Dogs don't learn from fear. They learn from safety.

Oct 4

8 min read

Have you ever seen how an agitated dog settles next to a particular person and, within a few minutes, breathes differently? Nobody said anything to it. Nobody told it off. It simply sat next to someone calm — and that calm passed into it.

And you've probably seen the reverse too: a panicked person at the end of the leash, gripping, pulling, shouting "stay, stay, STAY" — and a dog that becomes, second by second, more and more agitated.

Those two images say almost everything about how a dog actually learns. Not through fear. Through safety. Here I'll show you why — and why the most fashionable "shortcut" in training, the device that gives the dog a shock to "correct" it, breaks exactly the thing you're trying to build: the trust that next to you is a safe place.

Part 1 — Why it calms down next to a calm person

Let's start from something you already feel, but may not have put into words: your state passes into the dog. Not metaphorically. Literally.

Think of a small child who's been frightened. If the parent panics, shouts, gets agitated — the child cries harder. If the parent crouches down, speaks slowly and warmly, breathes deeply — the child calms down without anything being explained. The child doesn't process words. It processes the adult's state. It's what specialists call co-regulation: a calm nervous system helps another one, not yet formed or frightened, to return to balance.

The dog works the same way. It doesn't understand your speech — but it reads the tension in your hand, your tone, your breathing, your posture. To it, you are what the calm parent is to the child: a living regulator.

And it's not just me saying so. Researchers gave dogs a problem to solve and noticed something clear: the dog works more, and with more motivation, when its owner is in the room than when the owner is absent. And the detail that matters — a stranger in the owner's place doesn't produce the same effect. Not just any person calms it. Its person. In you, the dog has what psychologists call a "secure base": the point from which it sets out to face the world.

What's more, over the long term: researchers measured the stress hormone accumulated in the hair — the one that shows the load over months, not over a moment — in dozens of dog-owner pairs. The result: the dog's stress level synchronised with the owner's. The dog mirrored the person's state, not the other way round.

Put the two together and you get a huge practical conclusion: half of your dog's "training" happens in your state, not in commands. A person who starts the walk tense, afraid of "what the dog will do today", transmits exactly that fear. A calm, attentive person, in control — not rigid, but present — becomes the base from which the dog can learn anything.

What you do at home, concretely: before you correct anything in your dog, check your hand and your breathing first. Loose leash, shoulders down, breathe. Be a lighthouse of control, not a headless chicken. If you are calm, you've given your dog half the lesson without a word.

Part 2 — Why fear ruins exactly what you want to build

Now the uncomfortable but important part. There's a seductive promise on the market: a device that gives the dog a small electric shock (or a prong collar, or a hard jerk) "teaches it to listen fast". It looks effective. The dog stops on the spot. The problem is what you don't see.

Imagine learning to drive with an instructor who slaps your hand every time you make a mistake. Yes, you'll tense up and avoid that mistake. But you won't learn to drive well — you'll learn to be afraid of the car, of the instructor and of everything to do with driving. You've suppressed a behaviour, but you've damaged your relationship with learning itself. When the brain is afraid, it stops analysing — it defends itself.

That's exactly what happens in a dog. In a serious study of companion dogs, those trained with methods based on fear and pain showed more signs of stress during training — lip licking, yawning, a tense body, panting — raised stress hormone after sessions, and, most importantly, they became more "pessimistic" in a test given outside the training context. In other words, the fear didn't stay in the lesson; it seeped into the way the dog saw the world in general.

What "pessimistic" means in a dog, briefly: it was shown a bowl that sometimes has food and sometimes doesn't, placed in an ambiguous position. The dogs raised in fear went to check less often and with less enthusiasm — they expected the worse outcome. A dog raised in fear starts betting on the bad.

On the shock collar specifically, studies that compared it directly with reward found the same pattern: the dogs worked with shock spent more time tense, yawned more often, interacted less with their environment — and in terms of results, the shock was not more effective. On the contrary, the dogs worked with reward responded better to commands. And there's a link that hurts: across thousands of dogs, the use of confrontational and pain-based methods was associated with more aggression — towards the owner and towards strangers — more problems when left alone and more barking. The tool meant "to calm" was linked, in the numbers, with more problems, not fewer.

