The most important thing first: how do dogs learn?

We need to spend a little time tackling the theory before we can start with meaningful training. It's like getting your driver's licence: in addition to practical skills, you also need to learn the road traffic rules. The thing to remember in the process is that the basic laws of learning are the same for all vertebrates with brains.

Classical conditioning: the bell rings, the dog salivates

Let's start with classical conditioning. Russian doctor Ivan P. Pavlov discovered this by accident: he noticed that the sight of food alone triggers the flow of saliva (reflex). If the trigger of the reflex (flow of saliva) is combined with another stimulus – bell-ringing, in this case – the stimulus alone will trigger the salivation, if it is repeated often enough. The dog salivates when the bell rings, even though there's no food nearby.

In simple terms, this means that a stimulus that was neutral before receives another meaning – but no new behaviour is learned. We're all familiar with this from everyday situations: anyone who has had unpleasant experiences at the dentist's only needs to hear the sound of a dental drill to start trembling and getting goosebumps. Anyone who works on a computer a lot knows that a certain acoustic signal announces a new e-mail. There are so many things that we learn unconsciously and associate with feelings without being able to do anything about it. This is precisely the problem with classical conditioning: we do not have it under our control, so it's hard to change it. One example of classical conditioning in dogs is quite simply the doorbell. We never consciously taught our dogs that they have to bark when the doorbell rings, and definitely not that they're supposed to race to the door when they hear it. At first, when the dog hasn't yet learned what happens after the bell, he may just come to the door with you (and this is where he learns that you go to the door after the noise). Once at the door, you attempt to gain some sort of dominance over him. Many visitors also think it's cute when a puppy scampers up to them. So, for the dog this means: ringing = fun at the door is on the cards. With enough repetition – barking is added as the situation develops – we have unconsciously taught the dog that ringing = barking. It works exactly the same way with the lead that we take off the hook when we're about to go for a walk and the dog starts jumping about. You're sure to know many more examples for your dog.

Operant conditioning: dogs that can run backwards

So, no new behaviour is learned with classical conditioning. Then how come dogs can do tricks – even run backwards – or simply learn how to walk on a lead? This can be explained by operant conditioning. Operant conditioning involves learning by trial and error. It has no recognisable direct triggers, but it does bring about a reaction in the environment. There's a phrase that you should note here: behaviour is determined by its consequence. We've been talking about behaviour all this time: so, what is it? Roughly speaking, behaviour is everything that we do. Sitting, lying, standing and walking are forms of behaviour. Our dogs can do this, too, of course, and lots more – but they only have control over forms of behaviour that belong to their natural behaviour repertoire and are physically possible. These were the findings of Burrhus Frederic Skinner, who performed research on pigeons, rats and other animals back in the 1920s. He coined the term 'operant conditioning'.

So, what happens with operant conditioning? If you want to teach a dog a new behaviour on command, you have two options: reinforcement or punishment. Reinforcements are things (stimuli) that increase the frequency of the shown behaviour after the behaviour: the dog sits down and gets a treat. Correspondingly, punishment is the opposite: the dog barks and gets a clip on the nose. There are two types of reinforcements and two types of punishment: positive and negative; but please view the terms 'positive' and 'negative' in the mathematical sense: positive = add something, negative = take something away. They're not meant to be regarded as 'good' and 'bad'.

Positive reinforcement means that something pleasant is added after the behaviour. The frequency of the behaviour rises if there is a positive consequence. If the dog comes to me and gets a treat or strokes, he'll come to me more often.

Negative reinforcement means that something unpleasant is removed in the process. For example, there are people who teach a dog how to sit using strangulation by lead. They tug on the lead that's attached to the collar until the dog sits down. The lead is slackened immediately – being able to breathe is reward enough for the dog. This is a method that is neither dog-friendly nor useful: negative reinforcement encourages avoidance learning. Learning is no fun this way. It makes working with the dog off the lead virtually impossible.

With positive punishment, again, something is added – something unpleasant. If my dog barks, I can interrupt him by giving him a clip on the muzzle. I have added something unpleasant.

Negative punishment is punishment by removing pleasant things: what can I take away from my dog that's pleasant for him? His ball, my company, social contact. If my dog jumps about around me, I turn away and wait until he behaves properly. Once he's got all four paws on the ground, he'll get his strokes. With this example, I have also explained to you how operant learning works: one form of behaviour gets the dog nowhere because he's getting no feedback from the human. However, once he's got all four paws on the ground he gets what he's been after (positive reinforcement) – my affection.

What are reinforcements? There are two types: primary and secondary. Primary reinforcements satisfy biological needs: food or social contact, for example. Secondary reinforcements come about through combining them (by means of classical conditioning, for example) with primary reinforcements – the 'click' in clicker training (see chapter, 'Clicker training') = food for the dog. Money is a secondary reinforcement for us humans. Secondary reinforcements can be put to use considerably more easily and accurately than primary reinforcements.

Our training method in this book will be based only on positive reinforcement and negative punishment/ignoring.

Swiftness is the name of the game

Now, if you want to reward your dog for good work – when practising the 'sit', for example – you need to be pretty quick for one thing, and for another thing the reward needs to be right for the dog.

Quick means really quick: 0.5 seconds is the ideal time, so that the dog can associate the food with what he's done. This is why working with a secondary reinforcement (the clicker, for example) makes communication between human and dog so much easier. I've explained the topic of rewarding specifically in the chapter, 'Rewards – your dog's wages'.

So, for you this means that you need to feed the treat, throw the ball or lavish the praise extremely swiftly. Five seconds later, once your dog has stood up again, praise for the sitting will get you nowhere: this way you're praising your dog for standing up and not for sitting. People often give...