Field Notes · An evolving learning library from Angier K9

The method · Chapter 8

A fair correction

Asked the question every owner asks (will correcting him make him frightened?), a trainer makes up a punishment he would find unbearable and imagines it happening to himself. He is ticketed for driving under the limit, a few times, with no reason given. He says he would be afraid to get in his car. That, he tells the family, is the whole difference, and it is all they have to get right.

because it's if i'm not speeding i don't get a speeding ticket if i had been driving 25 miles an hour in a 40 or in a 35 it's well under the speed limit but not too far under

Kelton · Watch at 17:29 ↗

The chapter in one paragraph

The trainers answer 'what makes a correction fair' with conditions the handler has to meet, not promises about the dog. The dog has to be able to tell what caused it and to avoid it next time. It lands at the moment of the act, and it is big enough to change his mind without being bigger or longer than what it answers. They make the case through speeding tickets, and then turn it round on themselves. Ticketed for speeding, you stop speeding and you are not afraid to drive. Ticketed at random for driving under the limit, you would be afraid to get in the car. Fear, on this account, comes from penalties you can't predict or avoid, not from penalties as such. From that come the exclusions: no correction for what you didn't ask, what you invited, what a stranger started, a reflex, an honest attempt, or a moment already past. Then comes the handler's control over what the dog blames: the word first, never while two dogs are face to face, in a neutral manner, and never in the places he has to live. A correction can be given length instead of force, and it ends in a release and a making-up, not in a command. The trainers count pressure over a whole life. One clear consequence on a command the dog knows and is refusing can cost him less than months of nagging, and a dog nobody corrects gets a smaller life, not a kinder one. They concede the price out loud. After a hard rep the dog gets back to ordinary before anything else is asked. And under all of it sits the claim they make as an ethical one: short of cruelty, the worst thing you can do to a dog is be unclear.

Principles

A correction is fair when it is clear, contingent on something the dog controls, easily avoidable, commensurate with what happened, and not overkill. The trainers argue it through a speeding ticket, and then turn it round. Penalised for speeding, you stop speeding and you are not afraid to drive. Penalised at random under the limit, you would be afraid of the car. What frightens is the penalty you can't predict or avoid. You know a correction worked by what the dog does next.

Said as a rule on film
On film

it's clear it's contingent it's something that's easily avoidable and it was commensurate with the thing that actually happened right and it wasn't overkill

Kelton · Bun · Watch at 32:24 ↗

so it's not like that because it's contingent and avoidable so that's all you got to do with penalties with dogs is if you penalize them and they understand why and then they learn how to avoid it they're cool everything's fine they're not afraid

Kelton · Zeus · Watch at 18:20 ↗

they have to be contingent on things we control yeah otherwise it just creates psychological trauma right if we don't know why it happened and we can't avoid it

Kelton · Kodak · Watch at 22:33 ↗

same thing with dogs if they understand why they don't mind they're like they're not afraid to be around you or to do anything else

Kelton · Roxy · Watch at 17:09 ↗

do it um the question is how do you know if it worked um you know it worked by what the dog does

Peque (PK) · Watch at 23:52 ↗

Concepts: Contingent, avoidable, and the right size

Exceptions

  • The trainers set two limits of their own before any technique. Kelton: 'Whatever feels ethical to you and whatever uh is not going to injure him. Those are the two boundaries that I have' (King, watch 29:16). 'There's so many ways to punish that don't have the potential for injury do those' (Zeus, watch 25:28).

Pressure is already in a dog's life, whether anyone chose it or not. A dog pulls, the lead stays tight all the way, and at the park the pressure comes off while the owner looks at a phone. That has taught him, hundreds of times, that pulling ends pressure. So the question is never whether to use pressure. It is whether the handler authors it, clearly and contingently, or leaves it to chance.

Said as a rule on film
On film

say they pull on the leash pull pull pull and then when they get to the thing they want the pressure turns off because you stand there looking at the phone

Kelton · Lola · Watch at 23:32 ↗

non-contingent that one too if you do nothing then yeah that's even worse or non-contingent pressure pressure that they have no way to avoid because they don't understand why the pressure's happening if it's unclear then that's majorly problematic

Kelton · Lola · Watch at 29:44 ↗

Concepts: Contingent, avoidable, and the right size

The fairness test rules things out. Don't correct what you never asked for. Don't punish excitement you caused, or something a stranger started. Don't punish a reflex, or an honest attempt that just missed, and don't correct a moment already past or out of range.

