When A “Positive” Trainer Goes to the Dark Side

Nothing sparks greater internet debates amongst dog people than training methods – are you force free, are you purely positive, do you use aversives, are you pack theory or dominance theory? We have a label for every type of trainer imaginable, and all of these labels carry a lot of emotion.

I was judgmental and emotional when I got into dog training – I went to a few seminars, took some courses, learned about learning theory, trained a few dogs, and I thought I could change the world with “positive” training methods. I could “fix” everything – nobody could possibly need a pinch collar or an e-collar? They cause pain! Who wants to hurt his/her dog?

Far too many years have passed by, and my experience has grown, along with my education. I have handled literally thousands of dogs, from feral to pet dogs to working dogs to extremely dangerous dogs. And while the majority of my training was and still is “positive”, I learned that there is a place for those two quadrants of operant conditioning that should never be named: positive punishment and negative reinforcement.

Let’s clarify what these two quadrants mean. Positive punishment means that I am adding an unpleasant consequence to a behaviour in order to decrease the occurrence/intensity of that behaviour. In negative reinforcement, I am removing an unpleasant stimulus in order to strengthen a behaviour. These two quadrants of learning certainly work, but they come with a lengthy list of rules that include precision of timing, consistency, strength of the aversive stimulus, proper association of the aversive and more. Sounds nasty and difficult, right?

So what prompted me to go to the dark side? A fast car that nearly collided with my beloved and very fast dog.   It was an early spring morning, when the temperature and moisture in the air is perfect for carrying scent. We had stepped out for a morning walk, and Mulder, who is confident and well trained, put his nose in the air, got onto a scent and sped off. I recalled him – he had years of good recall, all trained with fantastic rewards. He slowed to look at me, essentially gave me the paw, and sped off in pursuit of delectable roadkill. At the same time, a vehicle was speeding down my road and as that vehicle hit the curve of the road, Mulder was crossing the pavement at the same time, while I was running and screaming (helpful, I know). I was calculating the time it would take me to drive to the closest emergency clinic, which one I should go to, and what would be the consequences to Mulder being hit by a car – would he die? Could he still work? I played worse case scenario in my head.

Luckily, the car saw Mulder and braked. Mulder visited the road kill that was more exciting than myself, and then he returned to the proper side of the road. I promptly attached his leash, went inside, drank a coffee and got out an e-collar. Yes, a shock collar – call it what you want, but I was not going to lose my beloved dog to a car, as long as I could help it.

After my coffee, I went out with the dogs, established what level of stimulation Mulder responded to (starting at zero, of course, and incrementally working my way up), and he learned that if I recalled him, and he did not come, he would experience low-level stimulation, and he could turn that off by coming to me. It did not take more than 3 repetitions, and he was reliably recalling without me having to use the stimulation. I gradually added in distractions, and my problem was significantly reduced.

I felt relieved. I later discussed the training with a colleague of mine, and she asked me how I felt ethically about my decision and training – I had caused my dog pain, on purpose. I said I felt fine because I had potentially saved my dog’s life, and it was unethical for me to not use my knowledge and skills. The pain of a permanent injury or death far outweighed the pain of a shock, and he was still being rewarded for the desired behavior – it only took 3 shocks for Mulder to understand that not coming when called carried a consequence. We did his training out in the pasture – he is not afraid of the pasture, nor myself, nor putting on the collar.

I am going to be judged on this blog post – I might as well be tied to the stake and burnt because I use an e-collar on my dog, following the least intrusive, minimally aversive protocol and the science and application of learning theory. Judge all you want, but my dog still loves working with me, and I am able to continue to enjoy working with him.

Posted in Uncategorized.

36 Comments

  1. Sarah, this is a refreshing article. Balanced tools and training methods are our best chance of having good outcomes and well trained dogs. Thanks for writing this.
    Ray

  2. Love this… As one prong/e-collar using positive trainer, well said. I am glad someone else feels like I do. I use whatever tool I need to to mitigate/establish behavior. My dogs do not cower, they love to work and play and they love me.

