This New Years, I made a resolution. I am leaving a number of Facebook groups. It was triggered by a couple interesting exchanges, one that left me sobbing. I can’t say that I was sad or hurt. I have to respect someone in order to be hurt by them. I can’t say that was the case.
In hindsight, I suspect it was stress release. When you keep trying to not rock the boat, there comes a point where a straw breaks the camel’s back. Low-grade, long-term stress sucks. The relief of making up my mind and leaving just opened the floodgates.
Leaving the first group was difficult. It felt like I was shunning people.
Each subsequent group became easier. I needed less of a reason to go. I didn’t even need to feel personally threatened or offended. Seeing others bicker was enough.
You know something, it feels GOOD. Damn good. Click! You are gone. Click! No more Facebook baiting. Click! No more trolls. Click! You are banished from my little world.
Click … I have the power to leave.
It gave me an epiphany. Everyday social situations can be extremely aversive.
Some people will chime up that we should have the choice to leave abusive situations. Absolutely: Leave and stay away. This blog is not about the right to leave abuse. Although I do find it interesting that people go back to abuse. That just goes to show that returning to something does not mean it was pleasant.
Dogs should live a life free from abuse too. Let’s be really clear. This blog is not about leaving abuse. Nor is it about accidental situations where a dog becomes scared. Chalk that up to a mistake and don’t let it happen again.
This blog is about using aversives intentionally in training.
When you leave an aversive situation, you get a sense of relief. In social situations, it boils down to, “I feel good knowing that I can escape or avoid you.” That is negative reinforcement. All negative reinforcement involves aversives. That is what negative reinforcement means – the cessation of an unpleasant stimulus, increasing a particular behaviour.
There are two special forms of negative reinforcement.
Leaving an unpleasant situation is escape. With practice, you can hit the abort button soon enough, before things become unpleasant. That is avoidance. If you’re at all clever, you’ll learn how to avoid unpleasant situations entirely if you’ve had to escape them.
Tears notwithstanding, I am grateful that I felt the joy of leaving that first group. You don’t need shock collars, prong collars, choke collars or other tools to make something aversive.
I also realized just how good escape feels. Click…you are shunned. Woo hoo! Holy sh!t it feels good. I don’t have to face this anymore.
That creates quite a conundrum. Escape and avoidance legitimately feels good. If you feel good, you LOOK HAPPY once you have figured out the process. Looking happy as you leave does not mean that you didn’t feel lousy at some point in the past. The same goes for the dog.
If the dog is escaping, it could have felt lousy a few seconds ago.
If the dog has learned to avoid (escape before they feel lousy), then they might have felt lousy in the past. The “happy” dog you see is possibly avoiding an aversive that happened days, months or years ago.
Negative reinforcement is powerful. It is extremely resistant to extinction. You have to face what you fear in order to realize that you’re not actually in any actual danger. That fear, of facing what we fear, keeps us avoiding again and again and again.
Negative reinforcement is deceptive that way. If I I can avoid discomfort, I look happy, not because I have overcome a fear, but because I have become a master at avoiding fear. The embarrassing blowups might stop, but that does not mean the fear has disappeared. The fear of fear lingers and supports the avoidance.
The quadrant police, whoever they are, have it right. If you cannot trust your eyes, then trust the science. Blurring the lines of the quadrants obfuscates and hides the aversive control.
I care why my dog looks happy. In negative reinforcement your dog can look happy AND be under the control of aversives. They are not mutually exclusive. Perhaps I might not ever know for sure, but if I know which quadrant I’m using, I can make a reasonable assumption. If I can identify the behaviour, I can try to answer whether my dog is happy to approach or happy to leave.
As tempting as it might be to curl up and avoid social situations, I have decided that I dislike avoiding life as much as I dislike seeing dogs practice avoidance as a method of choice. I don’t feel truly happy escaping and avoiding a small handful of jerks because it deprives me of the smart, interesting, good, quirky people out there. So, I’ll continue reaching out.