Said as a rule on film
On film

If I had said like let's go she didn't pick up. She didn't listen then that's a different story because that's a direct like disobedience to a command which is not okay.

Connor · Luna · Watch at 48:59 ↗

i mean you didn't give him a command so like because you didn't say plots right he just kind of auto did it so don't correct him if you didn't ask him to do it yeah because then he'll get confused

Savage · Watch at 21:32 ↗

the dude did engage with him of course so i'm not like gonna make that like a punishment

Deuce · Watch at 6:15 ↗

you don't have to do a lot of yanking here for um for something so simple as you know she was trying and she just kind of missed um because she got distracted so just a little tiny tug there

Bubba Gum · Watch at 13:08 ↗

this would be way way too late for me to correct it so we're all the way at the truck over here now

Apollo (the Sharkie) · Watch at 19:36 ↗

Concepts: Contingent, avoidable, and the right size

Exceptions

  • Behaviours the family has decided are off limits (jumping, the counter, mouthing the other dog) are stopped with 'nope' even when no command was given. In the calm-liberty window of the four-windows map the trainers 'free shape with punishment' (King, watch 26:09). The exclusion is for commands never given and things never forbidden, such as a dog sniffing when nothing was asked. Where exactly the line falls between the two is an open question.
  • Credit for time served: a dog who has held place for forty-five minutes and stands to stretch is put back without heat (King, watch 11:04).

A correction teaches whatever the dog links it to, and the handler decides what that is. The word comes first, so he knows it came from the person. Don't correct while two dogs are face to face, or the other dog gets the blame. Keep your manner neutral, so he puts it down to his own act and not to you. Never make the places he has to live in, such as his crate, into places of punishment.

Said as a rule on film
On film

So, in this case, if you just hit the e-collar without anything while Cosmo's doing his, like, biting Simba, what could happen is he could think Simba did that.

Connor · Cosmo · Watch at 12:01 ↗

When you say the word and then the e-collar comes after, there he knows it was you not simba

Connor · Cosmo · Watch at 12:20 ↗

Yeah, I should have, if I had done it when he was sniffing, it probably would have been fine, but I did it when they were close, face to face. So he took it as coming from him, yeah.

Kelton · Murphy · Watch at 41:31 ↗

so the big thing is just don't be mean when you do it right don't really show any emotion in general because i don't want him to think it's coming from me i want him to think that's just bad behavior that doesn't work for him

Chunk · Watch at 5:50 ↗

if you e-collar everything they run up to them to the car then you create a different thing it's not neutrality it's fear

Simba · Watch at 13:15 ↗

that's happened to him so we can't send him to the crate for punishment we can't act like these places are punishments for him we don't do time out with the dogs in these cases that's a very very very very bad idea I want to emphasize that

Kelton · Bo · Watch at 9:52 ↗

Concepts: Control what the dog blames

Exceptions

  • Bo's crate history is Kelton's theory of what happened at home, stated as a theory, not the family's account (watch 02:53).
  • Whether the place cot is ever a time-out spot is not settled on film. Sputnik's go-home uses it that way, and Bo's lesson says 'we don't do time out'.

Intensity is only one dial. A correction can be made to mean more by making it last, without raising the level. The length is carried by the handler's voice: firm, then disappointed, then stop. No command comes straight after, because that turns punishment into a task. No instant 'good boy' either, because that blunts it. It ends with 'okay' and a making-up, so the dog learns not to do that again without learning to fear the people.

Said as a rule on film
On film

anything else right um and then even like um if i punish and then uh i want to make it more meaningful i don't only have the option of raising the intensity i can also make it last longer but not raise the intensity

Kelton · Max (Goldendoodle) · Watch at 59:36 ↗

the only way to increase the severity of it okay right another is duration not duration of holding the e-collar down no just how long does the punishment event take right he has to stay there right right and i don't even tell him a command because then it switches

Kelton · Thor · Watch at 31:04 ↗

authoritative but what did you notice towards the like last the last few little boops that i did my tone changed i started sounding disappointed

Kelton · Sputnik · Watch at 11:27 ↗

and then the other thing that i did was i made it last a few seconds yeah so a lot of times that's a mistake people have make they just kind of like we'll get on them and then immediately go but you're a good boy did you do and then that just blunts the whole thing right

