Victoria Stilwell Isn't “Purely Positive” — And She Told Us Herself
Like dogs behaving very badly Id never watched "Its me or the dog" or the most resent version "bad dog academy" so in the interest of fairness I thought Id better try and watch a couple of them.
One thing very quicky stood out to me. Victoria Stilwell isn't a force free purely positive trainer.
So why did I, and probably quite a lot of other people, think she was?
Apparently, I'm in my villain era.
I've started asking awkward questions about dog training.
I've been reading what dog trainers actually say about their own methods.
And today I've got a confession.
I was wrong about Victoria Stilwell.
For quite a long time, if you'd asked me to name a “purely positive” dog trainer, Victoria Stilwell would have been one of the first names I'd have given you.
And I suspect I'm not remotely unusual in that.
In fact, I'd wager that if you stopped a hundred ordinary dog owners in the street and asked them to name a positive dog trainer, Victoria would be somewhere near the top of the list.
And here's the awkward bit.
Victoria Stilwell herself has been telling us for years that she isn't a purely positive trainer.
So apparently I wasn't paying attention.
And that got me thinking.
Why did so many of us think she was?
Victoria Stilwell = positive dog training
For millions of people, Victoria Stilwell became one of the most recognisable public faces of reward-based dog training.
Through "It's Me or the Dog" and her wider work, she stood in stark contrast to what was seen at the time through people like Cesar Millan.
No hanging dogs from collars.
No “alpha rolling”.
No smacking the dog because it had been naughty.
No “show them who's boss”.
Instead:
Food.
Praise.
Toys.
Teaching.
Communication.
Understanding the dog.
And, importantly, a rejection of training methods based around pain, fear and intimidation.
Compared with some of the methods that were still commonplace when Victoria first became famous, that's a pretty significant difference.
So it's incredibly easy to make a mental shortcut:
Victoria Stilwell → positive training → positive reinforcement → purely positive.
Except those things aren't actually interchangeable.
And Victoria has explicitly said so.
“I'm not and never have been a purely positive trainer.” Victoria Stilwell
In March 2017, Victoria published an article with an unusually blunt title:
"Why I'm Not (And Never Have Been) A Purely Positive Trainer"
This wasn't somebody else trying to catch her out.
It wasn't an angry Facebook post from a rival trainer.
It was Victoria explaining her own philosophy.
She explains that she had never described herself as “purely positive” and that, when she started Positively, she struggled to find a label that accurately represented what she and other trainers were actually doing.
She mentions terms such as force-free, reward-based, positive reinforcement, fear-free and purely positive, but argues that each had problems or inconsistencies.
So she created her own term:
"Positive Training"
And there's an important little detail here.
Victoria explicitly says that “Positive Training” isn't an official scientific term.
She coined it herself to describe the philosophy she was trying to communicate.
That's worth knowing.
Because when you hear “Positive Training”, you might reasonably assume you're hearing a scientific classification.
You're not.
You're hearing **Victoria Stilwell's own label for her particular approach.
So what does Victoria actually mean by “Positive Training”?
This is where things get really interesting.
Victoria's definition includes:
* positive reinforcement;
* negative punishment;
* avoiding intimidation, physical punishment and fear;
* considering the dog's perspective;
* and rejecting the idea that dominance is the answer to unwanted behaviour.
So let's deal with the terminology.
Because this is where the entire dog-training world seems to have collectively decided that words should be deliberately confusing.
Positive doesn't mean “nice”
When trainers talk about the four quadrants of operant conditioning, “positive” doesn't mean good.
“Negative” doesn't mean bad.
They describe whether something is added or removed.
And reinforcement or punishment describes what happens to the behaviour.
So:
Positive reinforcement
Something is added and the behaviour becomes more likely.
Dog sits.
You give the dog food.
The dog becomes more likely to sit.
Negative punishment
Something is removed and the behaviour becomes less likely.
Dog jumps up.
The attention they want disappears.
Jumping becomes less likely.
Neither of those requires pain.
Neither requires intimidation.
Neither requires frightening the dog.
And Victoria openly says that her approach includes both positive reinforcement and negative punishment.
So when Victoria says she isn't “purely positive”, she's not revealing some horrible secret.
She's telling us exactly what she does.
And then it gets even more interesting
In a 2020 article about aversives, Victoria makes another distinction that I think is incredibly important.