  3. I had to do the same thing to save Nutmeg’s life. Her recall usually is very good, until she ignored my ” Nutmeg come” and chased the deer out onto the road and was nearly hit by a car. She bought herself an e collar, and it has worked extremely well. I tried it on myself before I put it on my dog. Doesn’t cause major pain at the low end (2). But does get her attention. Not all positive training works for certain dogs.

    • It’s a terrible feeling, to have your dog run out onto the road. I have a few rural clients with happy, confident dogs that I have helped with an e-collar to keep them off the road. I would rather show them how to use it properly than to deny them the use of it because it is “too aversive.”

  4. Dog training is like politics, you have two POLAR groups on opposite sides. Both sides have SOMETHING to offer, but their offerings are incomplete and have significant limitations. They spend most of their time propagandizing the wonders of their side as well as the horrors of their enemy. Meanwhile, what is truly best for the dog, what is supported by both real world practicality in combination with science, is left in the middle, overlooked. I started off Kohler, yank and crank and then over the years learned the science of behavior and picked up a clicker. I now find myself in the middle sitting silent as the idiots on both sides simply yell and scream at each other in tornadoes of emotion, bias, and subjectivity. Yankers have dogs who live day to day subdued, stressed, and in conflict……Butterfly trainers have exuberant and NAIVE dogs who are all one competing motivator away from a truck bumper…..Thank you for this article….its always nice to know there are others using their brains and logic 😉 Good for you and thank you for sharing!

  5. You could just use a leash around roads, so it wasn’t necessarily the least intrusive, minimally aversive choise.
    IMO if you wouldn’t put a shock collar on a child, you shouldn’t put in on a dog.
    Anyhow, a CO trainer will defiantly know how to use punishment MUCH more ethically and efficiently than a traditional trainer.

    • There is no place where I live that is not within sight of a road, which means, by your logic, that my dog is being punished for the occasional dropped recall by only getting what exercise she can at the end of a 6′ leash attached to a trainer with arthritic knees. I will never understand why life at the end of a short leash is better than a few well timed, well considered corrections. Just how long must a dog be “punished” in force-free training?

    • I wouldn’t crate a child, make a child eat from a bowl on the floor, ask them to pee/poop outside, castrate or spay a child, allow a stranger to bathe a child in a store behind a glass window so other strangers can watch, ask a child to go lay down while they watch me eat, allow a child to lick me on the face/hands, feed a child treats from my pocket for doing what I ask, etc…etc…etc…
      Dogs are not human.
      Children are not dogs.

    • A leash is an aversive, and it’s up to the dog to say whether it or the occasional use of an e-collar is least intrusive, minimally aversive. Try both methods (after having someone skilled with e-collars teach you how!) and observe the dog to see which it prefers. My dog’s joie de vivre increased exponentially once e-collar training enabled her to be regularly off-leash.

    • Oh jeez, the shock on a child argument. Not even remotely (ha) the same. I would say it’s intrusive to restrict a dog’s freedom simply because you don’t want to provide a few aversives for something. I’d say if you asked the dog, they would say they would prefer freedom. Every time.

  6. Thank you!! I am also a crossover back to the ‘dark side’…LOL I have always called it that. Cuz that’s sorta how it gets viewed, ruling with an iron fist, allowing no choice ever. I got good results teaching a bunch of stuff, and even getting great recalls, impulse control, etc. Except when I didn’t. Like that. I’m amazed I stuck with it, even after my dog slipped a collar and ran after a motorcycle down a freeway on ramp. Same exact things went through my head, I screamed til I lost my voice. High drive dogs are easy to motivate, certainly, but they also benefit from having some “doors shut”, and some options limited. Not having tools and techniques to do that, left education lacking. I also found that a trainer has to utilize a lot of control, in order to avoid aversives, which turns me off. That’s kind of why “force free” appealed to me, that the dog exercises self control, vs being controlled…But I’ve realized the tools help that, help the dog, with clarity. It’s farrrrrr less”positive” to get frustrated, and be in conflict.

  7. For many there is no reward that is of higher value than the bird, rabbit, coon, or the chase or herd that we have bred into their genes so in the right situation the genes will win. All we can do is use all our tools hope it is enough.

