Autonomy vs. Shame and Doubt

Autonomy vs. shame and doubt is the second of Erik Erikson’s eight psychosocial stages, running from about 18 months to 3 years old. If caregivers encourage a toddler’s new independence while tolerating failure, the child develops autonomy and the virtue of will. If caregivers over-control, criticise, or shame the child’s attempts, it develops shame and self-doubt.

How Does a Toddler Develop Autonomy or Shame and Doubt?

The Parents’ Role

A toddler has new abilities, but it still relies on the adults around it. In tasks such as dressing, feeding and toileting, an adult faces a choice. Let the child do it badly, or step in and do it properly?

That makes the adults the people who matter most.

They are both parents, or the parenting adults collectively. This crisis is about the child’s relationship to authority and permission. In other words, the adults set the limits and decide what the child may try.

They matter so much because they settle the outcome. The crisis is resolved by how they respond to the child’s attempts to act for itself (Erikson, 1963).

The child relies on them in two ways. Practically, they decide how much of each task it may try alone. Emotionally, it needs them to be there, and to tolerate its failures.

What the Toddler Is Working Out

The first stage, trust vs. mistrust, runs from birth to about 18 months and settles whether the world is dependable.

In Erikson’s scheme, autonomy is paid for out of that trust. A toddler tests its own powers only if it is reasonably confident the caregiver will still be there on return.

Now new abilities arrive. By about 18 months a toddler can walk, climb, grasp, speak and refuse. It can also control its own bowel and bladder.

The child discovers that it is separate from the adults, with a will of its own and the power to make things happen. That brings a new question.

Can I do things myself?

Put more fully: can I act on the world myself, and is it acceptable for me to do so?

The toddler finds out by trying things and seeing how the adults respond. There are two possible answers:

• Autonomy is confidence in your own ability to act, choose and manage yourself.
• Shame and doubt are the opposite. Shame is feeling exposed as inadequate in front of others, and doubt is private uncertainty about whether you are capable.

Erikson called each stage’s central conflict a crisis, a turning point where a person’s psychological needs meet what the social world provides (Erikson, 1950, 1963).

That meeting of the psychological and the social is why his theory is called psychosocial.

Initiative vs. guilt follows at about 3 years. You can read the full sequence in Erik Erikson’s stages of psychosocial development.

Why Refusals and Tantrums Are Part of the Process

In practice, a toddler asserts its will constantly. It insists on doing things unaided, refuses help and above all says no.

This can look like plain stubbornness. Erikson read it as developmental work instead. He stressed that the assertion is necessarily excessive at first. A toddler cannot yet tell asserting a preference from defying a real necessity.

So the battles over clothes, food, leaving the house and bedtime are not a malfunction. They are the process itself.

How Autonomy Develops

Autonomy grows out of a cycle that repeats many times a day.

First, the toddler tries to do something for itself, such as dressing, feeding or using the toilet. Next, the adults encourage the attempt.

They let the child explore what it can do, give it time and offer limited choices. They help with the part it cannot yet manage and treat the fumbling as ordinary rather than as failure.

The child draws a conclusion. Attempting is safe, difficulty is normal, and it is the kind of person who manages things.

Repeated across hundreds of moments, that conclusion becomes a settled confidence in its own ability to act, choose and manage itself. That is autonomy.

The confidence then spreads. It carries over from one task to the next, from clothes to spoons to stairs. It also shapes how the child sees itself, as someone who does things rather than someone things are done to.

The strength that grows from this is will, the determination to exercise free choice and self-restraint.

How Shame and Doubt Develop

Shame and doubt grow from the same cycle when the adults’ response goes wrong. There are three ways it can happen.

First, the adults may take over. They do everything for the child, forbid attempts, or criticise failures. The child concludes that its attempts are not worth the time and that competent action is something adults do to it. That conclusion is doubt.

Second, the adults may shame the child, with irritation or ridicule, especially over a toilet accident. The failure is now exposed to others, and the lesson carries humiliation. That is shame.

Third, the adults may give too little help. They leave the child with tasks far beyond its ability, so failure is constant. The child learns it is incapable.

Repeated across hundreds of moments, these conclusions harden. The child decides it cannot manage itself, becomes overly dependent on others, and carries shame about its own inadequacy.

