Does my dog have to like other dogs?

The case where the rule is stated as a rule, three times in one handover, because it is the standard a household of four other dogs will be living by. Avoidance is fine. They do not have to be best friends. They just need to not get in the fight. And then the hard half, which is the sentence that matters: friendship is not required, neutrality is required, that is not optional, he is not allowed to go after him. The goal was scoped before any of it was tested - not dogs in general but four specific dogs at one address, of which the two smaller ones are no problem and the two bigger, younger, more playful ones are - which is what made the case finishable at all. And the trainer's own summary of the ceiling is the least salesmanlike line in the corpus: sometimes all you can hope for is neutral. What he does when the criterion is met is the part owners never copy. The resident dog comes close, Zeus does nothing, and the trainer praises him out loud and then tells the owner, in as many words, that this is the thing you praise - because he is being neutral.

The dog on whom the criterion gets its two most useful refinements. The first is that neutrality is not a single skill: being neutral to people does not mean being neutral to dogs, to geese, or to random things in the air - so a dog who is superb with strangers may be nothing of the kind with the next moving thing, and in this case that is exactly where the residual risk sits. Her people-neutrality is finished; her dog-neutrality is the open question, and the trainer says plainly why it matters and does not soften it: she has to be safe around all other dogs, because if she is not that could be really dangerous, and in fact it already has been. The second refinement is that neutral has two sides on the human category too, given in one breath - he does not want any of the dogs to be frightened of people, and in the same breath they must know they cannot go running up to them. What the finish line looks like is narrated three times in this one session, and the evidence offered each time is the handler's own inaction: she thought about it, and he did not have to press anything, and the big thing is that she is correcting herself. Her second file supplies the intermediate state that makes progress countable before that: still fixated, still staring, but no lunge and no tension on the leash.

The pair case, where the criterion is a promise made in advance rather than a result claimed afterwards - and where the promise is split honestly into a floor and a ceiling. The standard is set before the first attempt and the owners are told it: best friends maybe, who knows, does not need to happen; peaceful coexistence required, because that is where escalation starts. At the first owner lesson the two possible outcomes are laid out side by side without preference - they could end up best friends, or they could tolerate each other as long as they leave each other alone. And then the line that is rarer than either: the worst case is obviously possible and he does not have much control over the best case, but he can set up the best possible chance for it. He guarantees the floor and only offers odds on the ceiling. The same file gives the permission an owner most needs and least expects - even if he is looking kind of hard, as long as he does not do anything, you are fine, let him look - and pairs it with the prior instruction that makes the whole thing work, which is to let the dog make the bad decision rather than pre-empting it with a stream of commands.

The dog whose file explains what the finished article actually is, and it is not a decision. The mechanism is given to the owner with the dog standing there: say the cue at every single instance of the trigger, until the dog's brain reaches the conclusion that every time we walk past a dog he says this, so I will predict it this time - and what an onlooker then sees is a seamless dog who does not care about the dog. The important part is the sentence after it, in the parallel passage from the first owner lesson: it looks like a dog who is indifferent, but psychologically there is a whole thought process going on. That is the honest account of most 'neutral' dogs in public, and it is the answer to an owner who wonders why their own dog does not just naturally stop caring. It also sets up the conflict on this page, because on this account the dog never chooses anything - the handler has installed a competing prediction, and the indifference is the appearance rather than the thing.

The dog whose file contains the warning that governs how neutrality may be built. Do not tap the collar every time she sees a dog: a dog with nervous tendencies will draw the obvious conclusion - all dogs are bad, they are dangerous, every time I see one I feel this - and you will have manufactured the aversion you were hired to remove. The alternative in the same lesson is the cue-every-time route, and the two together are why this page has to sit next to the emotional-change page rather than being read as a compliance standard: the same visible outcome, a dog who does nothing when another dog passes, can be built two ways and only one of them is worth having. Her handover also supplies the rule that keeps an owner out of the other ditch, because it is the mistake families make while trying to be diligent: never punish a position the dog was not put in.

Is staring a pass or the thing you intervene on?

Both are in the corpus within the same fortnight and one of them is in the same session as this concept's headline rule. They are probably distinguishable in practice - a look with a soft face against a head-up hard stare with a freeze - and the handover session does draw exactly that distinction elsewhere in its body-language teaching. But as instructions to an owner they read as opposites, and an owner told 'let him look' and 'a stare is what you don't want' has been given a judgement call rather than a criterion. Not resolved here; what it needs is the two pictures side by side on one page.

What the finish line is: a decision the dog makes, or a prediction you install.

Nobody compares them and the two produce the same footage. The difference matters in one specific circumstance, which is when the handler stops cueing: a dog who declines on his own has something a dog who has learned to expect a cue may not. It matters commercially too, because 'she corrects herself' and 'I say let's go every single time' are different promises about what life will be like afterwards. Left open.

How wide the goal is.

Not addressed, and both are defensible: a dog cannot be proofed against all dogs and can be proofed against four, which is what makes a case finishable - but a dog scoped to four is not a safe dog in a park, and one of these two dogs had already injured household dogs. The practical question a page has to answer is what a family is being told they have bought, and the corpus answers it differently in the two cases. Left open.

