Occasionally when talking to us dog trainer types you will hear us using what may seem to
be some baffling terms... the following are some of the most common and are all to do with how dogs and most species, including humans, learn.
Operant conditioning
Classical conditioning
Un/Conditioned stimuli
Un/Conditioned response
Positive reinforcement
Negative reinforcement
Positive punishment
Negative punishment
But what do all these terms mean, a good trainer will take the time to explain these terms to you but in case you find yourself confused by doggy mumbo jumbo then here is a brief introduction to the mysteries of learning theory.
Modern
dog training methods are founded on a concept called conditioning. There are two types of conditioning,
classical and operant – both of which dogs respond to.
Simply put, classical
conditioning is responsible for involuntary responses, e.g. a dog salivating
when food is served, while operant conditioning is responsible for voluntary
responses, e.g. a dog sitting for a treat.
Classical
conditioning was pioneered in the late 19th century by Ivan Pavlov.
Pavlov showed that when dogs smelled food saliva was released in
anticipation. He called the food an unconditioned stimulus and the
salivating the unconditioned response. Unconditioned because the dogs were responding automatically to the stimulus - they had not been trained to do so. Pavlov was an inquisitive kind of chap and soon discovered that if he rang a bell on a regular basis before feeding the
dogs the dogs soon started to salivate at the sound of the bell. The bell had become
a conditioned stimulus and the dogs’
salivating a conditioned response. Prior to linking the bell ringing to the food the dogs did not salivate when a bell rang - that is to say that the bell did not naturally produce salivation, the dogs had to be conditioned in order to salivate upon hearing the bell ring.
Pavlov
went on to show that it was possible to build an association between an
unconditioned and a conditioned stimulus so that either would elicit the same involuntary response. He called this association between the stimuli classical
conditioning.
At
around the same time that Pavlov was working on the theory of classical conditioning two Americans, Edward Lee Thorndyke and B.F. Skinner, were working on operant conditioning. Where classical conditioning forms an association between two stimuli to elicit a response, operant conditioning forms an association between a behaviour and a consequence. (You might also come across operant conditioning referred to as response-stimulus or RS conditioning because we are forming an association between the animal's response [behaviour] and the stimulus that follows [consequence])
Typically
classical conditioning is used to change a dog’s emotional state. It can help
the dog to feel positive about potentially negative situations or objects, such
as nail clippers or a visit to the vet.
Operant
conditioning is the science that lies behind the positive, reward based methods
of training most modern dog trainers prefer. At its most basic operant conditioning is the modification of behaviour through the use of consequences (reinforcers and punishers). Operant conditioning differs from classical conditioning in that it deals with changing operant or 'voluntary' behaviour as opposed to reflexive 'involuntary' behaviour. That being said, whenever you're dealing with changing behaviours, operant and classical conditioning can work hand-in-hand.
Operant conditioning has two main tools for modifying behaviour - reinforcement and punishment. Reinforcers increase behaviour, while punishers decrease a behaviour. These operate in two contexts - positive and negative. In this case, positive refers to addition; negative refers to subtraction.
What is a reinforcer?
A
reinforcer is a rewarding stimulus that will increase the desired behaviour. A reinforcer can be food or it can be a variety of other things including
toys, freedom to roam, walks, play, affection and much more.
Different
stimuli will have different results depending on a variety of factors such as
the dog, the trainer, the environment, the trigger event, how the stimulus is
applied.
How
and what a dog learns is also highly dependent on timing (when a stimulus is
applied), and frequency (how often a stimulus is applied). Generally,
we want to apply the reward as close to the target behaviour as possible.
There
are three main components of operant conditioning:
Reinforcement
Something that
makes the behavior occur again.
·
Reinforcement can be either positive or negative.
·
Positive reinforcement adds something, such as food, to increase the
occurrence of a behaviour.
·
Negative reinforcement removes something to increase the behaviour.
Punishment
Something that
causes a behavior to become less frequent.
·
Punishment can also be positive or negative.
·
Positive punishment is adding something to decrease the occurrence of a
behaviour - this is how shock or spray collars and other aversives work.
·
Negative punishment is removing something to decrease the occurrence of
a behaviour. So you stop walking with a dog when it pulls, the fun ends and the
pulling stops.
Extinction - No consequence to
behavior at which point the behavior becomes less frequent.