My dog shuts down during training - what do I do?

Finn is the case the existing concept was written from, and it is also the case where a handler films his own failure. On a store field trip he says, on camera, that he tried to gauge how much the dog could take that day and gauged it too high - and then describes exactly what that bought: not a permanent shutdown, but enough that the dog's brain cannot do commands any more, just kind of in survival mode. He names the general problem in the next breath: gauging how fast a dog can take each command is sketchy, and getting it wrong builds negative associations. The complication on this dog is that the usual readout is missing - his pupils are permanently dilated, so the trainer says outright that he can never tell when Finn is actually nervous. A backyard session gives the rule itself, and it has two halves, one about the dog and one about the business: we are not going to force him if it is just going to shut him down, because it makes my job so much worse and makes it more miserable for him - and if he keeps resisting, we come back to it another time. Which he then does: a down is asked for, refused twice, and abandoned without escalation.

Lola is a nervous dog and her file argues both ends of the range, which is why it is the most useful one here. The lower bound is set as a question: would you rather have a dog who pants a little through the learning process, or a dog you never make pant who never learns. The same passage removes the excuse owners hide behind - a nervous dog under stress is not psychologically different from a confident one, she just shows it differently - and reads the dog's face against the owner's interpretation, because the tongue people call a smile is a sign of anxiety. Who sets the dose is answered carefully: the dog decides, and the trainer's job is to look at her and make an educated guess about whether she has taken in too much. The upper bound is set harder, and against the trainer's own trade. It is easy to make a nervous dog look opened up when what you have actually done is repress her, and to open her up properly you have to not repress at all. At the handover it acquires a name and a picture: the velcro dog, produced by so much collar and reinforcement that the dog is frightened to leave your side for the rest of his life - repressed so hard the dog has no soul any more. Not a fan of that.

Stormie is a young, soft dog, and her file makes the anti-repression argument to an owner rather than to a camera - and makes it partly on commercial grounds, which is what gives it force. Some people get a two-week board and train back and the dog does stay when told, but she has been repressed into it and has lost her personality, which is not what anybody sent a dog away for. The second reason is more practical and is the one an owner should hear: a week or two after she comes home the repression wears off, and the training goes with it. So the program runs slower on purpose - he would rather she learn impulse control and be comfortable with it than be traumatised or repressed into it - and he tells the family she is on schedule rather than behind. A handler briefing another handler before a session says the same thing as an instruction: she is young, less pressure, because too much and she will shut down fairly quickly. And in another session you can hear it happen and be handled: he is a little frozen, so let's chill out and run around for a bit.

Dutchess supplies the mechanical cause and the corpus's single most unusual response. The cause first: too many repetitions will shut a dog down on their own, and the rest interval between them is treated as a variable that changes by breed and by dog. The healthy version of the same thing is named in another session - applying some pressure, but not driving her to the point where she decides she cannot do it - which is as close as the corpus comes to defining the line prospectively. And then the response. At the hardest repetition in the whole corpus, with the dog fighting rather than folding, the handler says it out loud to the dog: a lot of fight, man, don't do that, I don't want to fight you - and then finishes the repetition anyway, without escalating, and praises her. It is a different move from stopping before a shutdown, because this dog has not gone flat, she has gone hard. The answer given is to decline the contest rather than win it.

Bullet's file contains the corpus's only published aborted session, and it is titled as one. Four minutes in the trainer says he is counting a little bit of shutdown and is not really sure why; he lets the dog run around wherever he wants; he says the dog seems more suppressed than he would like to see; and he decides the answer is not a different exercise but more time spent chilling and bonding. A much longer session elsewhere states the principle behind it in a form nobody else uses - the emotional state of the dog is like God, it is all important - and concludes that pushing when there is nothing there for the dog at that moment is simply a bad idea. His first day out carries the same idea as a rule for first sessions: day one is not for running commands, it is for the trainer to learn what this dog is like.

Deuce is a large, robust adolescent rather than a nervous dog, and his file shows the same rule applied where nobody would expect it. Mid-session the trainer grades his own pacing and finds it slightly too hard - the dog is showing very light signs of commands being drilled too fast, and his brain is a little out of it - and the fix he chooses is not fewer commands but longer gaps between them. He then states the thing that makes early stopping bearable for a handler who feels he should be doing more: the session does not need to last forever if you accomplished something. Another session adds two more stop conditions that have nothing to do with the dog's mind - heat, because working him in the sun just feels cruel, and tiredness, read off an increase in pulling rather than off anything dramatic.