She says an aversive is something unpleasant that causes avoidance or suppresses behaviour.
But she also points out that "what is aversive can vary between individuals".
Something one dog finds unpleasant might be completely insignificant to another.
And she makes the point that even a trainer who considers themselves positive can encounter something the dog finds aversive.
She gives a very simple example:
Leaving a dog home alone might be unpleasant for a dog who is distressed by separation.
The person leaving didn't intend to punish the dog.
But from the dog's perspective, something unpleasant has happened.
This is a really important distinction.
Because it means Victoria's position isn't:
“Nothing unpleasant ever happens to a dog trained positively.”
It's closer to:
“We shouldn't deliberately use pain, fear or intimidation as training tools.”
That's a considerably more nuanced position.
And, interestingly, Victoria herself says:
“The learner decides what is rewarding and what is aversive.”
Which brings us rather neatly to the harness.
The no-pull harness
Now, this is where my villain brain started getting annoying.
Victoria has promoted her own "Positively No-Pull Harness".
Her 2016 description says the harness works by using the dog's centre of gravity to redirect the dog when they pull.
The redesigned version uses two chest attachment points, allowing the lead to be attached to the side furthest from the handler. When the dog pulls, the attachment point moves towards the centre of the chest, creating what Victoria calls the “no-pull action”.
She describes this as allowing the handler to redirect the dog “effectively and painlessly.”
And the product has been marketed with the wonderfully reassuring phrase:
“No pressure, no pain, no pulling.”
And I found myself thinking:
Hang on.
How does that actually work?
Because physics is a persistent little bastard.
How can a harness stop pulling without hurting the dog?
The answer is:
Complicated.
Personally for a strong puller I don't think it can. But in the interest of fairness the following is a what I understood from the sales pitch.
"A front-clip harness doesn't have to hurt a dog to change the mechanics of pulling.
Imagine a dog pulling directly forwards.
With a traditional back attachment, the force is largely aligned with the direction the dog wants to travel.
The dog can put its weight into the harness and continue forwards.
Move the attachment point to the chest and things change.
When the dog pulls forwards, the lead applies force from the front.
Instead of allowing the dog to continue efficiently in a straight line, the harness can redirect the front end of the body.
The dog may turn.
The dog may be brought back towards the handler.
The dog may simply find that pulling isn't as mechanically effective.
None of that requires pain.
None of that requires a tightening collar.
None of that requires an electric shock.
Masterful use of the Modal Verbs to say it might but might not happen.
So I don't think it's fair to say:
“Victoria's harness works because it hurts the dog.”
We don't have evidence to support that claim and like any new product we have no idea of the long tern implications.
But there's another question worth asking.
Does that mean the harness is completely neutral?
Again:
Not necessarily.
A 2023/2025 peer-reviewed study examined three commercially available harnesses in 66 dogs and found significant differences in several measures of locomotion depending on harness design. The effects also varied between breeds and breed types. The researchers concluded that harness selection and fitting should be individualised.
In other words:
A harness can change how a dog moves.
That doesn't automatically mean:
A harness hurts the dog.
Those are two different claims.
And this distinction matters.
The newer review is particularly interesting
Before I go into this one, while better than most as its a study of different studies, its still bad science. I've said it and ill continue to say it, until its gotten right.
A 2025 review examined the available research on collars, harnesses and head collars.
It looked at 21 full-text studies and two abstracts, considering things such as movement, pressure distribution and behavioural signs of stress.
Its conclusion wasn't:
“Harnesses are bad.”
Nor was it:
“Harnesses are completely neutral.”
Instead, it found that for dogs who pull, non-tightening front-clip harnesses appear to offer one of the better compromises between reducing pulling and discomfort.
It also found that different harness designs affect dogs differently and that tightening harnesses and head collars can present greater discomfort concerns.
That's a much more useful conclusion.
The equipment matters.
The design matters.
The fit matters.
The individual dog matters.
And perhaps most importantly:
“Harness” isn't a single piece of equipment.
A correctly fitted Y-shaped harness and a restrictive no-pull design aren't necessarily doing the same thing to a dog's body.
So I'm not calling Victoria's harness cruel
I want to be really clear about that.
I'm not.
There isn't sufficient evidence for me to make that claim.