Kelton · Rocky · Watch at 51:58 ↗

oh no i let him sit in like discomfort for a second and then i go okay and we reconcile and everything's cool because then he learns okay like don't do that thing again but everything's fine

Kelton · Rocky · Watch at 52:16 ↗

that might be the problem right so the way that you know that you penalize the dog you know you punish them properly like they the message was received not if they stopped doing it okay but if they look like they got punished

Kelton · Thor · Watch at 18:26 ↗

Concepts: A correction has length as well as force

Exceptions

  • 'One sentence and done' (Roxy) sits against 'until a complete stop, then disappointed mode' (Murphy), both from Kelton. See conflicts.
  • Every instance on film is Kelton's. No other trainer states it.

The trainers count pressure over the dog's whole life, not the single rep. A small, unchanging consequence becomes a fine the dog is happy to pay, and it never ends the argument. One meaningful consequence, always tied to the same act, can get close to single-event learning, and after it much lower levels do. The conditions are stated just as firmly. It is only for a command the dog demonstrably knows and is refusing. The same pressure on something he doesn't know 'is just gonna send him into a panic'. And it ends in release and decompression.

Said as a rule on film
On film

we're looking at all the reps that the dog ever does every time the dog ever feels the e-collar for the rest of its life we want to make that cumulative total lower

Louie · Watch at 13:46 ↗

using that same pressure for things that he doesn't know is just gonna send him into a panic, right? So I'm gonna turn it way down

Louie · Watch at 20:37 ↗

If they're driving, like, recklessly, say they're, like, hitting 100 miles an hour flying by you, a cop pulls them over, and it's the same fine every time, $10.

Kelton · Sputnik · Watch at 6:12 ↗

and so we just did like beeping or um we vibrated or and it worked okay and he stopped doing it but then he did he just keeps doing it though they just still do it like 20 times a day

Kelton · Murphy · Watch at 1:02:03 ↗

to punish him nearly as severely as the first time, because I punished him severely the first time. He was immediately like, oh crap, I made a mistake. And then the last time that I gave the opportunity, he didn't even take it.

Kelton · Zeus · Watch at 4:20 ↗

that's where the authority happens right but then right after that you want decompression

Marvin (Giant Schnauzer) · Watch at 10:58 ↗

Concepts: The lifetime pressure budget

Exceptions

  • The severe correction on Zeus was for going at the resident dog, and the curve is tracked on film: severe, then much less, then nothing needed. Nash's single long, strong correction was aimed at charging at other dogs. His trainer says that done that way it has come 'pretty dang close to single event learning' in his experience (watch 13:55). No footage follows a dog long enough to test the lifetime-total claim.
  • Making a dog afraid of a whole category of thing is named as a last resort, for life-or-death cases only. Connor gives the example of a car-chasing Doberman who had already been hit (Deuce, watch 00:18).
  • Severity is budgeted by what the behaviour costs the household the dog goes home to (chapter 7, severity-follows-the-stakes).

A dog nobody corrects does not get a kinder life. He gets a smaller one: the walks stop, he is put away when people come, the park is avoided, and the house is run 'like a cell block' of gates and rotas. The family often ends up resenting him. The trainers concede the price first and in plain words. Then they set it out as a choice they believe the dog would make: a handful of uncomfortable moments he can see coming and avoid, for a life he is allowed to live.

Said as a rule on film
On film

in a way they understand but over time you start to resent them because they're annoying you constantly they destroy things in the house so you put them in the crate they pull really hard on walks so you stop taking my walks and their life gets smaller and smaller and smaller and smaller and smaller and

Kelton · King · Watch at 50:17 ↗

life is full and fun and big but they feel some shocks here and there when they do something stupid your dog would choose option B every single time I would choose option B might punish me but make my life bigger

Kelton · King · Watch at 52:58 ↗

uncomfortable for you know you know one second up to like 45 seconds is very very uncomfortable but I get to go all over the place I can go anywhere versus don't punish me severely but my life is really small if it were me I'm picking the first option every time

Kelton · Zeus · Watch at 45:15 ↗

instead but it makes your dog's life a lot better too so they are uncomfortable a few times right because yeah I mean you shock them

Bullet · Watch at 18:34 ↗

that's what management does when you have to use management because your dog isn't trained you make their life uber small you can't trust them to do anything and so you're constantly restraining them managing them putting them in a tiny box

Kelton · Kodak · Watch at 25:44 ↗

because then his world doesn't end up being cuz I've seen it happen too many times is where like somebody will get a dog and they love the dog and then over time they end up resenting them.