It would break my heart to suspect that my dogs look happy because they are successfully avoiding the people and animals in it. I certainly don’t want to worry and watch for signs that my dog is recoiling from the people who love them.
Perhaps the idea that a dog can look happy and be under aversive control is a tough pill to swallow. I would rather face that uncomfortable idea. If my dog is a master of avoiding, I need to micromanage my dog, making sure no one assumes that looking happy means that the dog wants social contact. The implications of that would frankly scare me.
If you take a look at Mowrer’s two-factor theory of avoidance, some of the researchers that have done a lot of work in that area (e.g. Seligman, Weiss, Solomon) note how nonchalant animals that have mastered an avoidance behaviour can be about the imminence of something very aversive. If they can turn it off they just relax, turn it off at the last possible moment, then relax again. This nonchalance has been linked with quite significant reductions in blood plasma cortisol between the acquisition of the behaviour and the behaviour stabilising. In other words, actually, there may well be a reduction in fear with successful avoidance, and there is certainly scientific evidence to support that. But, there is also scientific evidence to suggest chronic, low level stress may still be a concern, and that repeated aversive experiences has a negative impact on welfare and emotional state. As with many things, it seems reasonable to consider avoidance training has its place in a humane behavioural modification program because mastery of stress and controllable aversive experiences and safety signals are all ways an animal can feel safer, and experience self-efficacy, and less fear. They are learned very quickly because they are highly relevant to a frightened animal. And when we have a safety behaviour, there is no need for us to be vigilant for the rest of the dog’s life. If they feel less fear, they need less safety, so engage in the avoidance behaviour less. One would hope we can be open-minded enough to realise that this is a REINFORCING process and whatever else one can say about negative reinforcement, it certainly can be associated with some neurotransmitters, brain regions, and behavioural observations that are in turn associated with a positive emotional state. If our goal is to improve welfare, we are surely remiss to dismiss that out of hand simply because it’s a quadrant we associate with fear, pain, and distress.
Negative reinforcement is very much the underpinning of some mental illnesses, where compulsive behaviours are used to avoid. R- reduces anxiety because avoidance is possible. The rituals and behaviours that support that avoidance are maintained through R-. That is why it’s so resilient to extinction.
It’s resilient to extinction when it works, because there’s no prediction error. In humans with anxiety disorders, yes, avoidance behaviours are major problems. In dogs without mental illnesses or anxiety disorders? What about dogs with anxiety disorders? The work has never really been done, although IME it’s not a huge problem. But there are a lot of dogs out there I bet you have safety behaviours that have been unintentionally trained and use them extensively while their trainers are happily telling us they are using R+ or CC/DS only.
Absolutely I think dogs have safety behaviours. Which is why I think it’s important to talk about it.
And it begs the question about whether we can create a ritual that is supported by avoidance?
I’d say it would be very easy. I just am not sure it’s wise.
Yes, we can create rituals supported by avoidance. It is indeed very easy. Wise? Depends on what they are avoiding and why IMO. I taught both my dogs social safety behaviours. I used R+, but I’m pretty sure if I stopped rewarding them with food they would persist. I REALLY like my dogs having these. It means when they genuinely are faced with another dog they SHOULD avoid, they handle it with confidence, competence and relative calmness. And, dare I say it, I usually get a happy grin. It’s a “Oh, mean dog, I know how to fix this!” grin. My reactive dog also has safety behaviours to tell me he is disturbed by something. I very much like this as well, because I’m not always super vigilant and it’s good that he can tell me well before he feels compelled to do something drastic. That is a safety behaviour I aim to see less of, though. It is not really a treatment on its own.
Great blog. I really don’t need to point out that escaping those groups didn’t make your emotions regarding those groups change for the better, did it?
And if those groups were instead a stimulus that you were afraid of, but something that you have to encounter in day to day life, just endurance of the stimulus isn’t a constructive goal. Any degree that your fear could be lessened, by pairing it with rewarding things, would be an improvement. Even simple endurance of things that you’re never going to enjoy can be had with positive reinforcement and D&CC too. But they’ve going to be tipped in the scales of being less negative via pairing them with positive outcomes.