  8. You did, as you always Do, the right thing for Mulder! The world of dog training is much to over opinionated with both sides being ‘right’s in their views. But at the end of the day you used a tool that was needed for Mulder and you know what is best for him!
    Your honesty, knowledge and experience is always informative and refreshing to read/hear!
    Thank you Sarah for another excellent blog!

  9. Good, short, clear post.
    I get it. You used LIMA. You had put the basics on Mulder and used the collar for a very specific and important behaviour. I am a positive trainer, as you know, and I don’t know if I will ever feel the need to use the tools at the other end of the spectrum, but I understand why some do.

    Knowing the dark side doesn’t mean you live there or choose to visit as a matter of course. I will though, respectfully disagree with the poster that says balanced methods are the best chance for the good outcomes. I think that generalization takes away from the skill of the trainer and the dog in front of you. Train the dog in front of you.

  10. Our dog training industry is going through a serious growth pain, and so all the drama is predictable and inevitable. Moving our industry forward, requires dog trainers like yourself to push against the norm, push outside what is considered socially safe, holy and devote within a limited set of perameters.

    In the end, free market capitalism will favor techniques that meet consumer demand. Demand (as it relates to your dog) means ensuring your dog avoids a premature death from a moving vehicle. Thank you for sharing your experience, Sarah.

  11. Why do people call this “punishment” it’s training? Plainly said. Sometimes I wished I had used 1 on my son.

  12. People who will burn you for using methods that they themselves would not use; don’t have your dog; and aren’t in your situation. I say do whatever works for you and your dog. I’m thinking of using an e-collar to stop my dogs from eating stuff they come across on the farm. Their “leave-it” used to be excellent… but not anymore it’s not.

  13. I have been training for over 40 years. Seen a lot of trainers, a lot of methods rise and fall in popularity. I train dogs and horses “positively” But just like in raising good children… there is a time and a place for discipline as well.
    The thing I found most interesting in your blog is that you are a “positive” trainer and this was your “turning” point to realize that there was a time and a place to show the dog that there is a consequence for not complying with your instructions, you went in the house had a coffee and strapped on the E Collar to go out and work on this. UMMMM…. honey, if you already owned the collar??? You had already given this some thought, I’d say.

    • Yes, we had had an incident with the road a couple of months prior, and I had borrowed a collar. Things had gotten better, no more issues, but this one was far too close. I was glad that collar was on my kitchen table.

  14. I am not opposed to the use of e-collars if used correctly and under the correct circumstances. I would never use an e-collar for dog aggression because such behaviors can be made worse by using the e-collar. I have seen it with and electric fence I had around my poultry and a foster dog who really wanted my chickens. The dog hit the electric fence trying to knock the fence over and was shocked and the other dog right next to her ended up getting attacked and injured because the foster associated the other dog with their pain.

    I am also against using e-collars if you don’t have a good understanding of your dog. I currently have a dog that was seriously damaged by someone using an e-collar on him to train him to herd sheep because it’s quicker than using other methods and they wanted results. This dog is such a soft dog that a word from me in a quiet voice and he will pee himself. He is now a wreck due to some idiots use of an e-collar on a dog that never should have been shocked in the first place. He would have been such an easy dog to train and now we are undergoing months and years of threshold training to get to the point where he’s not petrified of the whole world, listening to commands and simple things like moving from one room to another.

    I guess what I am trying to say is if you have a tough dog who needs a shock collar because they don’t listen any other way and you are experienced with how to use it properly that’s fine but it should be done as a last resort when other things have been tried and failed, and you should know your dog well enough to know that using an e-collar is right for him or her. Don’t ruin your dog for the sake of training quick rather than well.

