished, for the same thing! Plus as long as the reinforcement is there, you will never get rid of the behavior unless you are willing to escalate the level of the punishment massively, and perhaps not even then.
The thing about removing reinforcement is that it requires us to change our behavior, sometimes in some big ways. No wonder we resist the idea.
One of Many Examples
This reinforcement/punishment combo is very common and easy to fall into, because of ignorance about how learning works, but also because of the unwillingness of humans to change their own behavior even when they might know better. It becomes the norm in many dogs’ lives. Here is but one example.
It is a standard recommendation in traditional obedience training that when your dog pulls out of position and tightens the leash, you give a “correction” in the form of a quick jerk on the leash. This is translated to the dog’s collar and from there to the neck. This is a movement that humans quickly get very good at. And even if the dog is wearing a flat nylon or leather collar (i.e., not a prong or choke collar), getting a correction has got to be at least jarring, and probably in many cases painful. The most extreme form could cause injury.
But many say this is necessary. We are told anything from “you need to be the boss” to “that’s the only way the dog will learn what not to do.”
OK, so that’s half of the story. Now forget about the corrections and the rhetoric for a moment and visualize this. You are walking along with your dog on leash. The dog sees her best buddy and pulls you in that direction. Or you are walking along with your dog and there is evidently something just out of reach that smells wonderful, because your dog changes her course and goes over to smell it, tightening the leash momentarily on the way. Or you have just let your dog out of the car, on leash, and while you get your gear, she is walking the circumference that the leash allows, sniffing around and exploring, and tugging a little when there is something good just out of reach.
Now let’s think how positive reinforcement works. The dog does a behavior, receives something she really likes, and the behavior increases in the future. That means that in the case of each of those three examples, pulling on the leash (the behavior) probably got reinforced (access to the goodies). The dog pulled and got something she wanted. It’s as if she pulled on leash and you handed her a treat. She will likely pull again.
Keep in mind you are also probably having training sessions where you hand her a treat for the opposite behavior: staying by your side. And jerking on her for pulling. Again, the situation for the dog is that the same behavior gets reinforced sometimes and punished sometimes.
Dogs are good discriminators, so they can learn after a while which are the situations in which you are more likely to let them pull vs. when they will get the collar pop. This is one reason you will see plenty of dogs trained in competitive obedience who drag their owners to the training building (and again I will confess that I have been in this group).
But it’s just not fair. So many dogs live in this chaotic and ever-changing combination of punishment and reinforcement, yet we are encouraged to believe that the problems that arise are because we haven’t “taught them their place.”
Are We Just Robots, Then?
There are still a good many people who have a gut-level rejection of behaviorism, even when applied to animals. When presented with my epiphany, or particularly my vision of the “map of reinforcement,” some may have a negative response. It’s reductionist, they might say. It seems to represent our animals and ourselves as little machines, we don’t live in Skinner boxes, it’s too specific or not specific enough, it leaves out emotions, not everything fits in “quadrants,” ad infinitum.
For what it’s worth, everything I have learned in my very beginner-level studies of behavior analysis has shown me what amazing creatures our dogs are, what amazing creatures we are, and how varied and subtle the processes of learning can be. How do we interact with our environment? What effects do environmental stimuli have on us? It’s more like a wonderful, fractal dance than the cold, clinical image many people still have in their minds.
(Besides, it’s the new, sexy field of neuroscience that is presenting evidence that humans may not have free will, not the applied behaviorists!)
The tiny bit of the great field of behavior analysis that I have learned has taught me how to enhance my dogs’ happiness and fun in the world, and taught me to at least start to play fair.
****************************************************
Anybody feel like sharing a story about behaviors that you have (or have been tempted to) both reinforce and punish? Or any behaviors that continue even though you can’t figure out what the reinforcer is?
Related Posts
- I’ve written about the joy of living with dogs without worrying about dominance stuff here: Respect is SO Last Year!
- More about how it isn’t fair to reward and punish the same behavior: Three Behaviors I’m Tempted to Punish, and Why I Don’t
- My later analysis of crossover and the different stages I went through.
- Using trick training as an aid to crossing over: “Just a Trick?”
- Being tough on ourselves as positive reinforcement-based trainers: “Being Tough.”
Thank you Yvette Van Veen of Awesome Dogs for the graphic. It is an Awesome Dogs Shareable! It may be shared following these guidelines.
