Quadrants of what now?

“Oh, I don’t want to use FOOD to train my dog. They should obey me because they WANT to.”

I have no idea how many times I have heard something like the above statement, and it puzzles me every single time. Why should your dog want to obey you? What is in it for them? (Disclaimer – I do realise that some dogs are not as motivated by food as other forms of reward, but food does work for the majority.

Toys can also work as rewards!

Let me put it another way – would you do your job for free? No, I did not think so. That being so, why would your dog? Following our directives has no real benefit for the dog a lot of the time, they would far rather be doing their own thing, so it is up to us to make it worth their while.

Time to get a little bit technical…

The use of rewards in dog training is what is called positive reinforcement, and positive reinforcement is one of four ‘quadrants of learning’ based on something called operant conditioning. There is an explanation of what operant conditioning is, with mention of Pavlov and Skinner, but to cut a long story short, it relies on the principle that actions have consequences and these consequences can be good or bad. Good consequences make a dog more likely to repeat a behaviour, bad consequences make repetition less likely.

The consequences are described as positive or negative. These words are used purely in the scientific sense in this context, so positive means to add something in, negative means to take something away. Behaviours will responded to with either punishment or reinforcement.

The quadrants sound complicated when you start looking at all of the terms individually, but they really are not so bad once you start looking at the definitions for each of the four. So, here we go – from the least humane, useful and constructive to the ones that will get results in a force free manner:

Positive punishment:

This is really bad. Like really, really bad. Undesired behaviours are met with loud shouty voices that scare the dog, or smacks, kicks, ‘corrections’ on a collar (AKA jerking the collar). Something that the dog considers aversive. So something bad (punishment) is added (positive) after the behaviour to make it less likely to happen.

Negative reinforcement:

Also very, very not good. Something painful is taken away after the dog offers a desirable behaviour. An example would be the use of a prong collar for loose leash walking. The dog does not pull in the hope that the sharp metal prongs will stop digging into their skin and hurting. Something bad (the effect of the prong collar) is taken away (negative) to reinforce the behaviour and make it more likely to be repeated.

Negative punishment:

This one is not brilliant but can be used sparingly as part of a force free programme. If the dog offers an undesired behaviour, something he likes is taken away. A toy, a play session, something that he enjoys. The nice thing being taken away (negative) is the punishment as he no longer has access to the good thing, meaning the behaviour is less likely to happen again.

Positive reinforcement:

Here we are – the good stuff. The dog does something we like, he gets something great as a reward. He sits on cue, he gets a treat. He does what we want and is rewarded (positive) meaning the behaviour is more likely to happen again (reinforcement).

Here is a picture by the incredibly talented Emma Judson, the brains behind the wonderful informative anarchy that is Foul Mouthed Fido on Facebook (Warning - Fido swears a LOT), and illustrating genius found at Cartoon ‘Ems which explains the four quadrants with amusing simplicity.

‘But what if my dog does something I don’t like? What do I do if I don’t punish him?’

Teaching an alternative behaviour is a positive way to deal with so called ‘problem’ behaviours (they are not a problem to the dog!). Rather than yelling at the dog, turn away from him. When all four paws are on the ground, cue him to sit. Reward the sit. You are on the way to coaching your dog to sit rather than jump. It really is that simple.