  15. Good for you! Your dog’s life is too high a price to pay for commitment to an ineffective training ideology.

    • I would alter this a bit to say “Good for you! Your dog’s life is too high a price to pay for commitment to one specific training ideology.” Its not that LIMA is ineffective as a general rule, just that occasionally its not the best choice. I went from Jerk and Pop training to a zealous clicker and cookie convert. Then I started teaching my BC how to work stock (tough to cookie push in that situation), learned that cookies are a poor substitute for relationship and engagement (cookies should be paired with play and human engagement) and that one-size-fits-all seldom works. Nowadays, I recognize that not every dog and situation is the same, and that there are a lot of tools that work. I now work from the LIMA perspective but recognize that in some cases, a single strong aversive can be a good and effective choice for certain situations (assuming training has happened). I see dogs in my pet classes with fallout from bad e-collar training, and there are some dogs I would never use an e-collar on, but I also know many for whom an e-collar allows safe off lead activity. Even if its not something I would personally choose in a situation, I don’t think its helpful to be judgemental and cliquish about it.

  16. I applaud you for having the courage to write about your experience with all the sad judgmental people fuming about. Science based dogs training is just that, and not all in life is simple. My favorite quote “There are in nature neither rewards nor punishment – there are Consequences.”
    Robert G Ingersoll 1881

  17. Thank you for being brave enough to post this. We both know you’re likely to catch a bit of hell from the Purely Positive/Force Free camps.

    I do not understand why people have such a hard time realizing that the ultimate aversive lesson that a dog can learn is death. Any properly used tool that will prevent a dog from becoming road-kill, or being turned into a kill shelter is much preferable to the alternative. Modern Balanced Training, aka “the Dark Side,” is largely R+… rewards and praise. Tools such as an e-collar or prong collar are only used once the dog clearly understands the skill and then to build reliability. It is a system that works extremely well.

    Good luck with your future training endeavors.

  18. I don’t agree with walking a dog anywhere close to a road, without having it leashed properly. I don’t care how great the owner thinks their recall is with their dog as I’ve seen an extremely well behaved dog get hit…thankfully, she only suffered minor injuries. Putting your dog at risk and then ‘punishing’ it with the e-collar due to your own failure in protecting your dog in the first place…I don’t agree with that. I’m not promoting one type of training over the other…to each their own…but if you want to let your dog be leash free, go to a leash free park or take him/her on a hike on a trail. Besides being against the law to have your dog walk/run leash free, there can be issues that arise when coming across dogs that are on leash.

    • We maintain 100-200 feet distance from the road (I live on 60 acres) with either a fence or water obstacle between us and the road. Unfortunately, my dog is extremely fast, and obstacles are just minor inconveniences. If we do go for a walk “up the road”, then he is always leashed.

      • Thanks for the reply Sarah. I read some of your other replies as well in regards to how your property is laid out and I can understand the option to let your dog go leash free on your own property. I just didn’t want anyone to assume that e-collars would be an option to be having their dogs leash free close to roads. As I mentioned, I was going on the experience I had with the dog that got hit…and in that case, the dog was leash free on a sidewalk by a main road, and not on a property such as yours.

  19. What you actually told us is that you were 1. Too lazy to generalize the cue properly.
    2. Too lazy to maintain the cue properly
    3. Negligent about management of your dog near a busy road.
    Not that the averse method is is the only option

    • Thank you for your feedback Alex!
      1. I am not sure about lack of generalization – I can recall my dog off of deer, a decoy that he is biting, and in new environments. This morning, on our usual morning jaunt on the farm, it just didn’t work.
      2. I always carry food on me, as well as a ball. Sometimes a copper pipe, which is his favourite, but this is not reasonable in winter as his mouth will freeze to the pipe.
      3. We live in the country. I maintain 1-200 feet away from the road as well as a water obstacle &/or fence between us and the road. Unfortunately, with a driven, working line Malinois, those two things are minor inconveniences. Also, it is not a busy road – there are maybe 10 cars per day on it.

      As far as laziness, I drive 2 hours each way per week to PSA training. I host protection seminars at my property. My dog does not eat out of a bowl – he earns all of his food each day with training. Prior to purchasing him, I completed a 5 day course (each day being 10-12 hours in length) at his kennel. So I guess I am pretty lazy?

      No, the aversive isn’t the only option, but when requesting 100% reliability, proofing with it, particularly with confident dogs who are not anxious/fearful/nervous and are extremely high drive and prey driven, it is. Most pet dogs do not need to perform at 100% reliability, but my dog does.