© Eileen Anderson 2014 eileenanddogs.com
Note: My text in the graphic strongly resembles a sentence in Jean Donaldson’s Culture Clash, page 158: “If a certain behavior is occurring in the first place, it is, by definition, being reinforced somewhere, somehow….” This was not deliberate on my part. I love that book and it’s possible that that sentence sank into my psyche, or it could be a coincidence. Anyway, a nod to Jean Donaldson for saying it first and best!
hahaha, this made me laugh coz i immediately see myself reinforcing things inadvertently but sometimes with knowledge of what i am doing, but no will power to change. i think the ‘out of the kitchen’ sometimes is one of them. but though i feel frustrated sometimes, it is more at my own lack of will power when the sous chef comes to encourage and help….. lol
Sharon, I think every one of use does it. And you’re right, it takes all sorts of will power to refrain. Thanks for the comment!
Thank you for this. Another great blog post. Were you sitting in on Susan Friedman’s course? 🙂
Thanks, Laurie. I didn’t sit in this year. I was there in 2012 and 2013. Plan to return next year. Were you there this year?
I do this. I know I do. My nine-year-old Basenji has been rocket powered all his life. He was born to run. He runs to the end of the leash, or to the point when I hit the brake, then he throws his weight into the collar like a Malamute and tries to drag me onward. (I use a retractable when we walk around the acre I live on but can’t afford to fence completely).
Sometimes when my shoulder or elbow is sore from his efforts, I try keeping him on a shorter leash and praising him whenever he lets it stay slack for a few seconds. Apparently that is his cue to kick it into afterburner and take off again. I know it’s counterproductive that I knd of love that about him. Even though I want him to please, please slow down.
I’m thinking if I can find the right harness for him and use a different leash, and treats, I might be able to shift his mindset. I’m still searching for that harness.
I really enjoy your posts and I’ve learned a lot from them.
Hi crazybasenji! Teaching a dog to walk on a loose leash can be really difficult, I think especially the first time you try it. (I still struggle with consistency after four dogs!) For a skill this demanding of the dog–our speed is just not comfortable for most of them–I would definitely use treats. Good luck! Good idea about changing the gear, too. Perhaps you can change the whole picture and start afresh. Thanks for the comment.
Our 10 month old, Lab, Pointer, Whippet mix, loves shoes and socks. When he picks one up, we ask him to drop it and he will. He is not rewarded with a treat, just a “good boy”. We try to close the laundry room door or fold clothes when he is not in the area, but he always manages to find one. We say, Leave it, when he even “looks” at shoes or socks. That works for a few minutes. He plays with his toys, our 14 yr old Shelty and family members. He, also, goes for an evening walk. How are we reinforcing this bad behavior? Help.. We are running out of socks!
It sounds like just having the sock in his mouth is rewarding enough for him to keep trying the behavior. IMO in order to reinforce not going for the socks, you would have to come up with an alternative that’s more rewarding. When you tell him to “leave it” and he does, how are you rewarding him?
He is rewarded with a high pitched “Yes..Good Boy Coal, Good Boy”. So, should be go back to using a food treat for this leave it? He is rewarded with treats for tricks he performs and good behavior (not all the time). Thank you for your help. He is a sweet dog.
Food treats are generally lots more potent in changing behavior than praise, so yes, I definitely would. Just be aware that you’ll still need to be there him to cue a “leave it.” It takes a lot of work to generalize that kind of behavior to work when you aren’t there. Far better to prevent access to socks as much as possible in the first place.
This is a video about countersurfing, but you could switch in shoes and socks just as easily. It shows the steps you would take to build a **habit** of leave it for certain items (called a default Leave It). This is just to give you an idea of some of the other steps you would need to do to get Coal to leave stuff alone automatically, even when you aren’t right there to supervise.
He does sound like a sweetie, a lovely dog and lucky to be in your household.
Eileen and SaraJane thank you so much for your help. How did you know that we do have a problem with Countersurfing, too?? Eileen, thank you for the video. We have had a lot of changes in our household. Coal came to live with us shortly after our three year old Lab died, suddenly, from a Blood Clot the day after Christmas, and then our 16yr old Whippet died in July and now I am home after being laid off from my job, so I believe he is trying to adjust (and so are we). I will follow your suggestions and let you know our progress. Love your blog.
I’m sure he likes the praise BUT for the majority of dogs it’s never going to be as rewarding as having something in their mouth. You can go back to food rewards for dropping/leaving the socks, as well as giving him lots of other opportunities to earn rewards, and as Eileen says, limit his opportunities to practice the sock behavior as much as possible.
And I would be inclined to try to find a toy that he likes just as much as socks. I’m thinking if he loves socks, he’d love fleece toys, although at this point a plain fleece toy might actually be too similar and might confuse him about what’s allowable and what’s not. On the other hand a fleece toy with a tennis ball or two braided into it might be similar enough to a sock that he’ll love having it in his mouth, but distinctive enough that he’ll recognize it as his toy and not a sock. Does that make sense?