Dolce is a protection-program Rottweiler whose obedience ran ahead of her confidence, and her file holds the most useful distinction on this page. Watching a handler work her through a stuck heel, a senior trainer reads her out loud: she seems happy, she is totally fine, she is not shut down - she is unsure, she does not know how to get rid of the pressure, and she is trying stuff. That is the line. Unsure and experimenting is a working dog; stopped participating is not. Her own file also records the flat fact that she shuts down semi-quickly, and, from the same session, the summary that complicates this whole concept: it was a really good session, maybe not the everything-went-well kind, because they had a lot of healthy conflict. Elsewhere the break is asked for out loud by a handler - is there a three-minute break so she can calm down and not get pressured.

Sage is a nervous dog whose sessions show what stopping early actually costs in footage: long, apparently empty stretches. The trainer explains them. He gives her free time and lets her do whatever she wants for a few minutes because she needs extra time between repetitions, and points out that in earlier videos there are ten or fifteen minutes between some reps - because he would rather get fewer repetitions done with her in a good emotional state each time than push through, and pushing through does not do much good. In a store he says the quiet part to the dog rather than the camera: let's just walk for a bit, I don't want to get you shut down already. And in another store the whole exposure is made conditional on it - this is good exposure for her, as long as she does not go into a bad emotional state here.

Kodak is a fearful dog who wanted to bolt, and his file states the order of operations that sits underneath this whole concept. Could obedience be got out of him? Yes - and it would look gross, it would look disgusting. What is wanted instead is a happy-looking dog, and the happier the dog looks, the more that is the proof. The trainer is careful to say what that does not mean: not that they stop using negative reinforcement, or requiring things, or setting boundaries, or penalising - only that there is an order, and other things have to happen first. Fifty minutes into a first session he demonstrates the discipline of it. Asked, in effect, whether he could push a little further, the answer is no, and the reasoning is the most unusual thing in this concept: what would it accomplish, it makes me feel good, who cares - I want to make him feel good.

King had no authority problem and no fear problem, and his file is where the rule is applied in its least dramatic and most transferable form. Diagnosing any failure, the trainer says there are two components: the dog's emotional state, and how clear the thing was to him. From which follows the instruction that would change most owners' behaviour if they heard nothing else on this page - never default to putting on more pressure; default to suspecting play. In a store, working a dog frightened of ceiling fans, the method is simply to stop doing anything: I'm not going to pull on you, I'm just going to wait. And at the lesson he declines to create a problem he does not have - King offers a down when asked for a sit, is allowed it because a down is more comfortable, and the trainer says he does not have to manufacture situations to stamp authority with this dog.

Luna's file is the one where sessions are visibly cut short on a read of the dog, repeatedly, and the reads are named. Her tail is a little low, so the session is wound down - and the reason given is about what she is learning to expect: she thinks out here means a bad experience every time, and that is not good. A down is shelved mid-session rather than pressed, on the guess that she needs to open up a little more first. And a short session is explained rather than apologised for: he would usually go fifteen or twenty minutes, she was finished at nine, so this is where he gives her the time she needs. Another session shows the read that stops a session before it starts - a yawn that, he says, tells him everything.

Roxy's first lesson supplies the target zone, which is the closest the corpus comes to a prospective description of the line rather than a post-mortem one. There is knowing how to push, and then there is pushing to where the dog is close to threshold but not shut down - where the dog's reaction is something like whoa, that was a lot. What sits underneath it in the same conversation is the reason any of it is worth doing: pressure with a release teaches the dog to cope with pressure, because a great many dogs have no idea what to do with that feeling at all - they tighten up and feel weird - and what they get out the other side is relief, then resilience, then competence.

Louie is a rescue case with a bite history, and his file gives the other position on conflict - the one that has to be read next to Dutchess's refusal of it. In a programme with a dog like him, the trainer says, there are a few moments of conflict, and in those moments you make a lot of progress all at once. He describes what the conflict looked like: the dog's first reaction to a little pressure was not to give to it, he froze up, and the question in the air was whether they were going to fight now. The word used for the outcome elsewhere in the corpus is the same one used about Dolce - healthy conflict. Neither speaker addresses the handler who declines the fight, and none of the three sessions defines what makes a conflict healthy.

Zeus's session gives the cheerful version of the same discipline, on a dog who is neither nervous nor fragile. He had been biting at the leash; he learned that sitting is what gets him out of the pressure and back to freedom and fun, and that biting the leash is not. And then, with the lesson visibly landed, the trainer stops: that was plenty for him to learn that lesson, so he is not going to push it and go for another set. It is one of the clearest instances in the corpus of a session ending on a success rather than on a problem.

Bo's contribution is small and it is about quality rather than quantity. Watching a rep that produced the position he asked for, the trainer rejects it - he does not want the dog to just suppress and go into the down, he wants him to go smoothly into the position. That is the same distinction the rest of this concept turns on, applied to a single repetition rather than to a session: a dog who has complied because he has given up looks, from the outside, exactly like a dog who has learned. In another session the judgement is even shorter, and it is about continuing at all: it will just make it worse, and it will not be productive.