And frankly, I don't think that would be particularly interesting anyway.
The interesting question is:
If a piece of equipment changes what happens when the dog pulls, how should we describe that effect?
Is it management?
Yes, it can be.
Is it training?
Potentially.
Does it alter the consequences of pulling?
Yes.
Does that automatically make it aversive?
No. it depends on the dog.
Does calling it “management” automatically prove it isn't aversive to an individual dog?
Also no.
And this is where Victoria's own philosophy becomes useful again.
She says the learner determines what is rewarding and what is aversive.
So if a particular dog shows clear signs that a piece of equipment is uncomfortable, stressful or unpleasant, the fact that the manufacturer calls it “comfortable” doesn't magically override what the dog is telling us.
The dog gets a vote.
“Management” isn't a magic word
This is something I think we could all be better at.
We sometimes describe something as:
“It's just management.”
As though that settles the argument.
But management can mean lots of things.
A baby gate is management.
Changing the environment is management.
Increasing distance from a trigger is management.
A muzzle can be management.
A long line can be management.
A harness can be management.
But these things don't all work in the same way.
So rather than asking:
“Is it management?”
I'd rather ask:
“What does the management actually do?”
If a dog pulls and a harness redirects them, something has happened as a consequence of pulling.
That doesn't automatically mean pain.
It doesn't automatically mean punishment.
But calling it management doesn't make the physical or behavioural consequence disappear either.
That's why I think we should be cautious about absolute claims such as:
“No pressure.”
There is obviously a physical force involved when a lead attached to a harness redirects a moving dog.
The useful questions are:
How much force?
Where is it applied?
How is it distributed?
How does the equipment affect movement?
How does this individual dog respond to it?
Those are much more useful questions than simply asking whether something has a “positive” label attached to it.
And this is the bigger problem with dog-training labels
Consider the words we throw around:
Positive.
Positive reinforcement.
Reward-based.
Force-free.
Fear-free.
Purely positive.
Balanced.
They can sound like neat little boxes.
But they aren't.
A trainer can use positive or negative reinforcement without being reinforcement-only.
A trainer can use negative or positive punishment without using pain.
A trainer can use management without using an aversive tool.
A piece of equipment can alter behaviour without being designed to cause pain? In my mind the jurys still out on that one but...
....something a trainer believes is neutral can still be unpleasant for a particular dog.
The label doesn't tell you everything.
The details do.
So why did so many of us think Victoria was purely positive?
I think there are several reasons.
She became the face of reward-based training.
For a huge television audience, she represented the alternative to traditional punishment-based dog training.
The word “positive” sounds like “good”.
Nobody hears “positive training” and immediately thinks about the technical distinction between positive reinforcement and negative punishment.
Television simplified the methodology.
Viewers saw rewards, teaching and gentle handling.
They didn't necessarily see the underlying behavioural terminology.
The wider dog-training world uses the labels inconsistently.
“Positive”, “positive reinforcement”, “force-free” and “purely positive” are often used as though they mean the same thing.
They don't.
And then there's the biggest reason of all:
Victoria herself was very clear — but most of us didn't read the not very small print.
She published an entire article in 2017 asking people to stop calling her purely positive.
Nearly a decade later, many people still do.
Including, for quite a while, me.
Victoria's position today is even clearer
In May 2025, Victoria published a statement reaffirming the position of Positively and the Victoria Stilwell Academy.
She says they reject tools, methods and practices that have the intended or "unintended" consequence of causing pain, fear or purposeful discomfort.
She specifically includes e-collars, prong collars, choke collars, harsh lead corrections, hitting, pinching and similar techniques.
So this isn't a story about discovering that Victoria has secretly become a “balanced” trainer.
The evidence doesn't support that.
If anything, her recent public statement makes her position against aversive tools and methods even more explicit.
And that's worth saying.
Because if we're going to criticise trainers for misrepresenting themselves, we should probably avoid doing exactly the same thing to them.
I'm not here to cancel Victoria Stilwell
Sorry to disappoint.
Victoria Stilwell has played an important role in moving mainstream dog training away from methods based around intimidation, pain and dominance.
I don't think that should be ignored.
And I don't think her acknowledgement of negative punishment somehow makes her a “balanced trainer” in the way that term is commonly used.
That would simply be replacing one inaccurate label with another.
