Updated: 6 days ago
Anyone who's ever looked for a dog trainer or behaviourist may have come across a few terms being thrown about, things like "science based", "aversion free", "only kind" dog trainer or a quick google may have thrown up articles with headlines stating "dog trainer abuses dogs". Some very large UK based dog rescue organisations are very vocal about their hatred of certain types of training.
To understand what exactly they're referring too, we first must understand the 4 quadrants of operant conditioning (B. F. Skinner). The term operant conditioning simply means learning or training, and the 4 quadrants define how this learning is reinforced. This is a very simplified version that doesn't take into account lots of variables and I truly believe behaviours modification is a lot more complex, however It's a great starting point. Skinners work wasn't just for dogs, his experiments started with mice and is relevant to all living things, including us! Outlined below are the 4 quadrants.
Positive Reinforcement - behaviour causes something positive to happen
Perhaps the most commonly used technique, a lot of the time without even realising it. It is the most simplest to understand, it rewards the dog for a behaviour.
Dog does desired behaviour = reward.
This reinforces that when you do x, you get y. Happy dog, happy owner....right? If only life was that simple. It's important to note that rewards and reinforcers can come in many forms and your dog could be finding the behaviour more rewarding than the piece of chicken you're offering him.
Negative Reinforcement - behaviour causes an unpleasant thing to stop
Simply put, negative reinforcement is the term used to describe when something unpleasant stops as a result of the dog doing something. A simple example of this can be putting pressure on the dogs behind to guide them into a sit. The reinforcement part is the hand being removed from the dogs bum.
Dog does desired behaviour = unpleasant thing stops
A more complex example would be using lead pressure to ask for commands such as sit or walk to heel. Dog pulls forward, pressure is applied to the collar/neck and released when dog stops pulling, reinforcing no pulling means no pressure.
Again, it's very important to note, that dogs decide what they find pleasant and unpleasant. For some dogs, chasing after a ball for 4 days straight is Disney land, for others, they may as well spit in your face when you ask them to fetch. The amount of lead pressure applied to one dog in one scenario doesn't necessarily mean the same can be applied to all dogs.
Positive punishment - behaviour causes something unpleasant to happen
This is where things start to get a little confusing. Positive here is used to describe the addition of something following a behaviour, punishment meaning something the dog finds unpleasant, in order to decrease the likelihood that the dog will do the behaviour again. Remember, your dog will decide what they find unpleasant and punishing. The word punishment is defined as "the infliction or imposition of a penalty as retribution for an offence". An example of this is a pull on the lead when the dog breaks heel position. In order for punishment to be effective, the dog has to find the punisher sufficient enough to stop the behaviour you don't want vs how badly they want to do the behaviour. For example, a lead correction in the presence of sheep running will unlikely have the desired effect to interrupt the dogs perceived need to savage the lambs but will be effective in stopping your dog from pulling to a lamppost to pee on it.
Dog does unwanted behaviour = unpleasant thing starts
This type of reinforcement is often very difficult to get right and timing is everything when delivering positive punishment. A study done by Richard Solomon on punishing a dog found a very interesting correlation between timing of the unpleasant thing starting in how effective it is as a reinforcer. In short, the experiment involved allowing dogs access to a room with 2 bowls of food in, 1 with dry kibble and 1 with fresh meat. When the dogs approached the bowl with the fresh meat in, they were smacked on the nose with a newspaper (please do not try this at home!). Three different groups of dogs were smacked across the nose at different times, group 1 immediately as their nose reached the bowl, the second group 5 seconds after and the third group 15 seconds after they started eating the fresh meat. The study found that when group 3 were given free access to the meat bowl, they waited only 30 seconds before tucking into the meat, where as group 1 waited 2 whole weeks! This proves an incredibly important point when attempting to use this reinforcer and highlights really clearly why telling a dog off for chewing the carpet when you get home from work has absolutely no affect on his behaviour and will only affect the relationship you and your dog have.
Are you still with me? Positive punishment techniques absolutely rely heavily on the skill of the trainer/owner and this is where the issue lies, not in the reinforcement technique itself.
Negative Punishment - behaviour causes a positive thing to stop
This may also cause a lot of confusion because of the words used to describe it, however to reinforce what you want from the dog by negative punishment, you must remove something the dog wants. For example, a "time out" when a dog is becoming too boisterous or nipping. By removing the social reward, your presence, you reinforce the notion that if dog wants to be around you, he must behave in a certain way.
Dog does unwanted behaviour = positive thing stops
As owners, we often end up using a range of techniques that fall into all the above categories without even realising it. When we understand how a dog learns and the affects of different techniques can have on our dogs behaviour and their relationship to us we can make informed decisions on which techniques to use in which scenarios. The individual differences of each dog, the scenario they are in, the desired outcome and the motivators and stressors around them must be taken into consideration when deciding which technique to use. Every dog and goal must be taken at their own merit and training adapted for that scenario.
In order to teach our dogs, we need to teach them the whole picture. You can't learn a new skill without being told where you're going wrong, too. When the quadrants and tools are used incorrectly is when issues arise and even your "no pull" harness can cause damage and provide aversive affects on the dog. The phrase train the handler, not the dog, comes to mind. To be truly "force free", an entirely hands off approach would need to be adopted, which unless you live on 20 acres of fenced land and have a team of people to do the everyday jobs you need to do, is just not realistic - so all trainers use "force" of some form. The narrative that only positive reinforcement should be used to train all dogs is misinformation and dogs are being put to sleep rather than told that they're behaving incorrectly. This is beyond wrong.
I believe in using a range of techniques and all quadrants when training dogs.