4.3.1 - Watson and Rayner’s Little Albert

The question and Albert's initial responses

Watson and Rayner asked whether fear could be conditioned to a previously harmless animal, transfer to other stimuli, persist over time and later be removed.

Their participant was one infant, called Albert B. in the report, who had been raised in a hospital where his mother worked as a wet nurse. They selected him partly because they described him as unusually calm. This was a deliberately chosen single case, not a representative sample or random selection. The researchers reported that he was healthy; this is their original description, rather than proof of later claims about his identity or medical history.

Around nine months of age, Albert was presented with animals and objects including a white rat, rabbit, dog, masks and cotton wool. He reached towards or manipulated them without the fear reactions the researchers sought. A loud noise from a hammer striking a steel bar behind him did provoke startle and, after repeated presentations, crying. Establishing different initial responses matters: a fear response already present to the rat could not simply be credited to subsequent conditioning.

Before conditioning, the rat was a neutral stimulus in relation to this fear response. The loud noise was the unconditioned stimulus, and the startle/distress it produced was the unconditioned response. After pairing, the rat could become the conditioned stimulus, producing a conditioned response of fear or avoidance without the noise. These labels depend on the stimulus-response relationship; the rat did not itself become an unconditioned stimulus.

The investigation used controlled presentations and repeated observations of the same child. It is often described as a laboratory case study with experimental features: the researchers manipulated pairings, but there was no separate control child and no independent-groups experiment.

Acquisition and later testing

Fear acquisition occurred over two conditioning sessions, followed by further tests and additional pairings.

At 11 months 3 days, the researchers paired contact with the rat and the loud sound twice. Albert startled and then whimpered. They paused for a week. At 11 months 10 days, the rat alone initially produced hesitation and hand withdrawal; five further rat–sound pairings followed, interspersed with rat-alone tests. By the final rat-alone presentation, Albert cried and crawled away. Thus seven pairings established the initial response, rather than seven pairings covering every later session.

At 11 months 15 days, rat-alone responses still included withdrawal and whimpering, five days after acquisition. Blocks were repeatedly offered between trials, and Albert continued to play with them. The block observations suggest he was not simply afraid of every item or the whole experimental situation, although blocks were not a full experimental control condition.

At 11 months 20 days, some responses were weaker. The researchers added another rat–sound pairing to strengthen the reaction. They also directly paired the rabbit and dog with sound, then tested Albert in a larger, brighter lecture room. Responses there were sometimes weaker, and another rat–sound pairing was administered. A later dog bark also provoked strong distress. These extra events matter: later animal responses cannot all be treated as uncontaminated tests of transfer from the original rat conditioning.

After 31 days without further emotional-conditioning experiments, final tests were made at 1 year 21 days. Albert still showed some withdrawal, fretting or crying to several stimuli, but responses differed in strength and sometimes occurred alongside tentative touching. The report records persistence for about a month with some reduced intensity. It does not provide follow-up into adult life.

Generalisation and the untested removal plan

The early transfer tests support generalisation to some related stimuli, but the effects were neither identical nor universal.