Will my dog ever be able to play with other dogs again?

Maybe, and the trainers are careful not to sell it. What they commit to is the floor: the dogs do not have to be friends, they have to not fight. Avoidance passes, ignoring passes, sniffing and playing pass - and going after the other dog does not, ever. On a two-dog household the two possible outcomes are set out side by side without a preference: they could end up best friends, or they could tolerate each other as long as they leave each other alone. The most honest version of it is the trainer who separates the two halves explicitly: the worst case is obviously possible, he does not have much control over the best case, and what he can do is set up the best possible chance for it. So the answer is that friendship is a bonus you might get, and coexistence is the thing being bought.

Is it a failure if my dog just avoids other dogs?

No - it is one of the passes. The criterion is stated with avoidance named first: avoiding is fine, if he avoids the other dog there is no problem with that. Another session puts the same thing in the corpus's own vocabulary and says explicitly that they would rather the dog ignored than fixated and went after, so avoidance is a lawful choice and acceptable. This is worth taking seriously because avoidance is the outcome owners are most likely to be disappointed by and it is the one that keeps everybody safe. It is also the one the trainers will actively reward: a dog who declines to engage gets praised for it out loud.

He stares at other dogs. Should I correct that?

This is the one place the corpus gives an owner two answers, and both are on film, so it is worth knowing what each is pointing at. The permission is broad: even if he is looking kind of hard, as long as he does not do anything you are fine, and you can let him look - and a progress report on another dog counts a stare with no lunge and no tension on the lead as a good outcome. The caution is narrower and is about a specific picture rather than about looking: a head put up over the other dog, a freeze, and a hard stare is named as what you do not want. The practical reading is that a look is a look and a hard-eyed freeze over the top of another dog is a posture, but nobody in the corpus sets the two side by side, so an owner is being asked to make a judgement rather than apply a rule.

When he ignores another dog, should I do anything?

Yes - praise him, and this is the instruction owners almost never get. In a case whose deliverable is indifference, the behaviour you want is an absence, and an absence never gets rewarded by accident. So when the other dog comes close and the dog does nothing, the trainer says good boy, good job out loud, and then turns to the owner and says that this is the kind of thing he praises, because the dog is being neutral. It is twenty seconds of a long handover and it is probably the most immediately useful thing on this page. One caution from another file belongs with it: never tell a dog off for getting up out of a position you never put him in, because a dog punished for something he was not asked to do cannot learn anything from it.

My dog is brilliant with people. Why is he still terrible with dogs?

Because they are separate skills, and this is the sharpest thing anybody in the corpus says to an owner who believes their dog is simply good in public. Neutrality is treated as an association with a particular class of thing rather than as a general temperament: being neutral about people does not mean there will be neutrality about other dogs, or geese, or random things in the air. Each has to be installed. It is also worth noticing that the human category has two sides, given in one breath by the same trainer: he does not want any of his dogs to be frightened of people, and in the same breath they need to know they cannot go running up to them. A dog who hides from strangers and a dog who mugs them have both failed the same criterion.

He lunged at a dog and I corrected him. Why did it get worse?

Probably because of when, rather than whether. The corpus's instruction is to get the dog off the other dog first and correct him after: if you say the word and tap the collar while he is still in that position and still ramped up, he can take it out on the other dog, and the correction becomes an escalation because the other dog is what he is focused on and is what he will attribute it to. The same logic applies more broadly on a nervous dog - a correction that lands on the appearance of another dog rather than on a behaviour teaches the dog that other dogs are what hurt him. In both cases the fix is the same: change what the correction is attached to.

Aiming at friendship. The deliverable is that they do not fight; sniffing and playing are a bonus, and treating them as the target is what puts a household in the situation the criterion exists to prevent.

Letting a non-event go unpaid. A dog whose target behaviour is doing nothing will never be rewarded by accident, so the trainer praises it out loud and tells the owner to do the same.

Correcting the sight of another dog rather than the behaviour, which teaches a nervous dog that other dogs are dangerous.

Correcting while the dog is still on the other dog. The corpus's instruction is to pull him off first, because a correction delivered while he is ramped up and focused on the other dog can be attributed to that dog and becomes an escalation.

Assuming that a dog who is excellent with people is neutral about anything else. Each category is separate, and a dog's public manners say nothing about dogs, birds or a thrown ball.

Pre-empting the choice with a stream of commands, so that the dog never actually decides anything and the handler never learns what the dog would have done.

Reading a frightened-of-people dog as a well-mannered one. The criterion has two sides: not scared of people, and not running up to them.

Your dog is faultless with people in public, and you have assumed that means he is fine with dogs.

The incidents in your house have all been with dogs your dog knows, not with strangers.

You have two dogs at home who have never worked out how to be around each other, and you have been keeping them apart.

On a walk you correct him the moment he so much as looks at another dog, and the whole walk has become a series of corrections.

He has got quieter around other dogs and also more wary of them - he used to be interested and now he avoids.

He stares at another dog with his head up over it and goes still, and you cannot tell whether that is fine.