Now, honest to the end, because otherwise I wouldn't be me: there is also a recent study in which the shock collar stopped a particular behaviour faster than the gentle method. It's real, I'm not hiding it. But — and this is the whole catch — the effect didn't hold in a new place: two out of three dogs went back to the behaviour when they were moved to an unfamiliar space. That's precisely the trap of force: it suppresses quickly, but it doesn't teach. The dog didn't understand "that's not done"; it understood "when I'm wearing this thing, here, it hurts if I do it". Change the place and the effect disappears — but the bill paid in trust and in its state of mind remains.

Hence a sentence I always say: when you apply force, you take the brain out of the equation. Force turns the human-dog team into a small war fought through the leash. And I don't want to be my dog's opponent — I want to be its best ally.

Part 3 — What to do instead (not spoiling, but minimum pressure)

Here I have to be very clear, so you don't misunderstand: the alternative to fear isn't "the dog does what it wants and we smile". A dog without structure is an anxious dog — it doesn't know who's setting the direction. A dog needs a calm leader, not a tyrant, but not a doormat either.

The method I use is balanced: it has rules, it has leadership, it also has corrections — but corrections without pain and without fear, calibrated to how sensitive the dog in front of you is. The difference between a correction and punishment through fear is the difference between saying firmly "no, not like that" and scaring it to the bone. The first guides. The second breaks.

The basic principle, in three words: minimum pressure, maximum thinking. The less I intervene, the more the dog thinks, decides and acts on its own — and that's exactly when learning sticks. Concretely, it's a ladder you climb only if needed, never jumping straight to the top:

First you catch it doing well. A dog sits, lies down and walks nicely dozens of times a day on its own initiative anyway. Instead of correcting it when it gets it wrong, you calmly mark it when it gets it right: "good". Zero stress, maximum thinking. It's the most powerful level, even though it looks the gentlest.
Then you guide it — with a reward or with your voice — towards what you want. You don't force it, you show it the way.
Only if it's really needed, a minimal, gentle physical pressure that stops the second the dog yields towards the correct position. Pressure isn't punishment; it's a cue that disappears as soon as it has been understood.
The leash, as a last resort, calm, without jerks — and as briefly as possible, so you quickly go back to letting it think for itself.

Notice what is completely missing from this ladder: pain, shock, fear. Not because "we're soft", but because, as you saw above, fear ruins precisely the mechanism by which the dog learns.

Once more, in plain terms: a good teacher doesn't make you afraid of the subject to "discipline" you. They give you structure, clear expectations, and correct you calmly when you get it wrong — but they keep you safe, so you keep the courage to try. A child who is afraid of the teacher learns to avoid the teacher, not the subject. The same goes for the dog.

What you do at home, concretely:

Always start at the bottom of the ladder. Ask yourself: "do I really need a correction here, or can I catch it doing well?" Most of the time, you can.
The correction, if it comes, is calm and short — a firm "no", a stop, a minimal pressure that fades instantly. Never in anger, never through pain.
Be the secure base yourself. If you are calm, small corrections are enough; if you are panicking, no collar in the world will fix the agitation you're transmitting to it.

The one-minute summary (if you remember only one thing)

Your dog mirrors you. Your calm state is half the lesson. You are its secure base — a stranger can't take your place.
Fear doesn't teach, it only suppresses — and leaves behind a dog that is more stressed, more "pessimistic" and often more reactive. The shock collar isn't a shortcut; it's a debt your relationship pays.
The alternative isn't spoiling, it's minimum pressure: structure, calm leadership and gentle corrections, without pain. You use less force, the dog thinks more, and what it learns stays.

A final word

Why do I insist on this so much? Because I see, in the field, day after day, what each path leaves behind. Dogs "fixed" with fear obey as long as they're afraid — and give way, or explode, exactly when you need them most. Dogs built on safety and structure stay by your side even when things get hard, because next to you is the place where they have always felt safe.

I'm not selling you any device or any shortcut. I'm giving you the compass: be calm yourself first, catch it doing well, and climb towards correction only as far as strictly necessary, never through pain. In the long run, a dog doesn't obey out of fear — it follows you out of trust. And that trust is something you build, on every walk, with a soft hand and a clear head.

If you want to understand why everything you've read here works, the first step is KNESsence, the free material in which I've put the essence of the system we work by at KNES. Create an account on knes.ro, you'll get a six-digit code by email and go straight to it.

With care for your dog,

Petre ROMAN

This material is for educational purposes and does not replace a consultation with a veterinarian or a behaviour specialist. A dog with intense fear, anxiety or aggression needs an individual assessment — don't start a harsh programme on your own with a dog that is already suffering.