What I think is much more interesting is this:
Victoria has spent years telling us that her "Positivity” approch doesn't mean what many of us assumed it meant.
And perhaps we should listen.
The real villain might be the label
Maybe Victoria Stilwell isn't the villain of this story.
Maybe I am.
Or perhaps the real villain is the ridiculously confusing language we've created around dog training.
Because the public conversation often reduces complicated behavioural science to:
Positive = good.
Balanced = bad.
Or:
Force-free = never say no.
Or:
Positive = never remove anything.
Or:
Management = doesn't count as training.
None of that is particularly helpful.
Dogs don't care what tribe their trainer belongs to.
They don't care whether the trainer's Instagram bio says “positive”, “balanced”, “force-free” or “science-based”.
They care about what happens to them.
They learn from consequences.
They respond to their environment.
And, annoyingly for all of us who like neat little categories:
They are individuals.
So here's my challenge
The next time you see a viral dog trainer online, don't immediately decide whether they're one of the good guys or the bad guys.
Don't take the label as the answer.
Look underneath it.
What happens when the dog gets it right?
What happens when the dog gets it wrong?
What happens when the reward isn't available?
What equipment is being used?
What does that equipment actually do?
What is being added?
What is being removed?
What does the dog appear to find rewarding?
What does the dog appear to find unpleasant?
And perhaps the most important question:
What does the trainer actually mean by the label they're using?
Because two trainers can use the word “positive” while doing quite different things.
And two trainers can disagree enormously about methodology while both genuinely believing they're being humane.
And here's the embarrassing bit
I used to think Victoria Stilwell was purely positive.
Not because I'd carefully studied her methodology and concluded that she used reinforcement exclusively.
I thought it because:
That's what I'd understood “Positivity” to mean.
And that's probably true of a lot of dog owners.
We hear:
"Positive"
and think:
"Kind."
We hear:
"Force-free"
and think:
"Nothing physical ever happens."
We hear:
"Reward-based"
and think:
"Nothing is ever removed or withheld."
But none of those assumptions necessarily follows.
And if we're going to have serious conversations about dog training, we need to stop pretending that the labels tell us everything.
Sometimes you have to believe the person you're investigating
And this is perhaps the strangest part of the whole thing.
I went looking for evidence that might challenge what I thought I knew about Victoria Stilwell.
And what I found wasn't:
"Aha! She's secretly doing something completely different.”
I found something much less exciting.
And much more useful.
She actually told me what she does.
She told me in 2017 that she wasn't purely positive.
She explained what she meant by Positive Training.
She explained that it includes positive reinforcement and negative punishment.
She explained her position on aversives.
She explained how she believes the learner determines what is rewarding or aversive.
She explained how her no-pull harness is supposed to work.
And her organisation's 2025 statement makes her opposition to pain, fear and purposeful discomfort even more explicit.
The problem wasn't necessarily that Victoria was hiding the truth.
"The problem was that I — and probably plenty of other people — had put a different meaning on the label she gave herself."
So perhaps the lesson isn't “Don't trust Victoria Stilwell”
Perhaps it's:
"Don't trust a label."
Ask what the trainer actually does.
Ask what happens when the dog gets it right.
Ask what happens when the dog gets it wrong.
Ask what the equipment actually does.
Ask what is being added.
Ask what is being removed.
Watch the dog.
And occasionally do something absolutely radical:
"Read what the trainer actually says about their own methods."
Because sometimes the biggest revelation isn't discovering that your hero has been lying to you.
It's discovering that you never actually listened to what they said in the first place.
---
Now, if you'll excuse me...
Apparently I'm the bad guy again.
I've questioned a dog-training label.
I've asked what a harness actually does.
I've mentioned behavioural consequences.
I've suggested that “positive” and “purely positive” might not mean the same thing.
Honestly, I'm surprised I haven't been stripped of my dog-training villain licence already.
I'll be over here.
Giving dogs treats.
Taking notes.
Asking awkward questions.
And waiting patiently for somebody on Facebook to explain to me that I'm doing dog training wrong.
Cue unnecessarily dramatic villain music.
Cavalli & Protopopova, “Review of Collars, Harnesses, and Head Collars for Walking Dogs”https://pmc.ncbi.nlm.nih.gov/articles/PMC12345489/
Comments