Kelton · Bo · Watch at 47:02 ↗

Concepts: A life that gets smaller, and the bargain

Exceptions

  • Not every dog is offered the bargain yet. With a frustrated, nervous dog the trainers hold the pressure back, because pressure there could produce a bite, and the bite would end the life the bargain was meant to buy (Lola, watch 28:43). What decides which dog gets the bargain and which gets the wait is open.
  • Some management is acceptable. A closed bedroom door is fine, and rotating dogs like a prison guard is not (Murphy, watch 13:38 onward). No rule for the line is stated.
  • An owner who would rather keep the dog in the back yard than ever see him uncomfortable is given that as a legitimate choice. It is the small life, and it is named as such.

After a hard rep, stop asking and let the dog get back to ordinary before asking again: walking, sniffing, running together, play. Go back in too soon and the two tries become one long stretch of pressure. Then the lesson the dog takes is that complying doesn't end the stress. The handler's manner straight afterwards is half of the rep: correct, then be pleasant, with nothing left over.

Said as a rule on film
On film

one of the ways that's the most reliable test learning is to do a second trial but before you do that second trial you need to let the dog return to baseline right

Finn · Watch at 14:25 ↗

instead of him learning from the first rep, he's going to be still in the first rep psychologically. He's still going to be a little bit stressed. So he never reset to neutral.

Louie · Watch at 23:39 ↗

Because if I go back to a command before they hit baseline, it's kind of like if I'm like, alright, new math problem, you'd be like, oh my god, because you didn't have time to decompress at all.

Sage · Watch at 10:30 ↗

correcting her but after you do it make sure it's really fun and it's really like there's no like residual like i hate you energy going on um because with her pressure could shut her down really fast

Lola · Watch at 9:29 ↗

happy endorphins. So let them be nice and free, let them do stuff, especially after you do a little bit of a harder correction, or a little bit of a harder negative reinforcement.

Phoebe (the Rottweiler) · Watch at 9:37 ↗

slow is fast with dog training slow is fast so um you have to let them return to normal before you ask them to do more stuff

Stormy · Watch at 28:51 ↗

so i cannot let stuff like that slide that cannot slide cannot slide but i can do a makeup with him

Wiley · Watch at 3:25 ↗

Concepts: Two events, not one - and what the handler does in between

Exceptions

  • A stopped behaviour is not a settled dog. Minutes after a hard correction Zeus is 'a little ramped up still' (watch 10:28).
  • How long is long enough is open. It runs from forty-five seconds named as too short, through four or five minutes for one hard down, to an hour's walk for a nervous dog.
  • It might be nothing, very little, the wrong lesson, or learning slowly.

Clarity is the kindness. Short of cruelty, the trainers call being unclear the worst thing you can do to a dog, because an unclear dog is frustrated for longer and ends up in a shelter or worse. For a nervous dog, a predictable handler is the treatment: structure makes the world predictable, and predictability lowers stress.

Said as a rule on film
On film

because in my mind the worst thing that I could do to a dog um other than like torturing them or something, right? Uh which would be evil. Um but the like the worst thing I can think of that I could do to a dog is not be clear because that's how dogs end up in shelters. That's how they get put down.

Kelton · Bo · Watch at 44:42 ↗

is this frustrating to the dog? Because there's two possible ways that you would be mean. One way is by being unclear. And then they're frustrated for longer because they can't figure out the puzzle.

Kelton · Sage · Watch at 27:33 ↗

then the other way is you could go way too hard yeah right and people often do the first one but they don't realize that they're prolonging this dog's lack of clarity

Kelton · Sage · Watch at 27:46 ↗

but negative reinforcement is clarifying to dogs um it makes their world predictable it makes it structured right

Kelton · Kodak · Watch at 16:34 ↗

the nervous dog right you don't give them clarity they'll freak out

Lola · Watch at 10:21 ↗

Concepts: Clarity is the kindness

Exceptions

  • The trainers agree both are errors and differ in emphasis. No session gives an owner a test for which side of the line a correction is on.

