Can another dog help - or test - my dog?

The case where the helper dog is standing practice rather than an improvisation, and where the mechanism is stated. A specific dog is named and the phrase used is 'open them up' - not tire out, not socialise, open up - and it is said as something the handler does generally rather than as a plan for this dog. The reason given is social and specific: she takes the lead, and he relaxes into it and concludes that life is good. And the effect is reported as a rate rather than as a state - she gets him a lot looser, a lot faster, than working him would. In one of the two sessions she also produces the day's best result indirectly, because the thing he is released back to is her. The same file carries the one fairness rule in the corpus: when the client dog is holding a position, the helper dog is put in a down as well, because the handler does not want her walking around while he holds a stay, and he says out loud that this is about fairness.

The household version, and the only lever in a nervous case that no handler can supply. It is stated flatly, in front of the owners, while the dog trots back with a normal tail carry: Archie is good for her, because Archie is confident and happy to be alive and to see people, and having her out with him does good things for her. The pair's dynamic is then treated as an outcome in its own right and tracked across the programme - he used to want to be ahead of wherever she was, and by the end of a long lesson the trainer's observation is simply that they are walking together. A later session gives the honest state of it: he is very confident, she is still a little less confident, and she has got better. The reason this dog sits awkwardly on this page is that her own file also contains the opposite finding, which is that the presenting problem existed in the pair rather than in either dog - and that tension is laid out under conflicts rather than smoothed over here.

The most permissive use in the corpus, and the one that shows what the priority order really is. With a shut-down dog the trainer will not discourage behaviour he does not ultimately want - prey chasing included - because normal dog behaviour is the prerequisite and obedience can take it back later: he just wants him to act like a dog, be normal, do dog stuff, not be all freaked out. The partner dog is then chosen for a trait that would be a fault in a tester: persistence. Boogie gets the job because she is not going to give up, which is exactly what a frightened dog who will not initiate needs. And the arrangement has precisely one rule on the other dog - no humping - with everything else allowed, and the handler reports having corrected her for it in the plainest terms. The result is graded honestly rather than claimed: it was better than before, he did not run and hide when released, he let her stand on him and let the handler do things with him.

The dog whose sessions contain two of the four jobs, on three different partner dogs. As a tester: another handler's dog is walked past her held sit to see whether it survives, the pass is repeated until it does, and the passing dog is named out loud as a tester - which turns a chance distraction into a repeatable procedure. As a draw: a dog who otherwise stays close to her handler most of the time has no reason to leave, so another dog supplies the reason, and the handler says plainly that this is the only realistic way to teach that recall. Note what the reward then is - the release back to the other dog, which makes the helper dog the payment as well as the motive. And at the end of one of these sessions the criterion is stated in the praise itself: you did not let him bother you at all.

The ambient use, and the only place in the corpus where two dogs in one room is stated as a design rather than as a compromise of space or time. The shepherd holds a place while the handler teaches the retriever a recall a few feet away, and the reasoning is explicit: it makes a high-distraction environment because he is going to be visibly, deliberately engaged with the dog that is not her, and she has to learn to sit there and chill out. That is a harder distraction than an environmental one, because it is the handler's own attention. Then the arrangement is justified a second time from the other dog's side: the retriever has a bit of nervousness, and forcing him to lock onto the handler rather than let his nerves get to him makes it a growing experience for both dogs simultaneously. One setup, two problems. The same session also shows the plan being downgraded on arrival - the kennel is too loud for the nervous dog, so keeping the environment low is worth more than the exercise that was planned.

The case that explains why a dedicated tester exists at all. The business has a dog kept for the job and says so - she is our doggy tester - and the reason she is used here is stated in the same breath: the dog being assessed has been sketchy with other dogs and launches at them, and they need more repetitions where the behaviour can be penalised. The deeper reason appears two days later in a lesson, and it is the most useful single observation on this page: how bad an encounter gets is only partly a property of the dog you are testing. A dog who comes back at him makes it a fight. A dog who shows a lot of submissive body language probably does not, though the trainer says that is still not good. So a known, stable tester removes one of the two variables, and a pass in a staged meeting is a measurement of your dog only to the extent that the other dog was a constant. His household's own version of the same finding, from the pair case, is that a dog who screams and acts like a victim does not stop anything.

The dog whose only recorded role in the corpus is in somebody else's programme, and it is as an instrument. One handler walks him past another handler's dog specifically to test a held sit, and in the same session the two handlers exchange the honest note about him that goes with the job: asked whether he is a nipper, the answer is that he will bite you. That pairing is worth keeping, because it is the corpus's clearest illustration that a helper dog and a safe dog are not the same thing - for the job he is doing here, which is to walk past on a lead at a distance, his own bite history is not disqualifying, and the handler working alongside him knows about it and keeps his hands clear. A dog used as a tester is being asked for reliability under one narrow condition, not for general soundness.

Is a second dog the medicine or the problem? Lola's own file says both.

Both are the same trainer about the same two dogs, three weeks apart, and he does not remark on the reversal. The reconciliation is probably in the second dog's own state - an over-aroused Archie transmits arousal, a trained and settled Archie transmits confidence - and the household was told to send both precisely so that the second dog could be changed too. But nobody says that, and it is the single most important thing this page would need in order to advise an owner with two dogs. The risk is named in the same file from the other direction: one dog winding the other up is why sending a pair is harder, and another household is given as the example.

