Every week someone tells me “I've tried everything and nothing works,” and when we walk through it, the pattern is the same: three months of interrupting the behavior and no time spent teaching an alternative.
This isn't about being soft. It's about what actually holds up under pressure, and what happens to your dog's warning system when you punish it.
What punishment does and doesn't do
Punishment, in the technical sense, is anything that makes a behavior less likely in future. Yelling, leash corrections, spray bottles, alpha rolls, shock. Sometimes it does reduce a behavior in the moment. That short-term result is exactly why people keep using it — it reinforces the human, not the dog.
The problem is what gets left behind. Behavior happens for a reason: the dog is bored, scared, excited, hungry, in pain, or has been paid for it before. Suppressing the output doesn't touch the input.
- It teaches what not to do, never what to do — so the dog keeps guessing.
- The dog often learns “this is unsafe when you're watching,” not “this is unsafe.”
- Whatever your dog was focused on when the punishment happened can become the thing they now distrust — the child, the visitor, the dog across the street.
- Escalation is built in: the intensity that worked last month stops working, so it has to go up.
Welfare and safety notes
Start with what the behavior is getting your dog
Bark For Attention
Every unwanted behavior is paying for something
Name the payoff and you'll know what the replacement has to deliver.
Access
Counter surfing, garbage raiding and begging pay in food. No amount of scolding beats a sandwich.
Attention
Jumping, barking and nudging often pay in eye contact and touch — including the negative kind.
Distance
Growling, lunging and snapping pay by making the scary thing go away. That payoff is powerful.
Management: stop the rehearsal
Every repetition of a behavior makes it stronger. Before you teach anything, make the old behavior hard to practise — not with a correction, with the environment.
- Counters cleared and bins behind a cupboard door instead of a scolding after the fact.
- Gate or leash at the door instead of a grab as the guest arrives.
- Cross the street or turn early instead of correcting a lunge you could have avoided.
- Feed the dog before dinner, in another room, instead of pushing them away from the table.
- Put the shoes away. Management is not cheating; it's the part that buys you the time to train.
Name the behavior you actually want
“Stop jumping” isn't a plan. “Four paws on the floor while a person says hello” is. The replacement has to be specific, physically incompatible with the old habit, and worth doing.
- Jumping → four on the floor, or sit, paid by the greeting itself.
- Barking at the window → come find me in the kitchen, paid with food.
- Pulling to a smell → loose leash gets you there; tight leash stops the walk.
- Growling over a chew → a trade routine, so your approach predicts something better arriving.
- Counter surfing → go to your mat while food is prepared, paid from the counter.
The step-by-step marker reward plan
Step 1 — Write down the ABCs
Antecedent
What happens right before? Doorbell, food on the counter, dog across the road, you picking up keys. Be specific about distance and time of day.
Behavior
Describe it as a camera would see it. “Barks four or five times, front feet leave the floor” beats “goes crazy.”
Consequence
What does the dog get? Food, attention, the scary thing leaving, a chase game with you. That's the payoff you have to match or beat.
Step 2 — Change the setup
Remove the trigger where you can
Film on the window, bins away, feeding done earlier, walk route changed. You're buying quiet weeks.
Lower the intensity where you can't
More distance from other dogs, quieter times of day, a helper instead of a real visitor. Your dog should be able to eat and move normally — that's the sign the level is right.
Step 3 — Teach the replacement away from the trigger
Build it in an easy room
Mat, hand target, come-find-me, trade — whatever you chose. Ten clean reps a day for a few days, no trigger anywhere near.
Pay like it matters
Real food, delivered fast. The replacement has to out-earn the habit, and early on that means chicken, not kibble.
Add a cue once it's fluent
Name it only after your dog is offering it happily. Naming a shaky behavior gives you a shaky cue.
Step 4 — Reintroduce the trigger at low volume
Start ridiculously easy
One knock from the next room. A dog visible at eighty feet. A single crumb on the counter. Ask for the replacement, pay, end.
Two misses means back off
If your dog fails twice in a row, the level was too high. Add distance, reduce noise, shorten duration.
Interrupt kindly when life goes wrong
It will. Use a cheerful noise and a food scatter to break the moment, get your dog out of the situation, and log it as management to fix — not as a discipline problem.
Where this fits in the Training Transformation Foundation™
Give the behavior somewhere legal to go
- Digging dog: a sand pit with buried toys beats a decade of arguing about the flower bed.
- Chewer: rotate appropriate chews and keep one in every room you can't supervise.
- Barker at the window: a designated watching spot plus a paid recall away from it.
- Shredder: cardboard and paper-wrapped food puzzles on a schedule.
- High-drive dog: a flirt pole or a structured tug game with rules, three times a week.
Common objections and honest answers
- “But the correction works.” It works on the symptom. Check whether the behavior moved elsewhere, or whether your dog is now avoiding you in that context.
- “I'll have treats in my pocket forever.” No — you pay heavily while learning, then shift to a variable schedule and to real-life rewards like the door opening.
- “My dog knows he did wrong, look at his face.” That posture is appeasement in response to your body language, not guilt about the shoe.
- “The old trainer said be the alpha.” Dominance framing doesn't match how dogs learn and it puts you in conflict with your dog for no gain.
- “Nothing motivates him.” Something does. Try higher value food, a favourite toy, sniffing access, or check whether the environment is too hard for eating at all.
- “It got worse when I stopped correcting.” Extinction bursts are real and brief. Keep management tight so the burst can't get paid.
When to get professional help
Go deeper
Print & keep
Your what-to-do-instead checklist
- Veterinary exam booked if the behavior appeared suddenly
- Warnings — growls, freezes, lip lifts — never punished
- Trigger written down with distance, time and context
- Payoff identified: access, attention or distance
- Management in place so the old habit can't be rehearsed
- One specific replacement behavior chosen
- Replacement trained away from the trigger first
- High-value pay used while learning
- Trigger reintroduced at a level where your dog can still eat
- Two misses in a row means the level drops
