Dog training

There are two quite different things called by this name. One is the practical business of getting a dog to a state where it can live in a house, walk down a street and meet other dogs without anybody having a bad time. The other is a hobby in the full sense: clubs, syllabuses, competitive disciplines, a technical literature, and people who have been doing it for thirty years and are still learning.

Almost everybody arrives at the first and a substantial minority discover the second. What connects them is that both are built on the same small set of ideas about how animals learn, and that those ideas are considerably less obvious than they look. The single most common realisation among people who get seriously interested is that they had spent years accidentally training the opposite of what they intended.

The other thing worth knowing at the outset is that this is a field with a genuine, unresolved and quite bitter methodological argument running through it, and that the argument is about ethics rather than about what works. A page that presented the technique without naming that would be misleading by omission, so this one names it early.

Session length

Five to ten minutesShort and frequent beats long and occasional, by a wide margin.

What decides success

Consistency across the householdMore than method, more than talent, more than the dog.

The legal frame

A statutory duty of careFive welfare needs under the Animal Welfare Act 2006.

The profession

Not a protected titleAnybody may call themselves a dog trainer. This matters more than it sounds.

What actually matters, in rough order:

1. 1

   Step 1: Get the household to agree the rules first

   Written down, before the dog arrives if possible. Is the sofa allowed? Who feeds? What is the recall word? Inconsistency between people is a bigger obstacle than any training problem.

   You will know it worked when: One set of rules rather than four.

2. 2

   Step 2: Start with the name and a marker

   A name that reliably gets attention, and a marker word or click that predicts food. These two underpin everything else and take a few days.

   You will know it worked when: You can get the dog’s attention and mark a moment.

3. 3

   Step 3: Spend the socialisation window on experience, not obedience

   For a puppy, controlled positive exposure to people, surfaces, sounds, traffic and other dogs is time-limited and far more consequential than whether it can sit. Sit can be taught at any age; a fear acquired at fourteen weeks can take years.

   You will know it worked when: A dog that finds the world unremarkable.

4. 4

   Step 4: Train recall as the priority behaviour

   More effort than anything else, on a long line, and never punished on return. It is the behaviour with legal consequences, the one that keeps a dog safe, and the one most reliably destroyed by handlers who call their dog and then do something the dog dislikes.

   You will know it worked when: A recall that works at distance and under distraction.

5. 5

   Step 5: Find a class, and watch one before booking

   A good class is calm, uses food or toys, and has an instructor who adjusts to individual dogs. Walk out of one where dogs are being physically corrected or where a struggling dog is being pushed through.

   You will know it worked when: A class you and the dog both leave in a good mood.

6. 6

   Step 6: Train in short sessions, several times a day

   Five minutes before a meal, two minutes in the garden, one minute in the hall. The total matters far less than the frequency, and long sessions actively make things worse.

   You will know it worked when: Training as a habit rather than an event.

Three legal points bear directly on training, and the first of them frames everything else.

The third point, and some related duties worth knowing:

• Control in public — Separate legislation makes it an offence to allow a dog to be dangerously out of control, including on private property. This is the legal backdrop to recall training, and it is why recall is the one behaviour worth spending disproportionate effort on.
• Livestock — Worrying livestock is a specific offence and farmers have particular rights in relation to dogs among their animals. This is a countryside-access matter as much as a training one.
• The trainer is not regulated — There is no protected title and no licensing requirement for a dog trainer or behaviourist in the United Kingdom. Voluntary registers of assessed practitioners exist and are the practical answer, but nothing prevents anybody from advertising.

The genuinely useful list is short:

• A well-fitted harness and a plain lead — A harness that does not restrict the shoulder, and a fixed-length lead of about two metres. This is most of the equipment question answered.
• A long line — Five to ten metres, attached to the harness rather than the collar. The single most useful training tool there is: it makes recall training possible in real environments without risking the dog.
• A marker — A clicker at £3, or a word, or a visual signal. The device does not matter; the precision does.
• Food worth working for — Small, soft, quick to eat. The commonest reason a dog "is not food motivated" is that the food on offer is dry kibble in a distracting environment.
• A treat pouch — Unfashionable and genuinely useful. Reinforcement that takes ten seconds to get out of a pocket arrives too late to mean anything.
• Something to work on — A mat or bed used as a target. Teaching a dog to settle on a mat is the single most transferable behaviour in domestic life.

What is worth naming explicitly is the category of equipment sold to stop behaviour rather than to teach it: devices that tighten, pinch, spray, vibrate or shock. Our position on the shock end of that range is stated above and is the veterinary profession’s. On the rest, the more useful observation is structural — every one of them is marketed as a solution to a problem that training addresses at the cause, and buying one is usually a decision to manage a symptom for the life of the dog.

Retractable leads deserve a specific mention because they are ubiquitous and they work against training rather than merely being neutral: they teach a dog that pulling extends the lead, which is precisely the behaviour most owners are trying to remove.