How do I know the difference between a dog who's struggling and a dog who's shut down?

The examples encourage looking at the whole sequence rather than using one visible sign as a test. During Dolce’s lesson, a trainer interprets her attempts as uncertainty while she is still trying options (yoecbjfaOa4, 08:57). Finn’s trainer describes the difficulty of interpreting features that are common for that individual (0JRh17nttVw, 16:08). Neither excerpt establishes a diagnostic checklist. Record movement, responsiveness, exploration and changes from the dog’s own baseline, and ask the trainer how those observations affect the next step.

Isn't any stress bad? Shouldn't I stop the moment he looks uncomfortable?

The trainers say no, and one of them argues the point directly at the gentler end. Would you rather have a dog who pants a bit through the learning process, or a dog you never make pant who never actually learns? The same session removes a comfortable illusion: a nervous dog under stress is not going through something different from a confident one, she just shows it more visibly - and the tongue hanging out that people read as a smile is a sign of anxiety, not of enjoyment. What is being defended is not discomfort for its own sake. It is that learning to cope with pressure is itself the skill a nervous dog is missing, and the sequence described is pressure, then relief, then resilience, then competence. The limit is the one everything else on this page is about: close to the edge, not over it.

What do I actually do when it happens?

Stop asking, and then do one of a handful of things the trainers do on camera. Give him free time - let him do whatever he wants for a few minutes, or longer; one trainer describes leaving ten or fifteen minutes between repetitions with a nervous dog, on purpose. Space the repetitions further apart rather than doing more of them. Make the next ask small enough that he can win it. Change the activity to play instead of adding pressure. Move him out of the heat, or simply walk for a while. Or end it: the session does not need to last forever if you accomplished something, and if he keeps resisting you come back to it another time. What none of them do is escalate to get the repetition finished.

My dog came back from training very obedient and very quiet. Is that good?

The trainers in this corpus would want a closer look, and they say so about their own trade rather than about a competitor's. A dog can be repressed into behaviour: she does stay when you tell her to, and she has lost her personality along with it. Two reasons are given for treating that as a failure. The first is that it is not what anybody sent a dog away for. The second is practical and it is the one to watch for: a week or two after she comes home the repression wears off, and when it goes the training goes with it. The same warning is given a name in another lesson - the velcro dog, produced by so much collar and reinforcement that the dog will not leave your side ever again. If your dog has come back quieter, flatter and less himself, that is the thing to ask about.

He's fighting me rather than shutting down. Is that different?

Yes, and the corpus handles it differently and does not entirely agree with itself. At the hardest repetition anywhere in these files, with the dog fighting hard, the handler says it out loud to her - a lot of fight, man, don't do that, I don't want to fight you - and then finishes the repetition without escalating, and praises her. That is declining the contest rather than winning it. On another dog the opposite view is stated: with some dogs there are a few moments of conflict, and those moments are where a lot of progress gets made at once. Both are in the record and neither addresses the other. What is stated clearly is the limit at the other end: if the dog has no authority problem, do not invent a confrontation to prove one.

If I keep stopping, will he ever learn anything?

The trainers' position is that stopping is how he learns anything, because repetitions run past the point of shutdown teach nothing and cost the next session as well. The arithmetic is stated: fewer repetitions with the dog in a good emotional state each time beats pushing through, and pushing through does not do much good. A session does not have to be long if something was accomplished, and one trainer stops the moment a lesson lands rather than running another set to be sure. Two of these dogs took longer in the programme as a direct result, and in one case the family were told so, and told the dog was on schedule rather than behind.

Doing more repetitions to fix a repetition that is going badly. Too many reps will shut a dog down all by themselves.

Adding pressure as the default answer to a dog who is struggling, instead of suspecting that play is what is missing.

Reading a panting dog with her tongue out as a smiling happy dog.

Taking a compliant, quiet, easy dog as proof that the training worked, when what has happened is that the dog has been squashed - and losing the training a week or two later when the pressure lifts.

Correcting and then staying cross afterwards, so the residual mood does the damage the correction did not.

Working the dog when the conditions are against him - in the sun, when he is already tired, or when his state is simply not there that day.

Pushing for one more repetition after the lesson has already landed, instead of ending on it.

He pants heavily, his eyes go wide, and he stops responding to words he knows.

He goes still and will not move - not resisting, just frozen.

His tail drops when you get to the place where you usually train, before anything has happened.

He runs out of steam much sooner than he used to - what was a twenty-minute session is now nine.

He is obedient and somehow flat - he does what he is told and does not seem like himself any more.

He drops into the position the instant you ask, but it looks like collapsing rather than moving.

He has started leaving - wandering off, going to his bed, or finding somewhere else to be when you start a session.