Unlearning of children's fears

1. Introduction

Version followed: Mary Cover Jones's two 1924 papers, "A laboratory study of fear: the case of Peter" (Pedagogical Seminary 31, 308–315; text as hosted by Classics in the History of Psychology) and "The elimination of children's fears" (Journal of Experimental Psychology 7(5), 382–390; as reprinted in the Behavior Therapist 47(3), 2024). Where secondary accounts differ (e.g. Peter's age given as "three"; "candy" as the food), the 1924 texts are followed.

What it is, plainly. This is not a manualised therapy and has no practitioner training track. It is a 1923–24 research programme run at Teachers College, Columbia University, with J. B. Watson advising, that tested seven ways of removing learned fears in young children (70 children aged 3 months to 7 years, most treatment cases under 4) and documented one intensive case, Peter (2 years 10 months at the start), whose fear of a white rabbit had spread to rats, a fur coat, a fur rug, cotton wool and feathers. [source: Jones 1924a; Jones 1924b] This guide gives the clinical procedures Jones describes, then adapts them for a present-day clinician; every adaptation is marked [synthesis].

Core idea of change. Fears in young children are learned (conditioned) responses, so they can be "unconditioned". The feared object is brought into the presence of a stronger competing positive response, eating food the child likes, or the calm, interested behaviour of admired peers, gradually and never faster than the positive response can outweigh the fear, until the object evokes approach and pleasure instead of withdrawal. Of the seven methods tested, only direct conditioning and social imitation gave what Jones called "unqualified success"; the others were at best useful in combination. [source: Jones 1924b] Jones did not use the later terms counterconditioning, reciprocal inhibition or graded exposure; that her work prefigures Wolpe's systematic desensitisation is a later reading (Wolpe called her "the mother of behavior therapy"). [source: Rutherford 2001 intro; Wikipedia]

2. Framework

Jones described no formal phases. The sequence below is reconstructed from what she actually did; the ordering and the move-on criteria are drawn from her texts, while the consent/safety step and modern timing are [synthesis].

Phase 0. Consent, safety and set-up [synthesis]

Goal: parental informed consent, the child's ongoing assent (stopping whenever the child signals "no" in words or behaviour), screening for contraindications (see §4 Direct conditioning and §5), a safe and calm animal or object, a room separate from the child's usual play space (Jones used a laboratory upstairs from the nursery). [source for room: Jones 1924a]
Move on when: consent documented, contraindications excluded, the feared object can be presented in a controlled way.

Phase 1. Fear assessment

Goal: identify which situations actually evoke fear. Jones exposed children to standardised situations (being left alone, a dark room, other children showing fear, and sudden presentation of a snake, white rat, rabbit, frog, false faces and loud sounds) and selected only those who showed specific fear. Most children showed none. [source: Jones 1924b]
Also map spread: for Peter she tested related objects (rat, rabbit, fur coat, cotton wool, feathers, wooden blocks) to see which carried the fear and which did not. [source: Jones 1924a]
Move on when: one target object and its related ("transferred") fears are identified.

Phase 2. Build the graded toleration scale

Goal: an ordered list of observable steps from "fear whenever the object is in the room" to full affectionate contact. Jones's 17-step scale (A–Q) for Peter ran from fear of a caged rabbit anywhere in the room, through tolerating it caged at 12, 4 and 3 feet and then close, loose in the room, being touched while held or loose, contact on the high-chair tray, squatting beside it, helping carry it, holding it on his lap, being alone with it, having it in his play pen, petting it, and finally letting it nibble his fingers (paraphrased). [source: Jones 1924a]
Move on when: steps are ordered and observable. Jones checked her ordering by having six graduate students and instructors rank shuffled steps (average agreement .70). [source: Jones 1924a]

Phase 3. Treatment sessions

Format: individual child, sometimes with one to three fearless peers present; brief sessions in the laboratory; for Peter twice daily at first, then once daily; direct-conditioning sessions timed to the existing mid-morning lunch because meal schedules could not be changed. [source: Jones 1924a; Jones 1924b] Session length is not reported [uncertain].
Count: Peter was first seen 4 December; illness and quarantine removed about two to three months; by late April he patted and played with the rabbit. Jones 1924b gives roughly two months of daily or twice-daily conditioning sessions. The texts name sessions up to at least no. 21 but give no total [uncertain]; secondary accounts' totals were not verified.
Content: direct conditioning (§4) and social imitation (§4) as the main agents; distraction, negative adaptation and verbal appeal only as adjuncts. Jones "rarely used any single method in pure form" (paraphrase) and chose whichever device suited the current stage. [source: Jones 1924b]
Criterion for moving up a step: the object is brought only as close as possible "without arousing a response which would interfere with the eating" [source: Jones 1924a]; i.e. the child keeps eating/playing calmly. Move back when eating stops, the child withdraws or cries.