      Thank you again for your feedback!

    • Oh, I love it! The “lazy” argument, because when all else seems to make sense and the person telling the story has the experience to back it up, there is ALWAYS an out by just writing them off and telling them they didn’t work hard enough! I WONDERED when this ad hominem card would be drawn! A very hearty congratulations to Alex Millatiner!

      If this is “lazy” training, then what is this rule about keeping a dog on-leash its entire life because either a) dogs should never be off-leash because they can never ever be trusted to recall of ANYTHING more motivating than that piece of liver in your hand or b) one cannot train a dog to off-leash standards, so therefore the easiest thing to do is just never ever remove the leash?

      Damn if that isn’t lazy. Keep a leash on the dog its entire life WOOOOOO. Putting in the time, knowledge, learning and effort to actually train a dog to one-command proficiency, at liberty and around distractions? If that’s “lazy,” I have some prime swampland in Florida as well as a bridge in the Rockies in addition to my beautiful mountain-view cabin in Alaska for sale.

      Although I guess I do get it to an extent, since a dogs’-lifetime of constant environmental management, tiptoeing around issues, BAT/CAT, constant reinforcing and marking as well as the emotional strain of trying to convince other skilled, experienced trainers (some with successful businesses!) that they are somehow wrong–after all, you are the Light, the Truth, the WAY!–takes a certain amount of effort and dedication, not to mention a whole pantry full of Kool Aid. I can’t call that “lazy” at all. Cheers to your dedication, misguided though it may be. I hope you one day enjoy life with a truly well-trained dog that shows you the errors of your way and the ease/speed that could come with the actual training process, leaving you years of enjoyment with a mentally, physically, emotionally and socially-balanced companion animal that has the confidence, resilience, coping skills and self-control necessary to live in our society instead of years of stress, anxiety and frustration attempting to mold a small corner of one’s highly-managed environment to a dog that has no concept of how to behave without such restrictions.

      Carry on, though. Sarah is “lazy.” Bad Sarah. Bad!

  20. Such an interesting story, Sarah! I read it when you first posted as I was running out the door. There were so many things that jumped out at me from the get go. Now that I’m back in my chair and reading again, I feel the need to unpack those first reactions. This is me just spilling out what’s in my head — observations and questions. You know I have admired your work. And I have huge respect for your honesty and guts publishing this piece.

    Here goes… the top 10 things that made me go hmmmm…

    1. You mention your early years as a dog trainer — “I thought I could change the world”. Then you suggest that many years later you’ve learned and experienced more, and this has lead you to justify the use of shock collars. Are we to conclude that positive trainers who haven’t gone to the dark side are naive?

    2. I wonder if your title is misleading. From what you describe in your article and comments (you do e-collar training with your rural clients, you use a prong collar), you’re not a positive trainer to begin with.

    3. Do trainers really use the label “purely positive”? I don’t know any educated positive trainers who do.

    4. You say “I learned that there is a place for those two quadrants of operant conditioning that should never be named: positive punishment and negative reinforcement.” Technically, it’s impossible *not* to use all 4 quadrants. I think a better way of looking at it is that some aversives are more severe than others. The main question is whether you are causing fear, pain, or intimidation. And if you are, there were probably many missed opportunities for using positive reinforcement.

    5. Did Mulder already have a recall? You mention that he is well trained, yet go on to say it only took 3 repetitions for him to be reliably recalling without using shock. Do you mean he didn’t have a 3-time reliable recall using positive reinforcement?

    6. If roadkill is the strong attraction, not something more challenging like a pouncing deer, why not use the immobile dead animal as part of your training plan? So easy to do training set-ups with an animal that is dead! LOL Why skip many excellent training possibilities and be so quick to jump to shock collar as a solution?

    7. In your final paragraph you say that you were following LIMA. Logically, the death-is-the-alternative argument can be used to justify just about any action.

