21/08/2026
Punishment: what does it actually teach?
I’ve been thinking about punishment this week.
Does punishment actually teach a dog anything other than “don’t do that or something unpleasant happens”?
For me, that’s where things get interesting.
Punishment can suppress a behaviour. It can teach avoidance. It can change what a dog chooses to do in a particular situation.
But it’s doesn’t automatically teach the dog what we want them to do instead. And that’s particularly important with reactivity.
Once a dog is already reacting, we have potentially missed the point at which we could have effectively taught the alternative behaviour. Learning a new response becomes much harder once the dog is over threshold.
So the work I want to do is before that point. My training starts at home. Then the garden. Then the street. Then the park. We gradually increase the difficulty while continually reinforcing the behaviour we actually want.
And we don’t stop practising just because the dog is doing better, we build the relationship. We build enjoyment in being around each other. Most importantly, we build value in trusting the handler.
Sometimes that means avoiding certain situations or environments for the first month or two while we work. That seems like failure but it’s not, it is simply management while we’re changing the behaviour and setting the dog up to succeed.
And this is where the owner’s contribution becomes absolutely massive.
Because I can give someone the best training plan in the world, but if they come home from work exhausted, can’t be bothered practising, are constantly busy or only want to work with the dog when they’re already frustrated, we’re probably not getting very far.
Then there’s the question of control.
Does punishment have a place in dog training?
Honestly, I don’t know.
What I do know is that having control in the moment can be incredibly important, particularly when we’re dealing with a physically strong dog and an owner who simply cannot safely manage them. That’s where I’d first look at management and equipment that gives the handler more control and confidence.
For example, an anti pull harness can be a fantastic tool for someone struggling to physically manage a strong dog during those difficult moment. It doesn’t necessarily change the underlying emotional response, but if it gives the owner enough control to prevent rehearsal of the behaviour and enough confidence to keep working, that’s valuable.
Then we get the dogs that aren’t merely reacting. They are actively looking for the smoke.
The harness is still useful, but perhaps it doesn’t give us everything we need to keep everybody safe before that reaction happens.
So do we use equipment that can deliver an aversive consequence?
A prong, a grot, an e-collar? I’ve used a grot and an e-collar in the past.
And I’ll be honest: They work.
But “it works” isn’t the end of the conversation for me, how often do we actually see these tools being phased out?
Social media would have us believe that a dog can go from being completely out of control to perfectly behaved and somehow the tools magically disappear.
Sometimes they probably do.
Sometimes they don’t.
And there’s an important distinction between having a tool available and actually using it.
With something like a prong or grot collar, the opportunity for physical pressure is always there whenever the lead is used. An e-collar gives you considerably more control over when the aversive is actually delivered.
But even then, the question shouldn’t simply be:
“Does this tool work?”
It should be:
“Why am I using it, what am I trying to achieve, and what am I teaching alongside it?” Because tools aren’t the only things that can be aversive.
My voice can be aversive.
A loud clap can be aversive.
An “eh-eh” can be aversive.
Even my facial expression can be aversive to a particular dog depending on the dog, the situation and the relationship between us.
That’s just life with another living creature.
The important thing, for me, is understanding what we’re actually doing rather than pretending that putting a particular label on ourselves makes every interaction automatically positive or automatically harmful.
I’m not interested in pretending punishment doesn’t work. I’m also not interested in pretending that because something works, we’ve automatically demonstrated that it is the best way to train that dog.
Control matters. Management matters. Learning matters. The relationship matters. Welfare matters.
And sometimes the hardest part of dog training isn’t deciding which tool to use.
It’s knowing when not to use one at all.
