The Little Albert experiment: how Watson conditioned fear, and what the records changed

In the winter of 1919 and 1920, John B. Watson and Rosalie Rayner taught a 9-month-old boy to fear a white rat. They paired the animal with a loud clang. After seven pairings the boy cried at the rat alone, avoided furry objects, and still avoided them a month later. His mother then removed him from the hospital, so the fear was never undone. Today the study could not run as Watson ran it.

• Watson and Rayner conditioned Albert B., a healthy 9-month-old described as stolid, by striking a steel bar behind his head each time he touched a white rat.
• Seven joint pairings produced a full reaction: crying and crawling away so fast that staff caught him before the table edge.
• The fear spread to a rabbit, a dog, a fur coat, cotton wool, and a Santa mask, but not to wooden blocks.
• Thirty-one days later avoidance persisted in weaker form, with new signs such as head nodding and thumb-sucking that blocked the reaction until staff removed the thumb.
• The identity of the boy is disputed: Beck and colleagues proposed Douglas Merritte in 2009, while Powell and colleagues proposed Albert Barger in 2014, with later reviews favoring Barger but no agreement.

What Watson and Rayner actually did

Watson and Rayner worked with one infant, Albert B., reared in the Harriet Lane Home for Invalid Children, where his mother served as a wet nurse (Watson and Rayner, 1920). At nine months he weighed twenty-one pounds and staff described him as stolid and unemotional. That stability was one reason the team chose him. The paper states plainly that work so far covered only one child.

At about nine months the team ran baseline tests (Watson and Rayner, 1920). Albert met a white rat, a rabbit, a dog, a monkey, masks with and without hair, cotton wool, and burning newspapers. He reached for the objects and showed no fear. His mother and attendants confirmed that he practically never cried.

At eight months and twenty-six days the team tested sound. One researcher held his gaze. The other struck a suspended steel bar behind his head with a hammer. The bar was four feet long and three-fourths of an inch thick. He started violently and raised his arms. On the second strike his lips puckered and trembled. On the third he broke into crying. Removal of support, by dropping and jerking the blanket, was tried exhaustively the same day and did not frighten him.

Conditioning began at eleven months and three days (Watson and Rayner, 1920). The rat appeared, Albert reached, and the bar sounded as his hand touched the animal. He jumped, fell forward, and buried his face, without crying. On the second pairing he jumped again and whimpered. The team stopped for a week to avoid serious upset.

At eleven months and ten days the rat alone produced fixation without reaching, then withdrawal when it nosed his hand (Watson and Rayner, 1920). Blocks served as a control between trials, and he played with them normally, which showed that the room and table had not become frightening. Five more rat-and-noise pairings followed. On the last rat-alone test he cried at once and turned left. He fell over, rose on all fours, and crawled away so fast that staff caught him before the table edge. The paper counts seven joint stimulations in all for the complete reaction.

The conditioning map

The study is a textbook case of classical conditioning, in which a signal predicts an event. The behavior of the learner did not change what happened. The clang arrived whether or not Albert reached. The learning was classical rather than operant.

Generalization means the learned reaction spreads to similar stimuli. Discrimination means it does not spread to all stimuli. Albert showed both. Fear spread to furry animals and objects, while blocks and the hair of two observers left him calm and playful.

What the fear spread to, and what it did not

Five days after conditioning ended, at eleven months and fifteen days, the team tested transfer (Watson and Rayner, 1920). Blocks came first and he played as usual. The rat alone drew whimpering and turning away. The rabbit drew a pronounced reaction: leaning away, whimpering, tears on contact, burying the face, and crawling away crying. The dog drew a milder start, shrinking back without tears, then crying when it neared his face. A seal fur coat drew withdrawal, fretting, crying, and crawl-away. Cotton wool in a paper pack drew kicking and hand withdrawal, though he later played with the paper while avoiding the wool. Watson's own hair drew a negative reaction, while the hair of two other observers drew calm play. A Santa mask with white hair drew crying and turning away.

At eleven months and twenty days the team freshened the rat reaction with one more pairing. Then, for the first time, staff paired the rabbit and the dog directly with the clang. That step confounds the design, because later fear of those animals may reflect direct pairing rather than spread from the rat. The same day the team moved to a large lit lecture room. Reactions there ran weaker. A quiet dog then barked three times about six inches from his face, and Albert fell over and wailed until staff removed the animal. Even adult observers started at the bark.

A month later

Thirty-one days passed with no emotional tests, only weekly lab visits for handedness and development checks. At one year and twenty-one days the team retested (Watson and Rayner, 1920). The Santa mask drew withdrawal and crying on sight and on forced touch. The fur coat drew nose wrinkling, withdrawal, whimpering, head nodding in a new peculiar pattern, and pushing with the feet while keeping the hands away. The rat drew intent fixation, hand withdrawal on touch, fretting with head nodding, and covering the eyes when it reached his chest. The rabbit drew pushing away with the feet, nodding, and wailing of da da. He touched it tentatively, shuddered, and sucked his thumb. The thumb-sucking blocked fear until staff removed the thumb. The dog, very active that day, drew intense fixation, crying, and hands over the face.