Phase 4. Generalisation probes and close

Goal: test the untreated related fears. At Peter's last interview, fear of cotton, the fur coat and feathers had gone; responses to rats and the fur rug with a stuffed head were much improved; novel stimuli (a mouse, a tangle of worms) drew brief hesitation and then interest. [source: Jones 1924a]
Close when: the target object evokes positive approach (not mere indifference; Jones's stated aim was acceptance), and related objects are tolerated. Plan follow-up [synthesis]: Jones herself flagged that response on re-exposure to the original fear situation was unknown. [source: Jones 1924a]

3. Guiding principles

[source: Jones 1924a; Jones 1924b unless marked]

Fears are learned and can be unlearned. Treat the fear as a conditioned response to change through new learning, not as something to talk the child out of.
Time away does not cure. Withholding the object for weeks or months left fears intact (disuse); do not advise parents to simply wait it out.
Words change words, not the reaction. A child can learn to talk happily about the feared animal and still react unchanged to the real one; work with the real object.
Gradual approach, paced by the competing response. Distance and contact are set by the strength of the fear relative to the strength of the positive state; never advance faster than the child can keep eating or playing.
The positive response must be stronger than the fear. Use something the child genuinely wants (Jones used food at lunchtime, when hunger was highest, "up to a point"); if it is weak, keep the object further away.
Delicate handling or the reverse happens. A careless pace can attach fear to the food instead of pleasure to the animal; err on the side of slowness.
Peers are powerful in both directions. Calm, admired children reduce fear; a crying child can create a new fear in a previously unafraid one (Vincent). Choose models carefully and control the situation.
Never shame or ridicule. Teasing and scolding suppress the display briefly, then fear returns stronger: "the emotion is re-suggested and entrenched, rather than stamped out." [source: Jones 1924b]
Aim for acceptance, not indifference. Mere habituation is a weaker goal; the end point is the child seeking out and enjoying contact.
Combine methods and match them to stage. No single method is used purely; adjuncts (distraction, repeated presentation) can bridge into direct conditioning.
Avoid building dependence on adult protection. Distraction needs a continually present adult and can leave the child reliant on that adult; fade it.
Measure, chart and check. Plot steps against sessions, have others check the ordering, and test related objects; observable behaviour is the outcome.
Drop measures that frighten. Jones abandoned blood-pressure measurement after Peter grew suspicious of the apparatus. [source: Jones 1924a]
Look at the child's wider world. Jones recorded Peter's unstable home and a mother who used fear as discipline, and an uncontrolled fright (a dog) that undid progress; context sustains or undermines gains. [source: Jones 1924a]
Child assent and welfare come first [synthesis]. Modern practice requires that the child can stop at any point, no food deprivation, and humane treatment of any animal used.

4. Intervention inventory

Jones's papers contain four assessment/monitoring procedures and seven treatment methods, plus two incidental factors she credited. All are listed, including the methods she found ineffective or harmful, because knowing why they fail is part of using the approach. [source: Jones 1924a; Jones 1924b] Example language is illustrative wording for a 2–5-year-old unless quoted; Jones reports almost no experimenter speech.

4.1 Standardised fear screening

Purpose / when: At intake, to find which situations truly evoke fear (most children show none) and to select the target. [source: Jones 1924b]
Steps: (1) With parent consent, present a short, fixed set of situations one at a time: being left briefly alone, a dark room, being near other children showing fear, and the sudden appearance of candidate objects (Jones used a snake, white rat, rabbit, frog, false faces, loud sounds). Today, omit the distressed-peer situation, since Jones's own data show fear spreads socially [synthesis]. (2) Modern adaptation [synthesis]: replace "sudden" presentation with announced, at-a-distance presentation; use the parent's report to choose only the relevant objects; stop at the first clear sign of fear. (3) Record withdrawal, crying, freezing, verbal refusal, and the distance at which they appear.
Example language (illustrative): "There's something in this box. Do you want to look from over here, or stay by Mummy?"
Watch for: Fear produced by the test itself; keep exposures brief and end each one on a calm note [synthesis].