When a correction was mistimed or bigger than it needed to be, the trainer says so on camera, in front of the family, and repairs it straight away: a release, play, the level turned down, a making-up.

Shown, not said
On film

I tapped the e-collar at the wrong time.

Kelton · Murphy · Watch at 40:57 ↗

i will say that was kind of like a negative experience for him a little bit there i didn't know he was going to be so sensitive i'm sorry ramsay

Roxy · Watch at 13:44 ↗

he came out of it as soon as i i did the okay and got him to be excited and i'll see if he's

Nash · Watch at 4:41 ↗

uh they definitely get it and it's only a problem if you're constantly restricting their freedom

Ace (Chihuahua-terrier mix) · Watch at 13:21 ↗

Concepts: Control what the dog blames, Two events, not one - and what the handler does in between

Exceptions

  • No trainer states this as a rule. It is how the sessions go when something goes wrong.

What owners are told

Before you correct anything, ask three things. Did I ask for something? Does he know what he did? Does he know what to do instead? If any answer is no, the problem is clarity, not consequence. Never correct something you didn't ask for, excitement you caused, a startle, or an honest attempt. Correct at the moment it happens, or not at all. Say your word with the consequence every time, so he knows it came from you and not from the other dog. Don't correct while two dogs are nose to nose, and never make his crate a punishment. Make it enough to matter: a small fine he pays twenty times a day is not kind. Then watch what he does. If he stops and stays stopped, it worked. If he goes straight back, change it instead of repeating it. You don't have to raise the level to make it mean more. Stay with it for a few seconds in words, firm and then disappointed, with no command and no instant 'good boy'. Then 'okay', and make up with him. After a hard moment, let him sniff, walk or run before you ask again, and be pleasant yourself straight away. Your two limits: it has to feel ethical to you, and it must not be able to injure him. Weigh honestly what doing nothing costs, because the dog nobody corrects gets a smaller life.

On film

If it's easily avoidable, if they know exactly what to do to avoid it, it's clear, it's contingent on something they control, they're not scared.

Piglet · Watch at 33:30 ↗

he was like well you didn't tell her to do anything right so you're correcting her on something you didn't tell her to do now you're gonna confuse her

Sage · Watch at 12:19 ↗

an event so it's punishment's not just about the intensity it can also be about the length okay

Kelton · Rocky · Watch at 32:38 ↗

sure that the release to freedom is you know immediate and then he needs plenty of time to return to neutral

Samson · Watch at 10:26 ↗

What trainers know

The test, in the order it is usually given: contingent, avoidable, clear, commensurate, not overkill. Deliver it through the speeding ticket, then turn it round. With a nervous client, concede the trauma point outright. The limits come in this order: your conscience, then injury. Run the exclusions before every correction: no command, invited excitement, a stranger's doing, a reflex, an attempt, late or out of range. Say 'nope' with every pop, never a pop without it. Exaggerate the mistake before correcting a heel drift. Don't repeat the command as the correction. Check clarity before raising the level. Size: one big pop, not five small ones, and the correction must outweigh the desire. Judge it by the next ten to twenty seconds. For attribution: marker first, time the correction to the act and not to the dogs being close, and use a lead signal instead of the collar on a dog who is only uneasy. No collar on the approach to people or cars. Places the dog must live in are never punishment sites. For length: firm, then disappointed, then release and reconcile. Carry a drag leash at home so the owner can follow up socially. Lengths named on film run from a few seconds to about forty-five. Pressure budget: the higher level only on known, refused commands, turned down for anything still being taught. End authority events in visible submission and immediate decompression. Say out loud which dogs are not getting the bargain yet, and why. For the reset, the dog sets the interval: forty-five seconds is too short, four or five minutes is the working figure for a hard down, and ten to fifteen indoors for a nervous dog. Early sessions run about ninety per cent walking and ten per cent commands. Don't load the dog into the vehicle straight after pressure. For clarity: frame every decision as 'is this frustrating to the dog?', and when he is confused, look for a clearer picture before more pressure.

