Two types of Learning
As a behaviourist and trainer, I follow a lot of dog training pages to keep up with new ideas and to find useful explanations of training concepts to pass on to clients. While “operant conditioning” and “the 4 quadrants” are referred to frequently by many modern trainers, what often seems to be missing is any reference to or understanding of another, equally important, type of learning: “CLASSICAL CONDITIONING” or “PAVLOVIAN CONDITIONING”. So, I thought it was time to write an article explaining what both types of learning (conditioning) involve and why this other type of learning cannot be ignored, as it so often is.
The study of how animals learn is vital for anyone who wishes to work with dogs or any other animals. Without an understanding of how animals learn, you cannot hope to train or change behaviour.
While there are many categories and subcategories when it comes to learning, there is one broad type of learning that is fundamental to understanding how behaviour develops and changes: ASSOCIATIVE LEARNING. Associative learning is what it sounds like – learning associations.
Associative learning is divided into Type 1 and Type 2. Many trainers focus just on Type 2: trial and error learning or operant conditioning. A basic definition of this would be “learning that behaviour has consequences”. In other words, this is how an animal learns that his own behaviour can make stuff happen or stop happening.
“The 4 quadrants” which are often mentioned by trainers, refer to operant conditioning and the 4 possible scenarios that can happen in this type of learning:
At this point it is important to mention that while many people call themselves “positive trainers”, what this often-misunderstood term is actually referring to is “positive reinforcement” i.e. trainers who rely on the first quadrant of learning, described above, to train dogs. Positive reinforcement trainers focus on reinforcing desired behaviour by “adding good stuff” whenever a dog performs the behaviour. They avoid using positive punishment and negative reinforcement, because both of these require adding something unpleasant (fear & pain) to get results (although negative reinforcement results in feelings of relief, it is relief from the positive punishment that is first applied, so goes hand in hand with positive punishment).
It must also be pointed out though, that by default, positive reinforcement training does involve a certain amount of negative punishment (removal of good stuff or withholding good stuff). This is only logical: if you are teaching “sit stay” you will reward the dog for staying in a sit position, but you will not reward the dog if they get up. If you are rewarding a dog for sitting politely when guests arrive (with attention from those guests), you will not reward the dog if they jump all over the guests (you will ask guests to withhold attention). However, GOOD positive reinforcement trainers will as far as possible SET A DOG UP TO SUCCEED to avoid the dog constantly failing and becoming frustrated and despondent.
Acknowledging the emotions that result during learning, brings us on to the other type of associative learning and the focus of this article. Classical or Pavlovian conditioning (Type 1 Conditioning) is when an animal learns which signals in the environment are associated with particular events in the environment. This has nothing to do with the animal’s own behaviour – it is simply figuring out what various smells, sights, sounds and situations predict. This type of learning is named after the scientist Pavlov, who found that consistently and repeatedly sounding a tone before feeding time, resulted in the dogs being fed drooling as soon as they heard the tone, even if no food was present. This type of learning happens all the time and involves subconscious, involuntary responses or reflexive behaviours. Some examples are:
Going for a walk = excited feelings
Picking lead up = going for a walk
Picking lead up = excited feelings
Cheese = drooling/anticipation of pleasure
Sound of fridge opening = cheese
Sound of fridge opening = drooling/anticipation of pleasure
Vet = fear
Going in the car = vet
Going in the car = fear
Being kneed in the chest = pain (fear)
Visitors = being kneed in the chest
Visitors = pain (fear)
As you can see, both positive and negative associations can be formed during this type of learning. The scary thing, is that these associations are being learned ALL THE TIME! You cannot switch off this type of learning or take a break from it. Everywhere we take our dogs, everything we do with them and everything we expose them to is creating associations. Understanding this is of vital importance, because feelings lead to behaviours.
Feelings of pleasure will lead to increased engagement and willingness to try new things.
Feelings of fear will lead to survival responses (fight, flight, appeasement)
Feelings of relief will lead to repeating behaviour that brought about relief
Feelings of frustration can result in angry outbursts
Feelings of disappointment can lead to lack of engagement
So, training methods (which quadrant of operant conditioning you choose to make use of) will lead to emotional associations being formed with you as the dog owner, handler or trainer, with the training field or environment that learning is taking place in and with other people, animals or objects in the environment that learning is taking place in. These emotional responses in turn will lead to voluntary behaviour responses designed to either maintain or increase pleasant emotions or alleviate and avoid unpleasant emotions. So, while most trainers only focus on trial and error learning (operant conditioning), Pavlovian conditioning is always at work and influencing operant responses.
Here is a quote from 4Paws University, as I cannot find a better way to express this idea:
“Animal trainer Bob Bailey is frequently quoted saying, "Pavlov is always on your shoulder." Regardless of what you think you are teaching, classical conditioning is always happening. The dog is always learning. Safe. Unsafe. Pleasant. Unpleasant.
So, if you say "Sit" and press on your dog's hindquarters, releasing the pressure only once his rear is on the ground, your dog might learn to sit on cue (an example of negative reinforcement - you release the pressure to reinforce the correct behavior). However, if your dog sits only to prevent you from pushing on him, do you think "Sit" carries a pleasant or unpleasant association? I once worked with a dog that ran away from the owners when they said "Sit." Not the behavior they were going for.”
https://www.facebook.com/4pawsuniversity/posts/10156137301927070
Why do we protect puppies during play, by immediately stopping any kind of bullying behaviour?
Why do we avoid placing reactive dogs in normal training classes?
Why do we set dogs up to succeed and lower our expectations if a dog is struggling with an exercise?
Why do we encourage clients to play with their dogs and have fun with them?
Why do we never use force in training?
The answers to all these questions can be found in Pavlovian or Classical conditioning and the forming of associations:
Why do we protect puppies during play, by immediately stopping any kind of bullying behaviour?
We want to avoid puppies forming negative associations with other dogs, which could lead to defensive behaviour.
Why do we avoid placing reactive dogs in normal training classes?
We do not want socially well-adjusted dogs learning to associate other dogs with lunging, barking and generally threatening behaviour. We do not want dog-reactive dogs learning that coming to training is frightening and stressful due to the presence of the very thing they are scared of.
Why do we set dogs up to succeed and lower our expectations if a dog is struggling with an exercise?
We do not want dogs associating training with feelings of frustration or despondency.
Why do we encourage clients to play with their dogs and have fun with them?
We want dogs to see their owners as a source of fun, pleasure and endless good feelings, because this will make them want to engage and work with them.
Why do we never use force in training?
We do NOT want dogs associating fear and pain with their owners as this leads to a breakdown of the dog-owner relationship and a host of behaviour problems.
So next time you are working with your dog, consider not only the reinforcement you are offering, but the associations being learned in the situation you have placed your dog in. Make sure that they are comfortable, make sure that your interactions with them create positive associations that increase their willingness to engage and work with you. Never forget that, while you may be focused on using operant conditioning to change behaviour, Pavlov really is ALWAYS on your shoulder!