Watson and Rayner read this as persistence with some loss of strength. They wrote that such responses persist and shape personality through life, especially in a phlegmatic child like Albert. An excitable child, they thought, would have kept the full form unchanged.

Why the study would not run today

The original report itself notes the responsibility such work carries (Watson and Rayner, 1920). By current standards the procedure fails on four points. There was no informed consent in the modern sense. His mother gave permission to observe and test, but the report does not show that she understood that staff would deliberately frighten him. Staff caused clear distress, including violent jumping, whimpering, crying, and rapid crawl-away. The team never removed the fear. Albert left the hospital the day of the last test, before any reversal. No follow-up tracked whether the fear faded or stayed.

Under standards set in the late 1970s, such work with an infant would not be allowed. The National Commission for the Protection of Human Subjects and the Belmont Report set the current frame (history of the experiment). The Milgram obedience study raises similar ethics questions about classic experiments. Modern rules require consent, minimal harm, and care for vulnerable groups, especially children.

Watson and Rayner listed four removal methods they never got to try (Watson and Rayner, 1920). Constant exposure without the clang might wear the reaction down through habituation. Tactile reconditioning would pair the feared object with pleasant touch. Feeding would pair the animal with food. Constructive play would build calm handling through imitation. The second idea, as worded in 1920 with reference to lips, nipples, and sex organs, reads as troubling today and must not be copied.

A student of Watson did carry the gentler idea forward. Mary Cover Jones worked with Peter, a two-year-old who feared rabbits and furry objects (history of the experiment). She moved a caged rabbit closer in small steps while he ate and played with non-fearful children. The fear faded. That graded method led toward systematic desensitization, still used in exposure-based care for phobias. Watson never tested any method on Albert.

Who was Little Albert

Watson called the boy Albert B. and never named him further. The paper gives the hospital, the wet-nurse mother, the male infant, and exact test ages (Watson and Rayner, 1920). Those clues launched a long archival search (Beck with Irons, 2011).

Beck, Levinson, and Irons reported in 2009 that the boy was Douglas Merritte (paper; DeAngelis, 2010). He was the son of Arvilla Merritte, a wet nurse at Harriet Lane. Douglas was born on 9 March 1919 on the Hopkins campus. That date fits the birth window of 2 to 16 March 1919 derived from the test ages (Beck with Irons, 2011). His mother received one dollar for participation. Douglas left Hopkins in the early 1920s and died on 10 May 1925 of hydrocephalus with convulsions. Later work argued that film stills showed hand-scooping, weak eye scanning, and flat affect. Those signs fit impairment, though other authors rejected that reading (history of the experiment).

Powell, Digdon, Harris, and Smithson proposed a different boy in 2014: Albert Barger, later known as William Albert Martin (Powell and colleagues, 2014). He was born within a day of Merritte. His mother also worked at the hospital, and his weight and health matched the paper's healthy description. He left the hospital at twelve months and twenty-one days, the exact age the paper gives for the last day (history of the experiment). His clinical file, the authors report, did not show the blindness described in Merritte's file, which fits the film, where Albert tracks objects and people. Barger lived into his eighties. Later reviews judge the weight of evidence to favor Barger, while the matter stays formally contested (Powell and colleagues, 2014).

The safe exam answer names both candidates and states the contest. No follow-up ever showed whether either boy kept a fear of furry objects. Douglas died young with no persistence data. Reports of Barger's later dislike of dogs cannot be tied to the experiment.

What textbooks get wrong

Ben Harris showed in 1979 that textbook retellings distort the report, and later surveys found the same errors still in circulation (history of the experiment). Common changes include naming the wrong animal, inflating the number of pairings, widening the spread of fear, and claiming that Watson removed the fear. He never did.

Surviving footage adds caution. The film covers only parts of the tests, and textbooks read the same film differently (history of the experiment). Albert often looks indifferent in the clips, and some cries plausibly reflect startle at touch or looming rather than learned fear. Titles that announce conditioning before the viewer watches may inflate what the eye sees.

Replication adds more caution. Later teams reported failures with the same loud-noise method and neutral objects such as blocks and curtains. The original used a furry moving animal. Fear links form more readily to such animals than to neutral objects, which helps explain the gap. The study used one calm infant, no control group, and descriptive notes rather than set measures. Maturation also confounds the result, since fearfulness grows naturally from eleven to twelve months, and repeated clangs alone may sensitize any infant.

None of this makes the study unimportant. It gave the first lab display of human emotional conditioning, used a real baseline plus the blocks control, and launched behavior therapy through Jones. Yet the fair label, with Harris, is interesting but hard to interpret. Fear learning alone also does not explain why phobias last. Avoidance brings instant relief. That relief rewards avoidance and helps keep the fear alive. The negative reinforcement account of avoidance covers this point in detail.

Little Albert versus Little Peter versus Pavlov

Exam questions mix three names. One table keeps them apart.

The four schedules of reinforcement belong to a different operant story. Little Albert turns on a stimulus that predicts an event. Operant work turns on a behavior that earns a consequence.

Students can save classical conditioning, generalization, discrimination, extinction, and preparedness as a set in Psychology World. Make one bookmark list for this article. Drill it with Flashcard, Question, and Typing games until the ages, the seven pairings, the transfer list, and the two candidates stick.