4.2 Mapping transferred (generalised) fears

Purpose / when: At intake and at the end, to see which related objects carry the fear and which do not. [source: Jones 1924a]
Steps: (1) List objects sharing features with the target (for Peter: rat, rabbit, fur coat, fur rug, cotton wool, feathers) and neutral controls (wooden blocks). (2) Present each calmly and record the reaction. (3) Choose one object to treat directly; leave the others as later probes.
Example language (illustrative): "Can you help me put these feathers in the bag?"
Watch for: A pattern (fur, white, moving) that tells you which steps to build; Jones could not identify Peter's original fear source and noted that only transferred fears were treated. [source: Jones 1924a]

4.3 Graded toleration scale

Purpose / when: Before treatment, to define the ladder of steps and the measure of progress. [source: Jones 1924a]
Steps: (1) Write observable steps from "fear whenever the object is anywhere in the room" to full affectionate contact. (2) Include distance steps (caged at far, middle, near distance), freedom steps (loose in the room), contact steps (touches while you hold it; touches it loose), proximity in the child's own space (on the tray, in the play area), active handling (helps carry, holds on lap), independence (stays alone with it), and affection (pets it, lets it nibble fingers). Jones's scale had 17 such steps, A–Q (summarised in §2). (3) Optional check: ask two or more colleagues to put the shuffled steps in order and compare. (4) Adapt the steps to the child's actual object [synthesis].
Example language (illustrative, to parent): "Each line is a small step. We'll only move to the next when your child can stay relaxed at this one."
Watch for: Steps that are too large (the child jumps from calm to crying); split them.

4.4 Charting progress

Purpose / when: Every session, to see trend, plateaus and relapses. [source: Jones 1924a]
Steps: (1) After each session record the highest step achieved. (2) Plot step (vertical) against session (horizontal); mark gaps, illness and incidents on the chart. (3) Note what was different on days with sharp gains (Jones linked rises to peer models and a liked visitor). (4) Do not add physiological measures that alarm the child (Jones dropped blood-pressure readings after one try).
Example language (illustrative, to self/parent): "Today: step H, rabbit loose, he touched it twice; peer model present."
Watch for: Drops after illness, a fright outside sessions, or a minor injury; see §5.

4.5 Direct conditioning (the main method)

Purpose / when: The core treatment for a specific object fear in a young child; one of Jones's two "unqualified" successes. [source: Jones 1924b]
What Jones did: At the existing mid-morning lunch, Peter sat in a high chair with food he liked; the rabbit, in a wire cage, was brought as close as possible "without arousing a response which would interfere with the eating", and moved nearer over repeated sessions until he touched and played with it. Hunger was the strongest motive, effective "up to a point". She warned the method "requires delicate handling" and that a careless operator could produce the reverse, fear attaching to food. [source: Jones 1924a; Jones 1924b]
Contraindications and safety [synthesis]: Do not withhold or delay meals to create hunger; use a preferred snack at a normal snack time. Exclude or adapt for feeding or eating difficulties (e.g. selective/restrictive eating), food allergies, diabetes or other diet-related conditions, choking risk (age-appropriate food only), and families who object to food as a reward. Postpone during acute illness or soon after a fright (Peter relapsed after illness plus a dog incident). For live animals: calm, healthy animal; experienced handler; barrier or cage in early steps; allergy and hygiene checks; hand-washing; the animal's welfare is respected (rest breaks, no teasing). Child assent: the child may stop at any time; parent present or nearby.
Steps: (1) Agree the snack and time with the parent. (2) Seat the child comfortably and let them start eating before the object enters. (3) Bring the object in at the current scale step's distance (start further than you think; Peter calmed only at about 20 feet when first tried this way). (4) Watch eating: if the child keeps eating and glancing calmly, hold; after a short calm period, move a little closer. (5) If eating slows, stops, or the child turns away, move the object back to the last calm distance and stay there. (6) End while the child is still eating and calm; remove the object before the food runs out [synthesis]. (7) Repeat daily or twice daily; carry the last achieved step into the next session, starting one step below [synthesis]. (8) Once close tolerance is stable, add contact steps (touch while held, touch loose) while the child is still in the eating or just-finished state. (9) Fade the food once the child approaches spontaneously [synthesis; Jones does not describe a fading procedure].
Example language (illustrative): "You have your crackers, and Bunny is way over there in his house. We'll just let him sit there while you eat." Later: "He's eating his lunch too. Want to watch him?"
Watch for: Working signs: steady eating, looking at the object, comments about it (Peter: "I like the rabbit"), reaching. Warning signs: food refusal, leaving food when the object enters, crying at the sight of the high chair or snack (fear transferring to food: stop and rebuild; see §5).

