Conditioning

Conditioning Definition

Conditioning is the process, central to behaviourist learning theory, of training or changing behaviour through association and reinforcement — the establishing of new connections between stimuli and responses. Behaviourists sought to study conduct as scientifically as possible by analysing only observable inputs (stimuli) and outputs (responses), and conditioning is a form of learning that can be observed in exactly this way.

Classical Conditioning

Classical (or Type-S) conditioning was identified by Ivan Pavlov in his experiments on the salivary reflex in dogs. When a neutral stimulus, such as a bell, is repeatedly paired with an unconditioned stimulus that already provokes a reflex response — food producing salivation — the two become associated, and the neutral stimulus acquires the power to evoke the response on its own. The bell becomes a conditioned stimulus and the salivation a conditioned response, or conditioned reflex. Classical conditioning operates only on involuntary behaviours governed by the autonomic nervous system, such as salivation, sweating, and heart rate. Reinforcement here precedes the response and is delivered regardless of it, being typically the unconditioned stimulus itself; if the conditioned stimulus is then repeated often without reinforcement, the response fades — a process called extinction.

Operant Conditioning

Operant, instrumental, or Type-R conditioning — associated with E. L. Thorndike and above all B. F. Skinner — trains voluntary responses by delivering reinforcement only after, and contingent upon, the response. In Skinner’s experiments, a rat’s bar-pressing was reinforced with a food pellet each time it occurred. Reinforcement using something pleasant is positive reinforcement; reinforcement through the avoidance of something unpleasant, such as an electric shock, is negative reinforcement; and a reinforcer whose value is itself learned — tokens exchangeable for food, for example — is a secondary reinforcer. Because reinforcement follows the response, behaviour can be manipulated by varying when it is given (schedules of reinforcement): learning is most resistant to extinction when the schedule is response-related and unpredictable, as in gambling on a fruit machine. Behaviour can also be shaped towards a desired end by reinforcing successive approximations to it — the principle behind teaching animals tricks outside their normal repertoire, much everyday social control, and behaviour therapy, in which people learn that certain patterns of conduct bring rewards. Extinction has its practical uses too: socially unacceptable behaviour, on this view, should be ignored rather than reinforced.

Ambitions and Decline

Some behaviourists hoped to explain aspects of human culture through the conditioning of infants and children, but the programme made little headway. Debate among learning theorists increasingly turned on how to interpret the experimental evidence: early behaviourists treated conditioning as a simple, unconscious, automatic process, but a series of experiments showed convincingly that cognitive processes are involved in forming stimulus–response connections. From the 1960s the rise of cognitive and information-processing approaches in academic psychology shifted attention away from conditioning studies and from stimulus–response models of learning altogether.