The conclusion spreads, just as confidence does. It carries from one task to the next, and the child reaches the next stage not yet confident it may act at all.

What Builds Autonomy or Shame and Doubt in Everyday Life?

Everyday Scenarios Compared

A toddler learns whether it may act for itself from how the adults treat its attempts, day after day. The table below contrasts four everyday situations.

Everyday scenario	Builds autonomy	Builds shame and doubt
Mealtime	The child feeds itself and spills some. The adult is patient and treats the mess as ordinary.	The adult takes the spoon and feeds the child because it is quicker.
A toilet accident	The adult treats it as an event, not a verdict on the child.	The adult shows irritation or ridicule, so the child feels exposed and humiliated.
At nursery	Water and real cutlery are within reach, so the child can manage alone.	Adults do every task for the child, so it gets no chance to try.
Putting on shoes	The adult holds the shoe steady while the child pushes its own foot in.	The child is left alone with a job far beyond its ability and fails every time.

In every row, the overall pattern matters more than any single moment. The lesson builds up over hundreds of moments like these, so one difficult morning or messy meal does not decide the outcome.

Some doubt is healthy too, as the section on healthy doubt below explains.

The Parental Balance: Self-Control Without a Loss of Self-Esteem

Erikson summed up the parents’ task in one phrase. The aim is “self-control without a loss of self-esteem” (Erikson, 1963).

The child must learn real self-control. But it must learn it without deciding that its own impulses, body or will are shameful.

That asks adults to do two things at once. They must encourage independence.

That means having the patience to let the child try until it succeeds or asks for help, rather than dressing it because that is faster. They must also protect the child against constant failure.

Failing the first half leads to taking over. Failing the second leaves the child to fail alone. Limits themselves are not the problem. The evidence section below shows that pressure is.

Why Toilet Training Is the Key Event

Erikson picked toilet training as the stage’s key event, and his reason was structural, not moral. It is the first arena where a child’s own bodily control becomes a matter of social demand.

The child has something the adult wants and can give it or withhold it. Failure is public, bodily and easy to shame. Erikson singled out shaming here as especially damaging.

That makes toilet training the stage’s question in miniature. Is my control over myself something I own, or something taken from me and used to judge me?

The point is not that toilet training determines personality.

Self-feeding, dressing and every other task pose the same question more diffusely. In each, an adult must choose between letting the child do it badly and stepping in to do it properly.

An Example: Rosa and the Coat

Imagine Rosa, who is two and a half and wants to put on her own coat. She is slow, gets an arm into the wrong sleeve, and the family is already late.

Two pulls meet in this ordinary moment. Rosa wants to prove she can do it herself. But she is bad at this, and there is an audience.

Pressed for time, one adult takes the coat and dresses her briskly: “You can’t do it, let me.” Across coat, shoes, spoon and stairs, the message hardens. She learns doubt.

A harsher adult shows irritation or ridicule, especially over a toilet accident. Now the failure is exposed as well as hers. She learns shame.

A third adult builds in five minutes and holds the coat so the sleeve is easy to find. Rosa does the part she can manage, and the fumbling is treated as ordinary.

She gets the coat on, mostly herself. She learns that attempting is safe and that she is someone who manages things.

With will established, Rosa is ready for the next stage.

In initiative vs. guilt she can move from “I can do things” to “I can plan things and lead others in doing them.” A child who learned doubt meets it unsure it may act at all.

What Are the Outcomes of Autonomy vs. Shame and Doubt?

Will, Shame and Doubt: What Each Outcome Produces

A good resolution produces will, the second of Erikson’s basic strengths, which he called virtues. Both halves of it matter: free choice and self-restraint.

Will is not simply getting your own way. It is the capacity to decide, and then to hold yourself to the decision. A child with will can assert a preference and also hold back an impulse.

A poor resolution leaves shame and doubt, which Erikson treated as two separate things. A child can carry either or both.

Shame is social and bodily. It comes from being caught out in front of others, and it tends to produce concealment and a worry about how you appear.

Doubt is private. It persists even with nobody watching, and it tends to produce hesitancy and reliance on other people’s judgement.

Why Some Doubt Is Healthy

Some doubt is useful. It brings appropriate caution and the ability to accept that you are sometimes wrong.

