Updated: Apr 7, 2021
I consider myself a LIMA (least intrusive, minimally aversive) trainer. I know some would certainly disagree and I will be honest, my viewpoints differ from many who consider themselves LIMA.
But before we jump into the nitty-gritty, my cardinal rule and a couple example scenarios.
My cardinal rule regarding LIMA training is: We should focus on what the dog in front of us actually finds aversive, not on what we perceive all dogs should find aversive.
I'll come back to what that means in a moment.
I want to now focus on a couple different scenarios to hopefully illustrate how complicated the concept of LIMA training really is.
As a side gig, I work at a dog daycare. I drive a dog bus where I pick up up to 20 dogs and drive them to the play yard. The logistics of this are actually rather complicated. Instead of crates, each dog is tethered to a spot in the bus there are spots on the floor of the bus and spots on the raised bench. You have to juggle where dogs are in the pick up process, where they sit, who they can/should sit next to, and where they will be the quietest.
I have an adolescent German Shorthair Pointer. This boy is non-stop energy. He wants to go go goooo. He does not want to settle for an hour plus long bus ride. I was struggling with where to put him on the bus.
I tried next to his best buddy so they could play non-stop, but then they would come into the yard so amped up that their energy would start fights. This would also tend to aggravate any dogs remotely near the duo.
I tried as an island unto himself, tethered so he couldn't reach any other dogs and they couldn't reach him. This was more successful except that he would whine and bark incessantly and anxiously for the next hour.
Over time, I realized more and more dogs were starting to pick up on his anxiety and frustration. I saw more dogs panting, more dogs trying to pace, more dogs hesitant to get on the bus. I had to make a change. I could try a squirt bottle, but that's dangerous to do while driving and the odds that I would miss and correct another dog were high. I could try verbal No, but all the other dogs are going to hear that too.
So one day when one of the dogs who prefers a bench seat was gone, I tethered the GSP in the back (still away from other dogs) on the bench. And you must have thought that I was murdering him. He was terrified to get up on the bench. He hated that he had to lie down on the bench. But he was quiet and eventually actually even napped. And the other dogs immediately started to relax. Over time dogs stopped hesitating getting on the bus. And over time, the GSP learned his spot and immediately hopped up and laid down.
Was this a LIMA solution? LIMA to who? The GSP? Surely not. He hated going to that spot at first. LIMA to the other 19 dogs on the bus? Most certainly. And, here's the piece many people never consider: was it a LIMA solution for me? I have sensory processing issues. High pitched anxiety barking is one of the biggest triggers for this and I'll be honest, that GSP was setting my nerves on edge. The barking was highly aversive to me. Shouldn't my needs in dog training matter too?
We had an 18 month black lab come in while I was still at the training center. It was a bit of a tragic story, the dog's owner's mother had been diagnosed with a fatal disease shortly after bringing home their puppy and so they never had time to train their dog. After his mother passed the dog's owner approached us for help. He'd tried gentle leaders, he'd tried harnesses. Let me tell, this dogs was one of the wildest I've seen. I've seen working line Belgian Malinios that were less of a perpetual motion machine.
One of my coworkers had him on the first day. Coming out of his kennel (in which he barked, anxiously non-stop) the dog had knocked my coworker over. He'd already knocked over the person who worked in the kennel room early that day. She'd put a slip lead on him (we usually started dogs on a flat collar or slip lead to see what foundations they had) and then he proceeded to panic and struggle and when she brought out treats (low value, just kibble) he chomped her had so hard she had bruises. She put him away (with help) and asked me and another coworker to help her work the dog the next time.
So she got him out again. He was slightly less wild, but more than anything, he was tired. She tried to get his gentle leader on and he started bucking and snapping wildly. I asked her to hand off the leash to me. In two seconds I stepped on his drag leash, clasped a prong on him (one of my small super powers is getting equipment on a squirmy dog in a hurry, thank you dog daycare), and when he went to redirect on me, I gave him a hard leash bump. Harder than I usually ever would with a dog. And you know what happened? He stopped. He sat. He looked at me. And then we went for a walk back and forth in the training room and he never even tested pulling on the prong.
Was this LIMA? Most would say certainly not. But once he learned how to walk next to us (we went back to teaching foundations) we never had to do anything more than ask for "Heel" and when he got out of position a simple, calm "No" and he would immediately get back into position. If he broke a Sit to go say hi to a person or dog, we could say his name, ask for the Sit again, and look at him thoughtfully and he would immediately hurl himself back into a Sit and look at us with lazer focus.
The traditional LIMA path would be to use a low-value reinforcer. Since in this case, kibble was too much then maybe just a simple "Good Boy." You would reward for any calm behavior- blinking, less frantic tail wagging, standing with four on the ground instead of bouncing, even reinforcing slow breathing through the nose. If he started to wind himself up, you would ask for an incompatible behavior or you would stop and wait for him to calm down if he was too high to listen to an incompatible behavior.
