Psychology › Section F · Learning › Topic F.2
Classical conditioning
Why does the smell of a dentist's surgery make some people tense before anything has happened? Classical conditioning explains how a once-meaningless signal comes to trigger an automatic response, and it underlies many everyday fears, cravings and aversions.
Pavlov's accidental discovery
Ivan Pavlov (1849-1936) was a Russian physiologist, not a psychologist; he won the 1904 Nobel Prize in Physiology or Medicine for his work on digestion. Measuring his dogs' saliva, he noticed a nuisance: experienced dogs began drooling early, at the sight of the bowl or the attendant who usually fed them. Rather than ignore these "psychic secretions", Pavlov studied them under controlled conditions, pairing food with signals such as a metronome or a tone, and produced the first systematic account of how organisms learn that one event predicts another.
The five terms
Food in the mouth makes a dog salivate without any learning: the food is the unconditioned stimulus (UCS, sometimes written US) and the salivation is the unconditioned response (UCR). A tone at first produces no salivation: it is a neutral stimulus (NS). If the tone is sounded just before the food on many occasions, it starts to produce salivation by itself. The tone is now a conditioned stimulus (CS), and the salivation it produces is the conditioned response (CR). The UCR and CR are often the same behaviour with different triggers: "unconditioned" means unlearned, "conditioned" means learned.
Acquisition and timing
Acquisition is the stage in which the association forms and the CR grows stronger with repeated pairings. Order and timing matter: learning is best when the neutral stimulus comes first and the UCS follows soon after, usually within a fraction of a second to a few seconds. A tone that comes after the food, or long before it, produces little learning. Robert Rescorla showed in the late 1960s that pairing alone is not enough: the CS must reliably predict the UCS. If food arrives just as often without the tone, the tone tells the dog nothing. Rescorla and Allan Wagner turned this into a mathematical model in 1972.
Extinction and spontaneous recovery
Presenting the CS again and again without the UCS makes the CR fade: extinction. Yet Pavlov found that after a rest the extinguished response returned, in weaker form, when the tone sounded again. This spontaneous recovery shows that extinction does not erase the original association. Researchers now regard extinction as new learning, "the tone no longer means food", that competes with the old learning. That is why a fear that has faded in therapy can reappear in a new place or after a long gap.
Generalisation and discrimination
A dog conditioned to one tone will often salivate to similar tones as well, and the more similar the tone, the stronger the response. This is stimulus generalisation. If only the original tone is ever followed by food, the dog learns to respond to that tone and not to its neighbours: stimulus discrimination. Generalisation makes an animal bitten by one snake wary of others; discrimination stops it fleeing from every rope. In higher-order (second-order) conditioning, an established CS is paired with a new neutral stimulus, which then also produces the CR; conditioning beyond this second level is usually weak.
Worked example: naming the parts
Every time Sam's mother gives him an injection for his allergies, she first wipes his arm with a strong-smelling antiseptic, and the needle makes him wince. After several weeks Sam winces as soon as he smells the antiseptic. 1. Unlearned link: the needle prick causes wincing without learning, so the needle is the UCS and wincing at it the UCR. 2. Signal: the smell came just before the needle and at first meant nothing, so it began as the NS. 3. After pairing, the smell alone causes wincing: the smell is now the CS and wincing at it the CR. 4. Predict: wincing at a similar-smelling hand gel would be generalisation.
Biology limits what is learned
Not every pairing is equally easy. In 1966 John Garcia and Robert Koelling gave rats sweet-tasting water accompanied by flashing lights and clicks. Rats later made ill avoided the taste but not the lights and sounds; rats given electric shock while drinking avoided the lights and sounds but not the taste. Taste aversions also form after a single pairing and despite delays of hours, which breaks the usual timing rule. Martin Seligman called this preparedness: evolution has made some associations, such as taste with sickness, easier to learn because they protected survival.
Worked example: a one-trial aversion
Priya eats prawn curry at a party, falls ill with a stomach virus about four hours later, and afterwards feels queasy at the smell of prawns. 1. UCS and UCR: the virus causes nausea without learning. 2. CS and CR: the prawn taste and smell, paired once with illness, now trigger queasiness. 3. Unusual features: one pairing and a delay of hours were enough, which fits Garcia's findings on preparedness. 4. Note: the prawns did not cause the illness, yet the association formed. 5. Prediction: eating prawns several times without illness should extinguish the aversion, though some queasiness may return after a gap.
Little Albert: what was shown, and its problems
In 1920 John B. Watson and Rosalie Rayner reported conditioning fear in an 11-month-old infant they called Albert. After a steel bar was struck loudly behind him each time he touched a white rat he had liked, he cried at the rat alone, and his distress spread to a rabbit, a dog and a fur coat. It is cited as evidence that emotions can be conditioned, but it is weak evidence: one child, no controls, no objective measure of fear, and others' attempts to repeat it in the following years gave mixed results. Ethically it fails modern standards: the researchers deliberately frightened an infant and did not remove the fear before he left. Historians still dispute who Albert was.
Undoing fear and other applications
In 1924 Mary Cover Jones, working with Watson's guidance, reduced a toddler named Peter's fear of rabbits by bringing a rabbit gradually closer while he ate food he enjoyed. Replacing a fear response with a calmer one in this way is counterconditioning, and it foreshadowed exposure-based treatments still used today. Conditioning also helps explain why some chemotherapy patients feel sick on arriving at the clinic, and why advertisers pair products with attractive people or music, hoping the good feeling transfers to the brand.
Common mistakes to avoid
First, labelling by behaviour alone: salivation can be a UCR or a CR, so ask what triggered it. Second, reversing the order: the neutral stimulus must come before the UCS, not after. Third, assuming extinction erases learning: spontaneous recovery shows the association survives. Fourth, treating pairing as sufficient: the CS has to predict the UCS reliably. Fifth, confusing generalisation (responding to similar stimuli) with discrimination (responding only to the CS). Sixth, citing Little Albert as solid proof: it is a single, uncontrolled and unethical case.
Key takeaways
In classical conditioning a neutral stimulus that reliably comes just before an unconditioned stimulus becomes a conditioned stimulus, producing a conditioned response. Pavlov discovered it; Rescorla showed that prediction, not mere pairing, drives it. Extinction weakens the CR; spontaneous recovery shows the association remains. Generalisation and discrimination decide which stimuli trigger it. Garcia and Koelling (1966) showed biological preparedness: taste aversions form in one trial over hours. Watson and Rayner's Little Albert (1920) suggested fears can be conditioned but was a single, uncontrolled, unethical case; Mary Cover Jones (1924) showed fear could be counterconditioned.
Flashcards
Flashcard
Triggers a response without learning
Test your understanding
Question 1 of 5
A cat runs to the kitchen whenever it hears the cupboard where its food is kept being opened. In this example, what is the sound of the cupboard?
Question 2 of 5
After extinction, a dog stops salivating to a tone. The next day the tone is sounded once and the dog salivates a little. What does this show?
Question 3 of 5
What did Garcia and Koelling's 1966 experiment with rats show?
Question 4 of 5
Which is the strongest criticism of Watson and Rayner's Little Albert study as scientific evidence?
Question 5 of 5
A researcher sounds a tone just before a dog's food on some trials, but gives food just as often on trials with no tone. Little conditioning occurs. Why?