The healthy outcome is therefore a favourable ratio: autonomy predominates, with some doubt retained. The aim is a balance, not the elimination of doubt.

Nor is the result a verdict. Erikson held that an unfavourable resolution at this stage can be reworked later. Even Rosa, in the example above, keeps some doubt, and that is healthy.

Impulsiveness and Compulsion: Two Ways It Can Go Wrong

Because the aim is a balance, the stage can fail at either extreme:

• Impulsiveness: Autonomy with no restraint. The child is shameless and reckless, unable to hold back or accept limits.
• Compulsion: Shame and doubt take over. The person is rigidly controlled and perfectionist, preoccupied with getting everything exactly right.

Impulsiveness counts as a failed resolution even though it leans towards the positive pole. Too much autonomy fails as surely as too little.

What’s the Evidence for the Autonomy Stage?

Erikson’s own evidence was clinical and observational. Research has tested his claims mainly through two other traditions.

One studies autonomy support, meaning how far a social setting offers choice rather than pressure. It comes from self-determination theory, a modern account of autonomy as a basic psychological need. The other studies self-regulation, a child’s ability to control its own impulses and behaviour.

Autonomy Support Versus Control (Deci & Ryan, 1987)

Aim: To examine how autonomy-supportive and controlling social settings affect motivation, behaviour and wellbeing.

Method: A review, which pulls together many earlier studies. These were experiments and field studies that had changed or measured how controlling a setting was, through rewards, deadlines, surveillance, evaluative pressure and the wording authority figures used.

Results: Controlling settings weakened intrinsic motivation, the drive that comes from interest in the task itself, and self-regulation. Settings that offered choice and acknowledged the person’s view strengthened both. Limits did no harm; pressure and coercion did.

Conclusion: Autonomy is a genuine psychological need. Whether the social setting supports or thwarts it has measurable effects on motivation and self-regulation.

This gives Erikson’s stage something he never supplied: a version of its central idea that can be tested in experiments. The match with his balance is close.

Structure without control helps, and pressure harms. That is much the same as “self-control without a loss of self-esteem”, reached independently and by better methods.

Two limits apply. Most of the research used older children and adults, not toddlers, so applying it to the 18-month-to-3-year window stretches the findings beyond the ages studied. It also measures immediate effects on motivation, not the lasting personality traits Erikson claimed.

The Largest Pooled Test: Bradshaw et al. (2025)

Aim: To test whether parents’ autonomy support predicts children’s wellbeing, whether parents’ psychological control (pressuring, controlling behaviour) predicts ill-being, and whether both hold across cultures and ages.

Method: A meta-analysis, which pools the results of many studies into one overall figure. It combined 238 studies covering more than 126,000 participants, with region, national individualism, age group and sex all checked.

Results: Autonomy support had a moderate link with wellbeing (r = .30, where 0 is no link and 1 a perfect one). Psychological control had a similar link with ill-being. Each effect held independently, across regions and ages.

Conclusion: Autonomy support and psychological control are separate dimensions, each with its own effects on children, and the pattern holds well beyond the Western samples the theory was built on.

At this scale the study sits near the top of the evidence hierarchy for the whole autonomy-support literature. Two caveats apply.

For the autonomy-support link, only four studies involved children aged 11 or under, too few to analyse as a group. So that link rests mainly on teenagers and adults.

Most of the studies are also correlational, meaning they show that two things go together, not that one causes the other.

An earlier meta-analysis of 36 studies, mostly of school-age children and teenagers, found that parents’ autonomy support went with higher achievement and better psychological health (Vasquez et al., 2016).

Two large meta-analyses built from largely separate literatures reaching the same conclusion is stronger evidence than either alone.

Does Early Parenting Predict Later Self-Control?

Erikson’s stage claims that how adults respond to a toddler shapes the child’s self-control. Studies that follow the same children over time test this better than a single snapshot.

A national American study followed 938 children. Sensitive mothering and a good home environment in infancy each predicted better self-regulation at 54 months (Birmingham et al., 2017). The link ran partly through a more secure attachment.

A smaller study followed 186 low-income mother-child pairs from age two and a half until age seven.