Sarahjaneb, thanks for this great advice to Mary Lynn!
Mary Lynn, you said it in your first sentence. He loves shoes and socks. I agree with sarajane. This falls under the “allowing the behavior to be reinforced” part, rather than your reinforcing it directly. He doesn’t need any external reinforcement from you to seek them out. Mouthing them is self reinforcing. Seriously, once this gets started it is hard to stop. It sounds like he gets attention for plenty of other stuff, and has toys to chew, and those things are important. But it is a lot to ask of a young dog to refrain from grabbing their favorite things. I would be trying to prevent his practicing it, absolutely, keeping those shoes and socks out of reach. Hard with a family, I know.
For some dogs, teaching them to bring you stuff instead of chewing it up is fun (my Clara does that–brings me anything she finds and gets rewarded richly for it). But you have to think through whether you are OK doing that for the rest of his life. Also, if that’s the only way he can earn a treat (I don’t know if you reinforce other behaviors around the house), that means he’ll go out of his way to find stuff to bring you. I’m OK with that with Clara–she has definitely slowed down now that she is three years old. But some people wouldn’t be.
Good luck. Good for you for even asking. Sorry I don’t have a magic fix!
Try phrasing any behaviour problem in a pattern.
Dog looks at socks – owner stops what they are doing – asks dog to leave it – gives attention. “Look for socks for attention.”
or if that doesn’t work…
Grab a sock, owner says leave it, drop sock, I’m a good boy…I grabbed and dropped a sock just like mom wants.
(Being a little tongue in cheek adding the dog’s interpretation into it. But hope it helps explain how a dog could think that socks lead to reinforcement.)
Try reinforcing something other than looking for socks. Otherwise you risk getting into a pattern of behaviour where you have to watch him like a hawk and continuously have to say “leave it.”
If it’s okay with Eileen, I have a blog post on creating an environmental cue for leave it. The trick is to reinforce other behaviour, and to add duration – time. It’s not enough to leave it alone. Leave it alone for a period of time. Then get addicted to doing other things (that you reinforce.)
I think is the blog Yvette was talking about: https://awesomedogs.wordpress.com/2014/10/06/context-as-cues/ It has two great videos attached that are pertinent as well.
I recently watched a trainer teach a student to step on the lead, immobilising her dog, if he barked during a training session. This was apparently an ongoing and annoying problem for the student. A short time later the student was chatting outside and I watched as her dog jumped up in front of her, barked and threw itself into heel position – the student clicked and treated. With the best of intentions the student not only reinforced the heel position, but also the barking, jumping and prompting for a click and treat. It struck me as very unfair that the trainer had not tried to establish the cause of the barking, that instead she moved straight to punishing the behaviour. Was this fair on the dog?
It is our responsibility to make sure (as best we can) that our communication is clear. Great article explaining this Eileen! Thanks!
> With the best of intentions the student not only reinforced the heel position, but also the barking, jumping and prompting for a click and treat. >
This inadvertent training of “chains” is one of the major problems which we must overcome with “positive training”.
I caught myself with a very bad one — I had trouble with my dogs enjoying bark-fests (with the neighbour’s dog) at the fence. Having “crossed-over” instead of yelling at them to stop, I would call them back, send them to their crates, shut the crate doors and then give them a treat for having gone into the crates.
SO, I ended up with dogs who would run to the fence, bark up a storm (about nothing!) run back into their crates to get a treat!
Hmmm. Rethink! I actually went back to “punishment” — into your crate and get ignored until I am no longer angry with you. It doesn’t work as punishment because they still love a bark-fest, but it DOES work as “management” 🙂
It is SO VERY hard to eliminate environmental reinforcement 🙁
It is indeed hard to eliminate or compete with environmental reinforcement. I’m glad you mentioned that re-crating and a timeout didn’t operate as punishment–folks, Jenny is making the distinction that though it was something the dogs probably didn’t prefer at the time, the timeout didn’t work as punishment because it didn’t reduce the barking in the future.
One way to combat the setting up of these types of chains is to give dogs a lot of opportunities for reinforcement in daily life. That’s why I cautioned Mary Lynn that the down side of teaching a dog to retrieve things it liked rather than keeping them and chewing them was that it would actively look for stuff to bring her all the time. But if the dog can also earn a treat by settling on a mat, going to his crate, giving you eye contact when another dog barks or the doorbell rings, or even playing with an appropriate toy, this will generally be mitigated.