On film

yeah what you do is you exaggerate the mistake you exaggerate the mistake and

Bane · Watch at 29:10 ↗

there is he didn't react and then also the other thing is if he was feeling at all like pressured with the other dogs and i tapped the e-collar he could take it as coming from them

Kelton · Astro · Watch at 33:29 ↗

it's really really sparse throughout the session sometimes um 90 percent of the session is just walking and 10 percent of the session is commands

Kelton · Kodak · Watch at 13:42 ↗

you the opportunity to do exactly what I did that's why I put the leash on because the social pressure afterwards is just as important as using the e-collar

Kelton · Sputnik · Watch at 41:22 ↗

In the dogs

Zeus

Zeus's first lesson turns the speeding ticket round, so the test is on the handler. His aggression work tracks the lifetime curve out loud: a severe first consequence for going at the resident dog, a much smaller second one, and on the third chance he leaves her alone. Minutes later the trainer won't call him settled. The bargain is spelled out over him: a big life, or a back yard.

On film

And then the last time that I gave the opportunity, he didn't even take it. He just left her alone and didn't do anything.

Watch at 4:20 ↗

Roxy

Roxy's lessons turn the fairness test into something a household can run: the speeding tickets, the difference between one sharp sentence and being shouted at for minutes, and a correction on Ramsey the trainer judged too big, apologised for and followed with a cool-down.

On film

there's a big difference between me going like hey like louder and like being like hey like I'm mad like what the heck guys and saying that one sentence and then that's it.

Watch at 3:08 ↗

Murphy

A three-dog house with a new baby. At the go-home Kelton blames a short clash between two of the dogs on his own timing, a tap while they were face to face. Minutes later he explains why the family's gentler settings had failed: a ten-dollar fine, paid twenty times a day. Then he lays out a penalty that runs to a complete stop and then into disappointment.

On film

That was my fault.

Watch at 40:45 ↗

Sputnik

A home visit to a household where one dog harasses the others. One strong correction, and the tone that turns from firm to disappointed, pointed out to the family. The owner says it is hard to watch, and the trainer says it hurts him too.

On film

makes it hurts my heart makes me feel bad but at the same time like this is a world of a difference

Watch at 8:02 ↗

Rocky

Rocky jumps. Kelton corrects him, counts it out, says 'okay', and the two of them make up. Then he turns to the owner and explains every second of it.

On film

right but did that look meaningful to him yeah it did right because um one it was clear why

Watch at 51:41 ↗

Thor

At the go-home Kelton adds the rule that keeps the length a punishment: no command straight after, or the dog's brain switches from being punished to doing a task. You know it worked when he looks like he got the message, not merely when he stopped.

On film

the only way to increase the severity of it okay right another is duration

Watch at 31:04 ↗

Lola

Lola is the dog who complicates the chapter. Her first lesson gives the owner the pressure already in the dog's life, and her trainer insists the correction is followed at once by fun. The bargain is also withheld from her for now. The pressure that would make a pushy dog's life bigger could make a frustrated, nervous one bite.

On film

you know you he'll grind and then all of a sudden nips at like a four-year-old it's not a great look for you or your dog that's how dogs get put down all the time

Watch at 28:43 ↗

Finn

A flight risk whose recall work gives the reset its name. A trainer grades his own timing on camera. He went back too fast, and the two reps became one long stretch of pressure. He spells out the lesson Finn would have taken instead.

On film

right back to the recall. Went right back to CLME. If I did that, the lesson that he would internalize instead is that coming back does not relieve stress it does not resolve the issue instead it belongs it right

Watch at 15:50 ↗

Louie

The accounting and its safety rail, narrated in one park session with a junior staff member shadowing. Keep the lifetime total low by being clear on known commands, turn it way down for anything he doesn't know, and don't go back within a minute, or the first rep never ends.

On film

by utilizing this level on known commands that he's resisting very quickly I won't need to use this level ever again and I can use a much much lower level for reinforcement when needed

Watch at 14:53 ↗

King

A family with an ordinary pet dog and no bite history is asked to imagine their dog could talk. Option A: never corrected, crated, the walks stopped. Option B: some shocks here and there, and a full life. Kelton says which the dog would choose, and which he would. He also sets his two limits: ethical to you, and not injurious.

On film

Now people can punish their dogs any old way. Whatever feels ethical to you and whatever uh is not going to injure him. Those are the two boundaries that I have, right?

Watch at 29:16 ↗

Bo

The moral argument, made to an owner who had stopped using the tools: being unclear is the worst thing short of cruelty. The file also carries the story of how a place got blamed. Bo, by Kelton's theory, was punished in his crate hours after the fact until he defended the crate.