Early attachment security predicted later self-regulation only indirectly. The route ran through how responsive mother and child were to each other at three and a half (Kim et al., 2026).

These studies are prospective, meaning they follow children forward in time rather than relying on memory. But none manipulated parenting, and none isolates the 18-month-to-3-year window.

The smaller study also drew on one low-income, mother-only sample. Both fit Erikson’s story without singling it out, because an attachment-based account predicts the same patterns.

Studies of children shaping their parents in return are covered under Causal Direction Unclear below.

Strengths

Erikson’s stage has four main strengths:

1. Real Transition: The toddler years do bring a marked shift towards independent action, and Erikson described its psychological content with unusual precision.
2. Shame Versus Doubt: Separating exposure before others from private uncertainty is a genuine analytic contribution.
3. Anticipated Evidence: His two-sided parental task matches what experiments later found independently: structure is not the problem, pressure is (Deci & Ryan, 1987).
4. Two-Sided Model: Counting too much autonomy as a failure too avoids the idea that the goal is maximum freedom, and fits real parenting.

Limitations

Before working through each one, here is a one-line summary of the main limitations:

1. Vague Mechanism: Erikson never specifies how much over-control produces shame, or how experience hardens into a lasting trait, so little could count against the theory.
2. Borrowed Evidence: No study has tested the stage’s actual claim. Support is borrowed from research on older participants and from adult questionnaires.
3. Causal Direction Unclear: A reactive toddler may draw out controlling parenting, and shared genes offer another route, so the arrow may not run only from parent to child.
4. Cultural Specificity: The stage values independent self-assertion, an ideal that suits Western, individualistic cultures better than collectivist ones.
5. The Age Window Is Too Rigid: Autonomy concerns recur across life, so treating toddlerhood as the time autonomy is settled overstates the case.

Vague Mechanism

Erikson (1964) himself described his model as a descriptive overview, not a causal explanation. The mechanism is left underspecified.

The crisis is described vividly, but nothing says how much over-control produces shame. Nor is it clear how a toddler’s experiences settle into a lasting trait.

Nothing would count as disproof, either. The theory allows any stage to be reworked later, so an autonomous adult who had a controlling childhood does not count against it. That flexibility protects the theory at the cost of its testability.

So the theory describes better than it explains. That limits the weight it can carry as a cause of lasting dispositions. Self-determination theory, compared below, offers a mechanism that can be tested: whether basic needs are satisfied or thwarted.

Borrowed Evidence

No study has tested Erikson’s actual claim: that how caregivers respond between 18 months and 3 years determines a lasting sense of autonomy. Testing it would take a study that follows children forward from infancy, with genes and temperament controlled for. Evidence of that kind is not available.

What exists is borrowed. The autonomy-support research mostly used older participants and measured short-term effects. Self-regulation research was not designed to test stages at all.

Direct measures fare no better. The main questionnaire for the stages, the Erikson Psychosocial Stage Inventory (EPSI), asks adolescents and adults about their current sense of autonomy (Rosenthal et al., 1981).

That answers the charge that Erikson’s ideas cannot be tested at all. But it measures the supposed result of the stage, decades later, not the stage itself.

The strongest long-term support concerns midlife. In the Harvard Study of Adult Development, healthier midlife development predicted better cognitive and emotional health in late life (Malone et al., 2016). That backs the idea that resolutions carry forward, but it says nothing about toddlers.

Causal Direction Unclear

One study followed 84 children who averaged about three and a half. Parents’ autonomy support and the child’s ability to hold back impulses each predicted the other over six months (Wang & Gai, 2024).

Another followed 306 children from age three. A child’s own ability to wait predicted more supportive help from mothers, though parenting’s effects were generally stronger (Klein et al., 2018).

So influence runs both ways, and temperament research helps explain why. Children differ from birth in impulsivity, persistence and reactivity. A difficult, highly reactive toddler may draw out more controlling parenting. The link between controlling parents and doubting children could therefore partly run backwards.

Shared genes offer a second route. Parent and child share genes as well as a home. So a link between parenting and a child’s outcomes cannot be read simply as the environment at work.

Neither study is large enough to settle the direction, and the designs are prospective but not experimental. Erikson’s model assumes a one-way arrow that this evidence cannot support.