    8. You say the “problem was significantly reduced”. Then how are you convinced you are potentially saving Mulder’s life?

    9. What happens now? For example, does Mulder always wear a shock collar? Just as with the prong collar? (From what you say in the comments, you always use a prong collar when on leash)

    10. What do you think our takeaways should be? As trainers? As pet owners? Who should use shock collars (and prongs) and under what circumstances?

    • Hi Sylvia,

      Thanks for your questions – I will try and do my best to answer them.

      1. In my blog, I did not speak for any other trainers but myself. So no, we can not conclude that positive trainers are naive. But I was, and I am more than happy to say that.

      2. Most people who are familiar with me, particularly when I was in BC, would consider me a positive trainer, and certainly, when I started into the realm of dog training, I considered myself one who only needed positive reinforcement and negative punishment. As time grew on, I built confidence and skill and recognized that one can use negative reinforcement and positive punishment. Look at CAT, for example – I only use that protocol sparingly, but it works very well. I do not consider myself a positive trainer, as that limits me. I am a trainer who applies science.

      3. Yes, trainers use purely positive and market themselves as such. Without any effort, I could list off 10 trainers/behavior consultants, who are educated and consider themselves purely positive. Many of these people have the same credentials as myself.

      4. Yes, I am using something unpleasant if I am using negative reinforcement/ positive punishment. There is no way around that. Did I miss opportunities for positive reinforcement? Probably, but I am requiring 100% reliability.

      5. Yes, of course Mulder already had a recall. I do not understand what you mean by a 3 time reliable recall. He lives off leash and does obedience. He has probably been recalled over 1 million times.

      6. In Ontario, things like roadkill have the potential to carry disease, such as leptospirosis, distemper, even rabies. It would irresponsible of me to bring the potential for these diseases onto my property and expose Mulder and Shorty to them (and yes, they are fully vaccinated and on year round parasite prevention). Some of these diseases can live in the soil. I actually remove all dead animals I find on my property due to my concern for disease. Mulder has chased deer in the past, but was aggressively charged by a doe, which has limited his fun in that.

      7. Death/serious injury is the potential side effect for poor recall. I am not sure of your question here.

      8. There is always the potential for error – I have left the house having forgotten to turn on the collar or the batteries could die. I could hit the wrong button. So I have significantly reduced the problem, but nothing is without human error.

      9. In competition, Mulder is only allowed to wear a fur saver on a dead ring, so in that training, he is proofed off of all collars. When we are just walking about the farm, I am more than happy to have him wear his e-collar. He has too much time invested into him and his training for me to get caught up in ideologies and not have him wear it.

      10. It is not my place to say what the takeaway “should” be -everyone has to figure that out for themselves. It is merely a blog post about my decision to use aversive stimulation with my dog, a working line Malinois, in a behaviour that requires 100% compliance and reliability.

      There is a fantastic thread on Bill Dotson’s share of my article, in which Mike Suttle sums up the use of aversive stimulation in proofing behaviours that have to be 100% reliable. As well, if you do some research into the Department of Defense’s training of their working dogs (in which they employ veterinary behaviourists and civilian psychologists to oversee), they build all foundations using positive reinforcement based clicker training but use aversive stimulation to proof all behaviours. This is on their website. These are highly qualified people training dogs that are not pets. Now, Mulder is certainly my companion, but he is not a pet, and I ask him to do extremely difficult things every day. I do not train Shorty using an e-collar or pinch collar, because she is not suitable for it. But Mulder is.

      I hope that helps!
      I hope that helps!

  21. Thanks for taking the time to respond to all 10! I actually did think that you were a positive trainer, and that this “dark side” was a big change from your normal practice. That title really did throw me. I bombarded you with all those questions not realizing your use of a shock collar wasn’t some big event.

    I had a peek at the further reading you suggested. I notice the rationale revolves around whether or not negative reinforcement and positive punishment work (just as you used the example of CAT). But whether or not it works isn’t really what I’m questioning.

    I realize your argument is that it it not possible to have 100% reliability with Mulder using only positive reinforcement. I still wonder about that though! And I’d sure perk up if you reported on that instead 🙂

    Thanks again, Sarah, for your thorough response.

Leave a Reply