Oh wow, Julie, good example. I agree that this doesn’t sound like a very fair situation. And those barking behavior chains are very hard to get rid of. The trainer will probably be trying to punish the jump/bark/heel thing too after it starts to get annoying. Glad you liked the post, and great example.
Hello Eileen, thank you again for your posts.
Now that automn is here, we have a different environment : the entrance door is closed. So Inouk has to learn she has to be either inside or outside. And it is a hard choice for her : she’d like to be outside with me. I am progressively letting her alone outside in hope she’ll learn to find distraction by herself (she has 8 hectares to explore). Please note I do walk her off leash one or two hours everyday, outside our property.
Not surprisingly, the behavior she has developped is jumping at the door and scratching in an attempt to get in. I may have started this inadvertively : our door doesn’t close well, it’s hard to close properly, and she discovered that jumping could open the door. I quickly made shure the door is correctly closed, again, I am not shure she got the behavior from this but most chances are she did.
Secondly, I think she learnt that if jumping at the door no longer opens it, it does draw my attention. So now, how do I handle this situation ? I have given myself 2 rules : 1- I don’t move or look at her until she has stopped scratching for a little while 2- before reaching for the door handle, I ask her to sit, stay, and release her to come in (I’m so proud of this behavior, I use it whenever applicable).
But Inouk still gets reinforced for jumping and scratching the door from time to time : that’s when my husband walks by and reaches for the door handle…. and as he once said to me ” you’re not going to train me like your dog “. Well, on this particular matter, maybe I will, because he will try to avoid an aversive : paint the door frame when it’s all ruined 😉
A good day to you,
Hi Josephine, always nice to get a comment from you. Oh wow, just think how reinforcing that was when she jumped at the door and it opened! AS long as there is a breath of reinforcement for that behavior, either directly from your husband opening the door, or indirectly if she gets your attention in any way, it’s going to stay alive. One thought I had–Would you be OK with Inouk asking to come in a different way?
If you are OK with her asking to come in, and are willing to open the door when she does, then you could probably set her up with a bell. If not…sounds like your husband will have to paint the door frame every so often.
Hello and thank you for your time Eileen.
A bell ? such a great idea though I don’t think I have the skills to teach her that. I don’t mind being her doorman, but I’d like her to cope with staying alone outside for a while (depending on the circonstances between 30mn to 1hour). We have a glass door, and I can see Inouk from my desk when she’s waiting to come in, but she doesn’t see me.
Why do I want her out sometimes ?
She is absolutly glued to me, and my idea is gradually help her being without me and ultimatly by herself. Secondly, she does get distractions when she’s outside : she gets up to locate a smell, the stands on a elevated position where she can check what’s happening on the road, she can bark at times, and eventually she will go for a walk along the neighbours fence. All reasons that make me think she’s better occupied outside than sleeping inside while I’m at my desk. But is she happier ….?
So I will do my best to teach her ” it’s not so bad do be outside “, you would say ” a neutral emotional response ” won’t you ?
Meanwhile, family board has adopted the decision not to open the door until Inouk is calmly sitting or lying.
I would go for a positive conditioned emotional response, rather than neutral. Here is a really good blog post about that: http://kimpikespositivepaws.wordpress.com/2014/11/05/what-is-a-cer-and-why-do-i-care/
Since your idea is for her to have a good time outside without you, the only easy road to that I can see is through food. A couple of ideas: Can you get toys like Kongs that you fill with food and freeze? (Freezing makes it more of a challenge to get the food out. If Inouk is not familiar with this kind of toy you’ll need to teach her first.) Another idea is to give her part of her meal outside, hidden so she needs to search for it. Again, start easy if she hasn’t done this, but it’s usually pretty thrilling for most dogs.
It’s quite possible that all these would do would be to teach her “it’s fun to find your food outside,” then she’ll want to come in again immediately, but you’d have a better chance of her having a good time for a while.
You really have two issues here, and both are challenging. One is for Inouk to be happier in the yard by herself. The other is for her to stop scratching the door. If you solved the first one, and she was happy to be outside, it would help the second, but it wouldn’t entirely fix it. She is still going to want to come inside sometimes when you haven’t thought of it.
Finally, a couple of cautions about the decision not to let her in until she is calm and quiet. One risk is that she will learn to make some noise to get someone’s attention, then be quiet to be let in. And that actually makes sense since it’s much harder to notice a dog who is being quiet than one who is scratching and barking. Another thing is that if family members suddenly stop letting her in when she is noisy, they are setting up the situation I described in my last post. Something that used to work for her doesn’t work anymore. That’s hard on the dog, and also it is likely a human will fail along the way. It only takes an occasional mistake to keep the behavior alive and well. Most people can’t outlast an extinction burst.