What the other dog in the room is allowed to do.

Not a contradiction once the exercise is taken into account - a place hold and a play assessment want opposite things from the second dog - but neither handler states the condition, and the two policies would look identical on a booking sheet. It is worth keeping both because the fairness rule is the kind of thing that never gets written down and the permissive rule is the kind of thing that gets criticised without its reason attached.

What makes a dog suitable for the job.

These are close to opposite traits and the corpus never notices, because nobody in it lists the jobs. A persistent dog is right for a play assessment and wrong for a neutrality test; a steady dog is right for a test and would achieve nothing with a dog who initiates nothing. Left open, and flagged as the thing a handbook would most usefully add.

Would getting a second dog help my nervous dog?

The corpus is genuinely encouraging about an existing confident dog and says nothing at all about acquiring one, so take the first part only. Where a household already had a confident, cheerful second dog, the trainer treated him as an active ingredient rather than as company: Archie is good for her, because Archie is confident and happy to be alive and to see people, and taking her out with him does good things for her. He tracked the pair as an outcome in its own right and reported at the end that they were simply walking together. The important caveat is in the same dog's file, from three weeks earlier: walking the two of them together is also where the reported reactivity came from, because the excitable one got excited and she could not place it. Which of those you get appears to depend on the state of the second dog, and the household in question was advised to send both dogs to training rather than one.

Can I test my dog with a friend's dog?

You can, but understand what you are and are not measuring. The trainers keep a dog for this and say so - she is our doggy tester - and the reason is stated elsewhere in the same case: how bad an encounter becomes is only partly a property of your dog. A dog who comes back at him makes it a fight. A dog who shows a lot of submissive body language probably does not, though the trainer adds that it is still not good. And from a different household's own experience, a dog who screams and behaves like a victim does not end anything. So a stable, known dog holds that variable constant and lets the meeting tell you something about your dog; an unknown dog turns the test into a gamble whose result you cannot attribute.

My dog won't leave my side. How do I teach a recall?

By giving him somewhere he wants to go. The corpus's answer to this is blunt and immediately usable: having the other dog there is the only realistic way to teach that recall, because the dog wants to go to her - and without her she stays close to the handler most of the time. A dog who never leaves you has never had the opportunity to come back, so there is nothing to train. The neat part is what the reward then is: the release back to the other dog, so the same animal is the motive and the payment. It is worth noticing that this is also the honest diagnosis of a clingy dog. Staying close is not obedience, and in one nervous dog's file the trainer says outright that a dog tied to him by nerves cannot have her recall tested at all.

I have two dogs and no time to train them separately. Is that a problem?

In this corpus it is treated as an advantage, and stated as a design rather than a compromise. Two dogs from one household are worked in the same room on purpose: one holds a place while the handler visibly, deliberately teaches the other a recall a few feet away, which makes a much harder distraction than an environmental one because it is the handler's own attention. And the arrangement is justified twice over - the same session forces the nervous one to lock onto the handler instead of letting his nerves get to him, so it is described as a growing experience for both dogs simultaneously. One setup, two problems. The fairness rule from another file belongs with it: when one dog is holding a position, put the other in a position too rather than letting him wander about.

Should I stop my shut-down dog chasing things, since I don't want him doing it later?

Not yet. The corpus's position on a shut-down dog is that normal dog behaviour is the prerequisite and obedience can take it back afterwards, so behaviour the programme does not ultimately want - prey chasing is the example - is deliberately not discouraged. The trainer's stated aim is that the dog act like a dog, be normal, do dog stuff, and not be all freaked out. The same logic governs who he puts the dog in with: a partner dog is chosen for persistence, because she is not going to give up and a frightened dog will not initiate anything. And the result is graded honestly rather than claimed - it was better than before, he did not run and hide, he let her stand on him.

Leaving the other dog loose while one dog holds a position, so the dog doing the work watches the other one enjoy itself.

Testing with an unknown dog. How bad an encounter gets is partly a property of the other dog, so a stranger's dog turns a test into a gamble and can turn a look into a fight.

Walking a nervous dog beside an over-excited one and blaming the nervous dog for the result. In the corpus's clearest case the reactivity lived in the pair, and the dog alone at a park did nothing at all.

Pushing a play introduction in a loud, chaotic setting. In one session the planned exercise is abandoned on arrival because the kennel noise is too much for the nervous dog, and keeping the environment low is judged worth more.

Discouraging a shut-down dog's first normal dog behaviour because it is not behaviour you want long-term. In the corpus that behaviour is actively left alone, on the grounds that obedience can take it back later.

You have a second dog who is genuinely easy and confident, and you have been keeping the two of them apart to avoid trouble.

Your dog never leaves your side outdoors, so you have no idea whether he has a recall - there has never been anywhere he wanted to go.

He will hold a position perfectly until you turn your attention to the other dog, and then he breaks every time.

He does not play, does not investigate, does not do normal dog things - and you have been working on his obedience instead.

He is fine with dogs on his own and impossible when you walk the two of yours together.

The encounters that went badly were all with dogs who screamed or scrabbled, and the ones with confident dogs ended in a second.