When we have more positive options to make the learning process less stressful and more enjoyable, why resort to aversives before exhausting your supply of positive means? Isn’t that what force free and positive reinforcement training is about?
It’s a lovely thought, but it still hinges on the idea that all aversives are BAD. What if the dog looks happy, NOT because “Hey, I just escaped feeling bad” but rather “Hey, I’m smart enough to have some control over my environment, and I so appreciate my owner making those choices clear to me”? Anyone who has participated in PAR training (pressure/active release, a la Tyler Muto and Chad Mackin) knows that the dog, who was happily straining at the end of a tight leash before, will gladly choose the loose leash, if he’s shown the option of a different way. In both cases, the dog was the one making the choice. I think there’s a lot of happiness to be found in free will, no matter the source.
I thought there were some other comments here? Were they removed?
Don’t think so – not unless it broke the rules – no cursing – no spam.
‘I need to micromanage my dog.’ I don’t think many beings are truly all that ‘happy’ being controlled to the Nth degree by another being, either. Just sayin’.
I agree. I think the owner’s stress and anxiety about whether or not their dog is getting stressed most likely creates that for the dog; then, seeing that, the owner becomes MORE anxious, more micromanaging, and the dog totally flips out. Complete and perfect negative feedback loop.
connect with your dog every moment you are with them
Yes. To all of it.
Wow very strong message. I hope you know that some of us love you and appreciate you for the amazing person you are !!
Sent from my HTC
“Perhaps the idea that a dog can look happy and be under aversive control is a tough pill to swallow.”
I think this is something a lot of people feel a very strong need to deny, and I have a very strong wish for them to stop. I think this denial is a big problem, because it allows people to get away with using aversive methods to train, while convincing many that the methods aren’t actually aversive. If people believe a dog under aversive control will always look unhappy, then if a dog looks happy, the methods being used must not be aversive! Easy! Only… not.
Also, when we make something the dog’s reality, they’re going to learn to make do with what they’ve been given as the adaptive animals they are. If one aversive is less severe than another aversive they might be given, they probably ARE going to appreciate the lesser of the two.
Exactly! Also, I know dogs whose owners believe beating them (serious beatings) is an essential part of making them obey, and even THOSE dogs are quite capable of looking happy. It’s easy to make them cower just by accidentally using the wrong tone/body language etc, but they certainly don’t cower or look miserable all the time. I believe part of it is relief when there are signs (happy voices, relaxed body language, etc.) no punishment is coming, and part of it would be that they are indeed very adaptive animals. That is the life they have, and they’ve adapted to it the best they can. (Doesn’t mean those beatings are any more acceptable or less horrifying than if the dogs did just cower in a corner all day, though.) Seeing this, the belief some people have that if a dog looks happy then the training it has had must be kind and gentle and non-aversive just makes me shake my head.
(And no – I’m not saying all forms of -R are the same/as bad as beatings; of course not. I’m just using an extreme example to make a point about the “happy looking dog” thing.)
Depends on what you think “happy” looks like. If a dog is routinely punished, I would absolutely expect that to be visible in the way they behave. It’s not necessarily going to be cringing and hypervigilance, but if we know what to look for I have every expectation we will be able to detect it. If welfare hinges on emotional states, we need to get better at detecting emotional states and stop clinging to quadrants.
I agree we need to stop clinging to quadrants. A blast of water to the face could be one dog’s reinforcer and another dog’s punisher. Depends if they like that water catch game…
But when it comes to BAT, it’s the specific aversive that I don’t agree with using. I don’t perceive a need for flooding, or too close for comfort proximity to what the dog is fearful or aggressive towards.
Interesting comment – see I take the water to the face switching quadrants as a reason to cling to quadrants. That if you don’t know how the quadrants work, then you’re at risk for seeing a strategy as something that is good/bad…fits in one spot. So if you know your quadrants well, you know when you get something jumping from R+ to an aversive. Interesting comment. Just pointing out that I take it differently. I agree 100% that no technique falls into one quadrants. Up to the animal.