That’s why I’m suggesting a bell. Think about it–the behavior is not a whole lot different from what she is doing now. If you hang some jingly bells by your door, it’s almost the same action for her to ring them as it is to scratch the door. I bet you could teach it (starting indoors first). Good luck!
Hello Eileen, can’t forgive myself to have left you without a thank you after your long and detailed answer : THANK YOU Eileen !
Actually, I came back to this page to give you an update.
I don’t think Inouk is borred outside, she wants me to be out with her, even if I’m doing nothing but sitting. And when I give her the choice, well, she comes inside to be with me, after a looooong hesitation. All this is very common, and I’am afraid very boring to you.
She’s starting to go on excursion by herself, especially when I’am away.
Kong : Inouk has one of her meals in a Kong, she often leaves the last kibble inside, she leaves a few kibbles on the floor, too far, she comes back to me for more… So Yes, I will need to teach her. Also in training cession, it happens often that she doesn’t find a kibble thrown to far… and because things have to go quickly, I lead her to it.
Consequently, I don’t think she is ready to search for food outside, but I’m probably wrong and I will definitely try. In fact, she finds some goodies : the seeds (or dry fruit) falling from the lime tree, she adores it.
It makes perfect sense 1/ be happpy outside 2/ stop scratching the door.
Well, I’m trying to teach her she can come in anytime she wants, when she has enough out garden, come to the door, I see you and let you in. The consequence to that, is that she is more patient, she waits because she knows the door will open. The scratching has almost completely desapeared, and I will be very carefull of that ” occasional mistake”. It’s only my husband and I, and we agreed on that.
As for barking : I think it must have happened twice, with no result for her. I must be super vigilant on that because she is a breed renown for barking.
This is a special chapter, and we’ve done pretty good so far.
So I will look for a ” tutorial ” to teach the bell (I suppose similar to a paw or nose touch), it will give me something to teach Inouk, and keep her busy learning.
I am sorry to write so long replies, you have far better to do with your research, articles, and Clara to look after. I hope she is well now.
Warmest regards
Hi Josephine,
It sounds like you are figuring out a good plan. I have a link, which I think you may have found, to a good tutorial about bell ringing. It’s in my own post about the problems to avoid when teaching that behavior.
It is always good to hear from you; I love your long replies! Take care!
I believe in positive training and never being harsh with a dog but I have a question, How do you handle a dog who shows aggression to people or children ?
Lori, rather than doing operant conditioning (dog gets treated for desirable behaviors and less desirable ones are punished or allowed to go extinct), with aggression problems the goal is to change the dog’s emotional response. Most aggression is fear-related and aggressive behaviors do get reinforced because they create distance. The dog threatens another dog, the dog runs away. The dog threatens a human, the dog is removed from the situation. But rather than try to train away these behaviors, we change the dog’s emotional response so they no longer are driven to perform them. This is done through a process called desensitization/counterconditioning. (Something that virtually never works is “obedience” training. It does not address the dog’s discomfort or fear.)
Please find a force free trainer who is experienced with these types of behavior problems or a vet behaviorist. In the meantime, you can read up on the methods on the CARE for Reactive Dogs site: http://CAREforreactivedogs.com . But again, if you have a dog that is aggressing, this is not a do it yourself project. Please get some help. One place to look is on the Pet Professional Guild Fetch a Professional page. Good luck!
This is why I feel sick when shelter dogs are killed for aggression: ‘But my very next realization was about ethics. How could it be fair for me to punish a behavior if it was there because it was getting reinforced, or worse, because I myself had reinforced it?’ Owners often inadvertently reinforce behaviors they don’t like, relinquish the dog to a ‘shelter’ because of those same behaviors, and often the dog is then killed because of the behaviors reinforced by the owner and sometimes by others who have handled the dog. It’s horrible. Imagine someone beating a child and then killing him/her because she/he cried. This, too, happens, of course. It’s all sickening. Sorry to be so gloomy. Must be the weather.
Yes, it does happen. Thanks for the reminder of the seriousness of this issue of human tunnel vision.
This is a perfectly articulate expression of a similar revelation, which hit me like a ton of bricks, that I had when I read an article by Suzanne Clothier in which she said that the dog’s behavior is the dog’s best guess (for what you want, or what will work in that situation). And it was precisely the same for me–how could it be ethical to punish my dog when she’s just trying her hardest to make sense of a world of human nonsense?
As for rewarding undesirable behaviors, if my dog finds laughter reinforcing, then I have inadvertently reinforced quite a bit of harmless mischief. Especially her propensity to rearrange (but not chew on) my shoes. 🙂