On film

because if you non-contingently punish a dog in the same area over and over again you know hours after the event happened he's not gonna understand why you're treating him that way and he's going to connect it to the only thing that he can see as a common denominator which in this case is what the crate

Watch at 2:53 ↗

Cosmo

Two brothers, one of whom mouths the other to keep him in place. At their go-home Connor explains why a silent correction could start a fight: Cosmo could think Simba did it.

On film

The reason I say we're going to work on that is because if you do it the wrong way, you can start a dogfight.

Watch at 8:07 ↗

Kodak

A fearful dog who treated the handler as a threat. His lessons carry the trauma line ('contingent on things we control'), the claim that structure makes a nervous dog's world predictable, and the payoff: late in a session Kodak walks close, not out of fear, but because the handler is predictable now.

On film

look at him walking close to me now um not because uh you know he's afraid to go out there because i might do something to him

Watch at 56:23 ↗

Wiley

Wiley goes after a runner on a path. The trainer says that 'cannot slide' and corrects it without ambiguity, then does a making-up straight away: sniffing, the ball, praise.

On film

so i cannot let stuff like that slide that cannot slide cannot slide but i can do a makeup with him

Watch at 3:25 ↗

Where the trainers differ

Issue. How long should a correction last: one sentence and done, or until a complete stop?
Sides.
Position. Short. One sentence and that's it. Two or three minutes of being told off is disturbing and teaches nothing.
Context. Kelton at Roxy and Ramsey's second in-program lesson.
On film

there's a big difference between me going like hey like louder and like being like hey like I'm mad like what the heck guys and saying that one sentence and then that's it. Versus, have you ever had somebody like yell at you for more than like 20 seconds?

Roxy · Watch at 3:08 ↗
Position. Until a complete stop, not a rolling one, then drawn out a little in a disappointed tone so that it weighs.
Context. Kelton at Murphy's go-home, about guarding and standing over the household's other dog.
On film

You don't want to just do a roll stop. You want a complete stop.

Kelton · Murphy · Watch at 1:03:47 ↗
Issue. How big: err small or err big?
Sides.
Position. Size it to the dog. Most dogs need far less than owners think, and an attention lapse in a trying, sensitive dog needs a tiny tug.
Context. Bubba Gum in a store.
On film

you don't have to do a lot of yanking here for um for something so simple as you know she was trying and she just kind of missed

Bubba Gum · Watch at 13:08 ↗
Position. Make it big enough to change the equation. A small fixed penalty is a fine the dog pays, and for reactivity it is safer to err high than low.
Context. Peque's follow-up in her adopters' home.
On film

more cautious to have it higher than it lower it's better if you accidentally go too high

Peque (PK) · Watch at 34:14 ↗
Issue. What does a serious behaviour sound like when you answer it: a big issue, or disappointment?
Sides.
Position. Make a big issue of it. A dangerous choice needs an unmistakable message.
Context. Bo, to the camera, on the choice window.
On film

big issue and you you want the message to be clear to them you're having a conversation a serious one

Bo · Watch at 3:07 ↗
Position. Sound disappointed, not authoritative, and keep your hand still. With a dog who growls in his crate, an authoritative correction is the fight he expects.
Context. Bo's crate work.
On film

and then just do like a like uh instead of um like an authoritative thing just do like a like just kind of disappointed like dude why are you doing that

Bo · Watch at 2:56 ↗
Issue. May the chance to get it wrong be manufactured?
Sides.
Position. Engineer it. The dog has to meet real choice points, so the trainer sets up situations (a cat, chickens) where the wrong choice can happen and be answered.
Context. Bo.
On film

if you don't practice that uh them being in that window on a regular basis they won't make the right choices because they've never had to try

Bo · Watch at 1:46 ↗
Position. Not with people. The people who happen to be in a park are fair, but staging them as distractions is not, and a stranger who starts the interaction takes the fault.
Context. Deuce on a park walk.
On film

there's some people so let's go see what he does i don't believe in using people as a distraction

Deuce · Watch at 6:29 ↗
Issue. Which dogs get the pressure bargain, and which get the wait?
Sides.
Position. Offer it. The trade is judged on the total outcome, the dog would take it, and withholding pressure is what produces the small life.
Context. King's lesson, an ordinary pet dog; Zeus, a reactive dog.
On film

life is full and fun and big but they feel some shocks here and there when they do something stupid your dog would choose option B every single time