4.6 Social imitation (peer modelling)

Purpose / when: The other "unqualified success"; used extensively and among the first methods to show results; combine with direct conditioning. [source: Jones 1924b]
What Jones did: She arranged controlled group sessions with fearless children chosen for prestige in the child's eyes. Peter's gains came when a peer said "Oh, rabbit" and went to look (Peter followed), when another child held the rabbit (Peter then wanted it on his lap), and when three children stroked it (Peter touched its back). Bobby approached a rabbit after two girls eagerly did. [source: Jones 1924a; Jones 1924b]
Steps: (1) Choose one to three children who are genuinely unafraid, calm, and liked or admired by the target child; screen them for allergies and consent [synthesis]. (2) Put the object at a distance the target child tolerates. (3) Let the models approach, look, touch and handle it naturally; do not pressure the target child. (4) Allow the target child to follow at their own pace. (5) Never include a child who is frightened of the object in the same session.
Example language (illustrative, to model): "Sam, would you like to give Bunny a piece of carrot?" (to target child) "You can watch from your chair, or come and see."
Watch for: Peer fear spreading: Vincent, unafraid at first, became afraid of a rabbit for over two weeks after playing near a crying peer; Jones called social induction "perhaps the most common source of maladjustive fear trends." [source: Jones 1924b] Teasing between children (see 4.10). Modern adaptation [synthesis]: a sibling, parent or video model may stand in when peers are not available; Jones did not test these.

4.7 Distraction

Purpose / when: An adjunct to calm a child at a step and to bridge into direct conditioning; Jones found it useful but often temporary, soothing rather than reducing fear. [source: Jones 1924b]
What Jones did: She redirected attention to another activity or topic near the feared object: Arthur reached for crayons placed beside a frog, then said the frog had not bitten him; Sidney's whining at a rabbit stopped when they talked about its name. [source: Jones 1924b]
Steps: (1) Place an attractive activity (crayons, toy, snack) at a distance from the object the child can manage. (2) Engage the child in it; talk about the activity or a neutral feature of the object (its name). (3) Once the child is absorbed, let the object stay or move slightly closer. (4) Hand over to direct conditioning or peer modelling as soon as possible. (5) Gradually reduce your own involvement so the child is not relying on you [synthesis, based on Jones's caution].
Example language (illustrative): "Let's draw Bunny. What colour is his nose?"
Watch for: Calm that vanishes the moment the activity or the adult goes; Jones warned distraction needs an adult continually present and fosters reliance on adult protection.

4.8 Negative adaptation (repeated presentation)

Purpose / when: Repeated, monotonous presentation until the child stops reacting; Jones saw it work (Godfried, 3, came to tolerate a white rat running near and touching him) but judged it suited mainly to infants and animals, and its goal (indifference) too limited. [source: Jones 1924b]
Steps: (1) Present the object at a tolerable distance, unchanged, for repeated short periods. (2) Keep everything else constant. (3) Use only as an adjunct alongside a positive competing response [synthesis based on Jones's verdict].
Example language (illustrative): "Bunny's just sitting there again today."
Watch for: Jones's main warning: repeated presentation without other measures more often increased fear (summation) than reduced it. If reactions grow across sessions, stop and switch to 4.5/4.6.

