Why did my dog become aggressive?

The case where the diagnosis is stated twice, from two directions. Early, in the dog's own run, the trainer rules something out on an observation anybody watching can check: he turns his back inside the run and the dog is untroubled and lies down, so this is not territorial. Weeks later, at the second owner lesson, the household describes a snap and the trainer takes the description apart question by question - does he growl, does he bark, what does he do afterwards - and lands on a name with a signature: instant snap, no growl, then fear, then the dog soliciting affection from the person he snapped at. He separates that from aggression and from protection and then ranks it, telling the family that of the things it could have been this is the one to want. At the handover the whole arrival behaviour is re-read the same way: insecurity, not intent. The film also records what that arrival behaviour was, in the trainer's own words - on the day the dog came he could not be got out of his crate and went for two handlers' hands - so the reading is not being offered in place of the facts.

The dog whose origin story nobody in the corpus tells, and whose trigger everybody names. There is no session in his file that explains what happened before he arrived. What there is, twice and independently, is the shape of the thing: a handler meeting him cold says he is much more reactive when he feels confined and cites a second trainer as having seen the same, and months later, at the placement, the man taking him on says the same thing from the other side - that dog does not like being cornered, and it only bites when it is scared. Neither speaker had access to the other. Alongside that sits a smaller, more practical note logged as a handling risk rather than as a temperament claim: he fixates a little and does not like hand movement. The bite history itself is stated on film as a fact rather than explained - the placement conversation refers to his having bitten a woman, and a handler in another dog's session answers a direct question with 'he'll bite you' - and it is left at that.

The pair's file carries the corpus's only dated, concrete mechanism for acquired dog-reactivity, and it is the owner's rather than the trainer's. Rocko lived sociably with other dogs - including his mother, and a house with four dogs - until he was two, and then dogs from next door came into the yard and fought his dogs. The account is given twice, months apart, once in person on day one and once from a text read aloud on a walk, and both times it ends the same way: it was when he got bitten that he stopped making dog friends. The household's own management response is in the same passage and is worth as much as the diagnosis - after a move next to two aggressive dogs, their own dogs went outside on a rota. The trainer's response to hearing it is a prognosis rather than a reframe: we can fix it, he will be fine.

The arousal origin, and the only case in this set where dog and owner arrive at the same account independently. The trainer's version is a claim about a breed group - they get tight, overstimulate themselves, run out of options and bite - and he applies it to the specific events in this household without being asked to. The owner's version, in the same conversation, is that the times she bit the other two dogs she was simply that excited. The consequence for the plan is the part that matters: because the bites came out of over-excitement rather than intent, the work is on the over-excitement, and the trainer says outright that he does not repress dogs and that if you fix the over-excitement there should be nothing left to repress. The same lesson set contains the other half of the household picture: the dogs she stopped for were the ones who gave a clean confident signal, and the two she injured were the most submissive.

The learned-strategy origin, named more precisely here than anywhere else in the corpus. The trainer does not say the dog is aggressive at the crate; he says two patterns have been installed in him, and he lists them: he either fights when people try to get him out, or he growls and threatens until people leave and back off. Both of those had been working, which is the whole point - the second one in particular is a behaviour that has been reinforced by every person who ever backed away. The treatment follows directly from the origin: violate the pattern by not leaving, and layer a complete alternative over the top. And the store session a couple of months earlier gives the before picture in the trainer's own words, offered to a stranger who asked whether the dog was friendly: he came out of his own front door bristling and freaking out at a man who had come to collect him, and he did it with everybody he saw.

The dog who is in this file as the counterweight. Her trainer is asked, in effect, what happened to her, and he declines to invent an answer: they got her from the pound and nobody knows what her life was. Then he argues the harder position - that for a great many dogs nothing happened, that this is baseline temperament, and that his evidence is that pound dogs are not the only ones who arrive like this. Later in the program he splits the same dog's nervousness in two and gives each half a different prognosis: the environmental part was essentially gone inside a fortnight, and the genetic part is not going anywhere and has to be managed. Nothing in her file is called aggression at any point, which is itself the finding: the same barking, growling and lunging that a household would report as aggression is handled here as fear with a leash attached.

Whether a dog's presenting state has a history behind it at all.

Not addressed by either speaker, and the two dogs are different enough that both may simply be right about their own case. The reason it is worth laying out is that the two answers point at different work and at different promises to a family: an acquired problem has a prognosis, and a baseline temperament has a management plan. Lola's own file later splits the difference for one dog - environmental nervousness gone in a fortnight, genetic nervousness permanent - which is the nearest thing to a reconciliation anywhere in the set, and it is still one trainer about one dog.

What the intake forms record about biting, against what the footage records.

Not resolved, and not necessarily a contradiction in every case: a snap is not a bite, several of the filmed bites happened after the form was filed, and a household answering a yes/no question about their own dog may be answering about people in their family rather than about strangers or handlers. What can be said is that the yes/no bite questions on these forms did not surface the histories the footage contains, and that in two cases the free-text incident field did carry the relevant fact while the tick-box next to it said no. Dutchess's form is the exception that shows the format can work: it reports bites on dogs and describes three emergency-vet visits in a household, and the footage never contradicts it. (intake facts paraphrased from bible/intake/zeus.json, kodak.json, bo-[family name withheld].json, lola.json, louie-2.json, dutchess.json)

How many bites are in Louie's history.