Kelton · King · Watch at 52:58 ↗
Position. Not to this dog, not yet. Pressure on a frustrated, nervous dog can produce the bite that ends the life the bargain was meant to buy.
Context. Lola, a nervous, reactive dog.
On film

you know you he'll grind and then all of a sudden nips at like a four-year-old it's not a great look for you or your dog that's how dogs get put down all the time

Lola · Watch at 28:43 ↗
Issue. After a hard rep, how long before the next ask: minutes, or a read within minutes?
Sides.
Position. Minutes, often many. Forty-five seconds to a minute is named as too soon.
Context. Louie's park session.
On film

If I go right back to it after 20, 30 seconds, or even about 45 seconds to a minute, then instead of him learning from the first rep, he's going to be still in the first rep psychologically.

Louie · Watch at 23:33 ↗
Position. In the aggression work the rep is followed at once by release, praise and an ordinary walk, and the next read of the dog's choices is taken within minutes.
Context. Zeus with the resident dog.
On film

why he got punished last time right and now he's thinking about me

Zeus · Watch at 10:15 ↗
Issue. After a fright, say nothing, or say 'it's okay'?
Sides.
Position. Say nothing. Reassurance tells the dog the thing was worth being frightened of, and the handler's unbothered calm is the information.
Context. Ramsey's startle at a sudden noise, at the go-home.
On film

Give him no input. Yeah. Because if you Oh, it's okay, buddy. He's gonna think big deal.

Roxy · Watch at 25:37 ↗
Position. Say it, and keep saying it, while the demand stays where it is: 'I'm not going to hurt you', 'I'm not mad at you'.
Context. Stormie, a soft dog who goes flat after refusing. The handler is alone with the camera.
On film

You're here to learn. Why would I be mad at you for learning?

Stormie · Watch at 8:29 ↗

As a course unit

The learner can state the fairness test in one sentence, run the exclusions before any correction, and deliver a correction the dog can attribute to his own act. They can use length and tone as a second dial, close with a release and a making-up, and judge a correction by what the dog does next. They can explain the lifetime pressure budget and the bargain honestly, including its price, its limits and the dogs it is withheld from, and hold the reset before asking again.

Lessons, in order

  1. Clarity is the kindness
  2. Contingent, avoidable, and the right size
  3. Control what the dog blames
  4. A correction has length as well as force
  5. The lifetime pressure budget
  6. A life that gets smaller, and the bargain
  7. Two events, not one - and what the handler does in between

Drills

  • The say-it-first session, from the fairness concept: before every correction the handler says aloud what it is for and how the dog could have avoided it. If he can't say both in one breath, the correction doesn't happen. Someone counts the near-misses against the exclusions.
  • At the first sign of trouble, separate with a pop and 'nope', repeat, and only then the collar, never while the dogs are face to face. Disappointment until a complete stop, then release.
  • The length exercise, as shown for Max: correction at the working level, then without raising it, 'ah, I said no. Why did you do that? No.' Shift to a disappointed tone, no command, then 'okay' and make up.
  • The one-rep session, as run on Finn and Louie: one hard rep, release everything the instant the behaviour appears, then walk, sniff and play for four or five minutes at least. Decide whether to go again only when the dog is visibly back.
  • The small-life inventory, as done with King's and Zeus's families: list what the dog no longer does (walks at busy times, parks, visitors, gatherings), then state the trade in plain words, including the price and the two limits.

Assessment

  • Your dog has been on place for forty minutes. You call a friend over, get excited greeting them, and he gets up and jumps. What, if anything, do you correct?
  • Your two dogs start to squabble. What do you do first, and when does the collar come in?
  • You've just corrected your dog for jumping on a guest. What do you say in the next ten seconds, and what don't you say?
  • Does he know this, is he refusing it, and is it serious enough to spend real pressure on? After it, is the requirement falling?
  • An owner says, 'I'd rather keep him in the back yard than ever see him uncomfortable.' What do you tell them?
  • You asked for a down, it took a real fight, and he did it. Twenty seconds later he looks fine. Do you ask again?

The ideas it draws on

← The e-collar · When the dog argues back →