4.9 Verbal appeal

Purpose / when: Talking about the feared object (pleasant stories, picture books, toy and model animals). Jones found it did not change the reaction to the real object; useful only as a supplement. [source: Jones 1924b]
What Jones did: Jean (about 4) had ten minutes daily of pleasant talk about rabbits with books and toy rabbits; she came to talk freely about rabbits, but a week later her reaction to a real rabbit was essentially unchanged.
Steps (adjunct use only): (1) Use brief, positive talk and books between in-vivo sessions to keep the topic familiar. (2) Never treat fluent talk as evidence of progress; check against the real object.
Example language (illustrative): "In this book the bunny is hungry. What does he like to eat?"
Watch for: Verbal change without behavioural change ("talks brave, still runs").

4.10 Repression (social ridicule, teasing, scolding): do not use

Purpose / when: Documented only so clinicians can recognise and stop it. Jones observed (and did not use in the laboratory) children being teased, ridiculed or scolded for fear in groups. [source: Jones 1924b]
What she observed: Arthur showed bravado about frogs in a group of boys, then fled and was teased; Sidney nodded when girls taunted him as afraid. Shame hid the fear briefly; it returned, "re-suggested and entrenched".
Steps for the clinician [synthesis]: (1) Ask parents and carers whether the child is teased, scolded or threatened about the fear (or with frightening things as discipline, as Peter's mother did). (2) Explain Jones's finding plainly. (3) Agree a neutral response at home: acknowledge, do not mock, do not force.
Example language (illustrative, to parent): "Teasing makes children hide the fear, not lose it. Let's keep the bunny practice here, gentle and step by step."
Watch for: Bravado followed by sudden flight; secrecy about fears.

4.11 Disuse (leaving the fear alone): do not rely on it

Purpose / when: Withholding the feared object for weeks or months. Jones found it ineffective and called it unsafe to rely on. [source: Jones 1924b]
What she observed: Rose (21 months), Bobby (30 months) and Eleanor (21 months) reacted as strongly after one to two months without contact.
Steps for the clinician [synthesis]: Advise against "wait and see" for an established fear; use breaks only for illness or welfare, and expect to restart lower on the scale.
Example language (illustrative, to parent): "Fears like this usually don't fade just by keeping away; small, safe practice works better."
Watch for: Unplanned breaks (illness, holidays) followed by lost ground.

4.12 Presence of a liked adult

Purpose / when: Incidental factor, not a designed method. Two of Peter's rises came on days when a student he liked ("Dr. S.") was present without giving any suggestions; Jones thought this added to his general well-being. [source: Jones 1924a]
Steps [synthesis]: Let a trusted adult the child likes sit nearby, calm and non-directive, during sessions; they should not coax, reassure repeatedly, or show fear themselves.
Example language (illustrative, to the adult): "Just be here and enjoy watching; no need to encourage."
Watch for: A frightened adult undoing progress (Peter's relapse involved a dog plus a frightened adult). [source: Jones 1924a]

4.13 Combining methods by stage

Purpose / when: Throughout. Jones "rarely" used a method in pure form; the aim was to cure the fear with whichever devices suited each stage. [source: Jones 1924b]
Steps [synthesis from Jones's case sequence]: (1) Early, high-fear steps: distance plus eating (4.5), distraction (4.7) to settle. (2) Middle steps: add fearless peers (4.6) to prompt approach and first touches. (3) Contact and handling steps: continue peer models and snack, fade adult support. (4) End: generalisation probes (4.2) with untreated objects. (5) Adjuncts (4.8, 4.9) only alongside, never alone.
Example language (illustrative): "Today we'll have snack with Bunny close by, and Mia is coming to feed him."
Watch for: Over-reliance on one device; if gains stall, change the combination, not the pace.

5. Handling problems

Jones addresses relapse after illness and frights, a minor injury, peer-transmitted fear, the limits of distraction and talk, and the food-reversal risk. Everything else here is [synthesis] unless a source is given.