Not resolved. The two accounts are not incompatible - one is a household's cumulative record and the other is a single incident discussed at a placement - but they are very different in size, and nothing on film indicates that the people receiving this dog had the larger figure. The detail that the bites never happened away from home is the most load-bearing thing on the form for this concept, because it is exactly the confinement-and-cornering read that two independent speakers reached on camera, arriving from the household's side. Whether anybody made that connection at the time is not recoverable.

My dog snapped at someone. Is he aggressive?

The trainers in this corpus would answer that by asking three questions about the event rather than about the dog. Did he growl or bark first, or was it instant with no warning? What did he do in the next few seconds - get frightened, or stay switched on? And did he then go to that person for attention? Instant, no growl, then fear, then soliciting affection is the pattern one trainer names a fear bite, and he separates it explicitly from aggression and from protection. He also says the honest thing about it: of the versions this could have been, that is the one you want, because a dog who means harm is a materially different case. None of that makes the snap acceptable or the dog safe around strangers - the handling instruction that follows it is about strangers, and it is to stop the person reacting - but it does change what the work is.

Did something happen to him before we got him?

Sometimes, and sometimes nothing did. The corpus has one clean case of each. In one, an owner supplies a dated mechanism for her dog's reactivity - he lived with other dogs happily until he was two, then dogs from next door came into his yard and fought his dogs, and it was after that that he stopped making dog friends. In the other, a trainer asked the same question about a pound dog says outright that nobody knows what her life was and then argues the less comfortable answer: a great deal of the time nothing happened, this is baseline temperament, and pound dogs are not the only ones who arrive this way. The practical difference is the prognosis. The same trainer later splits one dog's nervousness into an environmental part, which was essentially gone in a fortnight, and a genetic part, which he says has to be managed rather than cured.

He only does it when he is excited, not angry. Is that still aggression?

It is treated as a different problem with a different solution. On one case the trainer describes a pipeline rather than a temperament: the dog gets tight, overstimulates herself, runs out of things to do with it, and bites. The owner then confirms, without being led, that the times her dog injured the other two dogs in the house that was exactly the situation. What follows from that reading is the important part - the work goes on the over-excitement rather than on the biting, and the trainer says that if you fix the over-excitement there should be nothing left to suppress. He is not saying the injuries were minor; two dogs in that household went to an emergency vet. He is saying the lever is upstream of the event.

Is it the breed?

The corpus answers this two ways and both are on film. Breed anatomy is used to set the handling standard: with a Cane Corso the bite force is given to the owners before any behaviour is described, as a figure over 700 psi and a plain statement that it could snap an arm, immediately followed by the trainer saying he does not expect full force. And a breed-group tendency is named as a mechanism - that the bully breeds get tight, overstimulate and bite for want of anything else to do. But every actual diagnosis in this set is about a mechanism rather than a breed: insecurity, arousal, a learned pattern, a bad experience, or baseline nerves. Two Rottweilers in the same programme at the same time get opposite readings, and one of them is told his household's small dog is safe because of the dog's own temperament rather than because of anything training did.

What do I actually say about him to other people?

Name the condition, not the character. The most repeated practical move in this set is to replace 'he is aggressive' with a sentence that has a situation in it: he reacts when he feels confined; he does not like being cornered; he does not like hands moving near his face. That gives a household a rule it can follow and a stranger an instruction they can obey, which a label does not. It also happens to be more accurate more often - and one trainer's public version of the same honesty is worth copying: asked in a store whether the dog was friendly, he answered yes, and included the past tense.

Reacting to the snap. The one handling instruction attached to the fear-bite reading is to get the person not to react at all, and the trainer says in the same breath that this is the hardest thing to get people to do.

Backing away from the threat, which pays it. The crate case names this as one of the two patterns the dog had learned - growl, threaten, and people leave.

Managing an acquired problem forever instead of treating it. The household in the reactivity case had put their own dogs on a rota for the garden, which kept everybody safe and taught nobody anything.

Letting a child put their face on a dog whose limit nobody has found. Delivered as a general rule rather than as an accusation about the dog in the room, and the speaker says twice that he does not think this dog ever would.

The snap came with no warning at all - no growl, no bark, just an instant movement - and within seconds he was frightened and then trying to get affection from the person he snapped at.

It happens when he cannot get away - in a crate, on a short lead in a corner, at the vet, in a doorway - and not in open space.

The incidents happen at the highest-excitement moments of the day - visitors at the door, everybody greeting at once - rather than in standoffs.

He got on with other dogs until a specific thing happened, and you can name roughly when it changed.

People in the house have learned to back off when he growls, and it works every time - so he does it more.

He fixates slightly and dislikes hands moving near him, so reaching for his collar is the risky moment rather than the trigger itself.