Nearly all of the technique reduces to a small number of principles, and understanding them is what turns training from a set of tricks into something you can reason about when the tricks stop working.

• Behaviour that is reinforced increases — Anything the dog does that produces something good happens more. This is doing work every waking minute, not only during training — a dog that barks and gets attention has been trained to bark, by you, on a schedule you did not design.
• Timing is almost everything — Reinforcement has to arrive close enough to the behaviour for the dog to connect the two. A second late and you have reinforced whatever came next, which is why markers exist.
• A marker buys you time — A click or a spoken word delivered at the exact moment of the behaviour, followed by food. The marker says "that, there" and lets the food arrive a few seconds later without confusing the message.
• Behaviour is context-specific — A dog that sits perfectly in the kitchen has learned to sit in the kitchen. Generalising a behaviour to the garden, the street and a park full of other dogs is a separate piece of work and is the step people skip.
• Arousal and stress break everything — A dog over threshold — too excited, too frightened, too close to whatever it is reacting to — cannot learn. Most training failures at the difficult end are not failures of technique but of distance and timing.
• You cannot train an unmet need away — A dog without adequate exercise, mental stimulation, sleep or company will produce behaviour that no amount of training removes, because the training is fighting the need rather than addressing it.

The practical technique that follows from all of that is unglamorous. Break the goal into pieces small enough that the dog succeeds almost every time. Reinforce the piece until it is fluent. Add difficulty in one dimension at a time — duration, distance or distraction, never two together. And when it falls apart, go back two steps rather than repeating the failed one, which is the single most useful habit in the hobby and the one that feels most like giving up.

Broadly, one tradition trains by arranging for good things to follow behaviour you want, and by managing the environment so unwanted behaviour is not rehearsed. Another tradition also uses aversive consequences — physical corrections, pressure, or devices that deliver a shock, spray or sound — to suppress unwanted behaviour. A great deal of practice sits somewhere between.

It is important to be clear about what the disagreement is and is not. It is not primarily an argument about whether aversive methods change behaviour; they can. It is an argument about what else they do — to the animal’s welfare, to its stress, and to its relationship with the handler — and about whether an effect on behaviour justifies that. That is an ethical question, and the reason it is not settled by evidence alone is that people weigh the ethics differently.

One further practical point, separate from the ethics. Aversive methods suppress behaviour without necessarily changing the emotion underneath it, and a dog that has learned not to growl has not learned not to be frightened — it has learned not to warn you. That is a specific and well-known risk with corrections applied to fear-based behaviour, and it is the reason the professional advice on aggression and reactivity is so consistently to work with somebody qualified rather than to follow anything found online, including this.

Because "dog trainer" and "behaviourist" are not protected titles in the United Kingdom, the qualification question is genuinely difficult and is the practical problem most owners face at the point they need help most. Anybody may advertise, and confident marketing correlates poorly with competence.

What actually helps:

• Look for an assessed register rather than a certificate — Several voluntary bodies assess practitioners against published standards and maintain registers. Membership of one is a meaningful signal; a certificate from a course somebody paid for is not.
• Ask what happens when a dog gets it wrong — The most revealing single question. An answer about making the exercise easier is a good sign; an answer about corrections tells you which tradition you are in.
• Go through your vet for anything behavioural — Pain and illness produce behaviour changes that look like training problems, and a great many behaviour cases turn out to have a medical component. A vet referral is also how you reach a clinical behaviourist.
• Watch a class before booking a course — Reputable trainers expect this. Watch the dogs rather than the instructor: relaxed dogs and loose leads tell you more than any qualification list.
• Be wary of guarantees and of speed — Behaviour change is not a service that can be guaranteed, and anybody promising a fixed outcome in a fixed number of sessions is describing suppression rather than learning.

The competitive and club side is the part most people do not know exists, and it is where training stops being a chore and becomes a pursuit. Nearly all of it is open to crossbreeds as well as pedigrees, and a great deal of it is run by volunteers at village-hall scale.

• Agility — Timed runs over jumps, tunnels and contact equipment, handled at a distance. The most visible dog sport and the most physically demanding for both ends of the lead.
• Obedience — Precision heelwork, retrieves, stays and scent discrimination, judged against a standard. Technically the deepest of the disciplines and the least forgiving of sloppy foundations.
• Rally — A course of numbered signs each specifying a manoeuvre, worked as a continuous run. Widely regarded as the friendliest competitive entry point.
• Scentwork and tracking — Searching for a target odour or following a laid trail. Low-impact, suits older and less mobile dogs and handlers, and unusually satisfying for both.
• Working trials and gundog work — Longer-established disciplines combining control, agility and nosework, with their own traditions and their own long-running arguments about method.
• Canicross and dog-powered sport — Running, biking or scootering with a dog in harness. A sport for the handler as much as the dog, and one of the fastest-growing.
• Trick training and freestyle — The purest expression of the training skill itself, with no equipment and no venue, and a good deal of it done indoors on a wet afternoon.