But in the meantime, there would be times, in the beginning many of them, he would go over-threshold. He would continue to practice winding himself up, stressing himself out and everyone else around him. Yes, over time that habit would lesson and he would build new ones instead. But how is that intervening stress, frustration, and anxiety less aversive than one firm prong collar correction (because that is what the dogs' underlying emotions were: overstimulation and over arousal caused by not having alternate behavior choices available to him)?
Back to the matter at hand. What does Least Intrusive, Minimally Aversive truly look like? It's going to vary by trainer and how well that trainer/owner understands what stresses and frustrates an individual dog. It's going to vary by each person's understanding of what a dog finds intrusive and aversive. I have literally seen dogs who have never been one a prong collar before lean into the prong collar pressure in excitement. I seen plenty of dogs who found a simple collar or even a pet on the head aversive.
How we view a reinforcer or punishment doesn't matter. How a dog views a reinforcer or punishment does. If your dog flinches away every time you go to pet their head when they sit next to you, guess what? You're punishing your dog for sitting. Not rewarding them. If your dog's ears don't even twitch when you tell them "Good Dog" then you aren't reinforcing them.
"But aside from food, most dogs don't inherently find verbal or even physical praise motivating", you object. True. Food is one of the strongest primary reinforcers - think "primal" - the dogs need access to food in order to live. Things like fetch, verbal praise, and physical petting are secondary reinforcers (or conditioned reinforcers) - meaning we have to teach the dog that they are fun and desirable (via classical conditioning) before they will work for them as reinforcers.
"But you can teach a dog that a collar or a leash isn't aversive!" You object. True. Just like we can teach a dog that their e-collar means off leash hikes. And their prong collar means they get to play their favorite game of tug in the form of bite sports. And their head-collar (/halti/gentle leader) means calm, focused service work.
Tools ultimately are what we make of them. If we can teach a dog to understand how to turn leash pressure off on a regular collar without causing the dog to find it aversive, we can teach a dog to turn off the pressure of the leash on a prong collar gently and often quicker.
But why use an "aversive" tool when I don't have to?
Now that is a very, very good question. I don't start by using aversives because the dog doesn't know better. My baby puppy classes are almost entirely all positive. We introduce the word No as a no-reward marker, but that largely it. Most training dogs I get I start out with almost exclusively food and praise. I want dogs to understand the expectation, the framework of our social contract, before I tell them when they're wrong.
But what about the Lab I corrected on a prong when he had no idea what to do? That's another really good question. My one exception to starting with only positive reinforcement is in the case of dangerous behaviors. Wild thrashing and redirection (not aiming for me necessarily just aiming for the world in general) bites are dangerous. Jumping on children and the elderly is dangerous. Jumping on counters in the kitchen can be dangerous. Bolting out the door or toward the street is dangerous. I am going to correct dangerous behavior so it doesn't perpetuate. Yes, of course I'm going to manage with redirection and baby gates and crates and and and. But what if a dangerous behavior happens?
In both scenarios above, the dogs were anxious and frustrated and unable to settle. Very rarely is this addressed as an emotional state that is detrimental...perhaps even aversive...to a dog's long term mental health. Again, you can manage these scenarios so that the dogs aren't in positions to work themselves up. But let me be frank. I can do that as a dog trainer. Most owners are not equipped to be that consistent with their dogs, to be that aware of what anxious behaviors are. It is certainly a popular opinion, but in my experience using a correction, that wake-up slap you might give someone in true panic, is far kinder to the dog's long term mental health than allowing that anxiety and panic to build and perpetuate.
Something I touched on above but is rarely addressed in conversations about training style is: where do the owner's personal needs (not even wants, needs) figure into this discussion? As I mentioned, I find shrill barking incredibly aversive. Many people find begging or jumping incredibly aversive. Yes, as the being with greater capacity for understanding and the one with the dog in our charge, we can't just sit down and calmly explain to the dogs why their behavior hurts your feelings (or just plain hurts). I think when you approach the conversation from the position of: owners are never allowed to react angrily or get frustrated, that they should always approach interacting with their dog from a place of happiness, you accidentally deny owners the ability to feel the full range of human emotions. You start tone policing them and, in my experience, most people respond with sheepishness and shame. Should you stop training if you are getting frustrated? Yes. Should you cut an outing short if your dog is acting up and you're struggling to manage them? Yes. But here's the thing, that's not always an option. It's just not. And bottling your emotion and putting a happy smile on it and faking it till you make it just isn't emotionally possible for everyone. Is it kinder to give a measured leash bump to the dog to stop a behavior when nothing else is working in that scenario or allow the frustration and emotion to build? We don't expect dogs to be perfect. We expect them to make mistakes. There has to be room in the equation for dog owners to be human too. Extend them the grace to which we extend dogs. Kindness to dogs and kindness to people. That should be our goal.
I'll likely revisit this issue because there's so much to discuss. But there is so much more nuance to training and, more importantly, understanding dogs than a simple harness and head collar = good. Prong and e-collar = bad. Picture of my dogs enjoying time off leash in their e-collars.