Overwhelm in a session (crying, flight, freezing). Peter needed about 25 minutes to recover after his first session. [source: Jones 1924a] Response: remove the object or move it to the last calm distance at once; let the child return to eating or play; end on a calm note; next session start two steps lower and use smaller steps. Never hold the child near the object. [synthesis]
Stuck at one step. Split the step (half the distance change; caged before loose; touch while held before touch loose); add a fearless peer model at that step (Peter's jumps came with peers); check that the snack is genuinely preferred and the timing natural. [source for peers: Jones 1924a; rest synthesis]
Relapse after illness, a break or an outside fright. Peter lost most of his gains after two months of scarlet fever plus a dog jumping at him while ill, in an unfamiliar place, with a frightened adult. [source: Jones 1924a] Response: re-assess on the scale, restart from the current level (not the old high point), rebuild with direct conditioning and peers; expect faster recovery than the first time [synthesis]; ask parents to tell you about frights between sessions.
Minor injury (scratch, nip). A scratch while Peter carried the rabbit caused a brief drop that recovered quickly. [source: Jones 1924a] Response: give first aid calmly, acknowledge it, return to a lower step the same or next session; review handling (gloves, handler, calmer animal); inform parents. [synthesis]
Fear transfers to the food or setting. Jones's warning: careless pacing can produce the reverse. [source: Jones 1924b] Signs: refusing the snack, upset at the high chair or room. Response: stop pairing; re-establish the snack with no object present for several sessions; restart with much greater distance; consider swapping the competing positive state to a favourite play activity [synthesis; Jones did not test play as the competing response]. If eating generally worsens, stop food pairing altogether and review for a feeding problem.
Child refuses to come or says "no" (resistance/assent). Respect it; offer a choice of a lower step or a session with only peers playing near the object; reduce session length; check whether something at home has increased fear. Do not coax repeatedly or bribe beyond the planned snack. [synthesis]
Peer model becomes frightened. Fear spreads socially (Vincent). [source: Jones 1924b] Response: end the group part immediately, separate the children calmly, reassure both; check the model child is not left with a new fear (offer them a brief calm, positive contact); choose steadier models next time.
Teasing among children or from family. Stop it in the session; brief parents on Jones's finding that ridicule entrenches fear (4.10). [source: Jones 1924b]
Calm only while distracted or with the adult close. Fade distraction and adult proximity in steps; move toward independence steps (child near the object without adult). [source for caution: Jones 1924b; steps synthesis]
Talks bravely but avoids the real object. Treat words as no evidence; return to in-vivo steps. [source: Jones 1924b]
Non-adherence: missed sessions, family avoidance at home. Jones's data show long breaks lose ground. [source: Jones 1924b] Response: keep sessions frequent and short; agree what parents do between sessions (neutral, no forcing, no threats, no frightening discipline); if sessions must pause, plan a lower restart. [synthesis]
Rupture with the child (distrust of the clinician or equipment). Jones abandoned blood-pressure measurement after Peter grew suspicious of the instruments. [source: Jones 1924a] Response: drop the trigger, spend a session on play with no object, let a liked adult be present (4.12), resume lower. [synthesis]
Home stressors maintaining fear. Peter's mother used fear as discipline and his home was unstable. [source: Jones 1924a] Response: address parenting practices and family stress alongside the fear work; refer for family support where needed. [synthesis]
Fear that is not a discrete object fear (separation, dark, generalised worry, trauma-related fear). Jones's methods were tested for specific object fears; for other presentations use current evidence-based child anxiety treatment instead. [synthesis]
Ethical and modern-practice issues. Jones's studies used institutionalised children, deliberate sudden fear tests, and hunger as a lever; consent practices of the time are not described in the texts [uncertain]. Today: informed parental consent, child assent, no deprivation, no deliberate frightening, humane animal use. For clinical work, the modern descendants are graduated in-vivo exposure within cognitive-behavioural therapy for childhood specific phobia, intensive one-session treatment (Öst), and parent-involved CBT; these were not researched for this guide and are named only as pointers [uncertain]. Jones's procedures are best used as a historically grounded, play- and snack-based form of graded exposure for very young children within such a plan. [synthesis]

Sources

Sources:

Works cited:

Jones, M. C. (1924a). A laboratory study of fear: the case of Peter. Pedagogical Seminary, 31, 308–315.
Jones, M. C. (1924b). The elimination of children's fears. Journal of Experimental Psychology, 7(5), 382–390.
Watson, J. B., & Rayner, R. (1920). Conditioned emotional reactions. Journal of Experimental Psychology, 3, 1–14. [uncertain]
Jones, M. C. (1974). Albert, Peter, and John B. Watson. American Psychologist. [uncertain]
Alpers, G. W. (2024). Happy 100th anniversary, behavior therapy! Behaviour Research and Therapy. [known via press release]

← Previous · II-001Persuasion therapy Next · II-003 →Weekes's method for "nerves"