Cultural Specificity

Ochse and Plug (1986) tested Erikson’s theory on over 1,800 Black and White participants in South Africa. The model fitted Black participants less well than White participants.

That fits a wider worry. Erikson’s Western, individualistic assumptions are most exposed here, because the stage does more than describe development. It values one outcome over others.

In many collectivist cultures, where the group comes before the individual, the goal for a toddler is different. It is the start of interdependence and attunement to family expectations, not independent self-assertion. A child who insists on doing everything alone is not obviously succeeding by those standards.

Erikson’s framing risks reading appropriate socialisation as developmental damage. Shame also varies by culture. In many societies it plays a more central and less pathological role than Erikson’s account, which treats it as a bad outcome, implies.

The pooled study above offers a partial defence. Its findings held across degrees of national individualism, which cuts against the charge without settling it.

The Age Window Is Too Rigid

Deci and Ryan’s research makes no stage-specific prediction. Autonomy support seems to matter at every age. That supports Erikson’s central variable, but not his claim that toddlerhood is uniquely decisive.

Self-determination theory goes further. It treats autonomy as a permanent need, not the task of one stage. The need plainly recurs in adolescence, in illness, in old age, and in any situation that removes a person’s control over their own affairs.

Erikson’s model copes with this only awkwardly. It relies on the idea that a stage can be reworked later.

The evidence therefore supports a softer claim: toddlerhood is a period when autonomy concerns are especially prominent. It does not support the claim that autonomy is settled for good in these 18 months.

Real Life Applications

Parenting

The stage gives unusually concrete guidance, most of it following from “self-control without a loss of self-esteem”. Offering bounded choices, such as the red coat or the blue one, exercises will without demanding judgement the child does not yet have. It sits between open-ended freedom and bare instruction.

Building in time for the child to attempt self-care helps too. So does separating the deed from the child. Treating an accident as an event, not a verdict on character, is the line between correcting behaviour and inducing shame.

The same reasoning argues against making toileting a site of conflict or ridicule. The reasoning is that shame attached to bodily control lasts unusually long.

Early-Years Education

Nursery and preschool practice applies the stage directly.

It is built around child-initiated activity, with accessible self-service resources and manageable self-care routines, such as children’s own coat pegs, reachable water and real cutlery.

The aim is an environment where a very young child can act successfully without adult intervention. Montessori practice is the clearest example, with its emphasis on genuine practical-life tasks a child can complete alone.

The logic is Erikson’s. A child’s belief in its own competence is built by successful independent action, so the environment should make that action possible.

Clinical Work

Therapists use the stage to make sense of shame, perfectionism, chronic self-doubt and rigid control. Compulsion, the rigid, perfectionistic outcome, is a recognisable pattern.

An explanation tracing it to an over-controlled or shaming early childhood remains common in psychodynamic practice, the wider tradition Erikson belongs to.

But that causal claim is better supported as a description than as an explanation.

Vogel-Scibilia et al. (2009) applied the stages to mental-health recovery. They framed recovery partly as reworking earlier psychosocial tasks, with restoring a sense of agency at its centre.

How Does Autonomy vs. Shame and Doubt Compare with Other Theories?

Self-Determination Theory

Self-determination theory (Deci & Ryan, 1985, 2000) is the strongest modern relative of this stage.

It names autonomy as one of three basic psychological needs, alongside competence and relatedness. Meeting these needs supports wellbeing and intrinsic motivation.

The two agree on a good deal. Both treat autonomy as a genuine need, not a preference, and both find that controlling environments damage it.

They differ in three ways. Self-determination theory treats autonomy as a lifelong need, not the business of one window. It rests on experiments, not clinical observation. And it specifies a mechanism, need satisfaction versus thwarting, where Erikson offers a description.

Where the two make different predictions, the evidence currently favours the broader claim: autonomy support matters throughout life, not only in toddlerhood.

Freud’s Anal Stage

Erikson mapped this stage onto Freud’s anal stage (Freud, 1905, 1908), and both single out toilet training. They explain why it matters in incompatible ways.

For Freud, a child’s pleasure centres on the anus, and conflict over toilet training can leave energy stuck there, which he called fixation. The result is the anal character, an adult who is orderly, stingy and obstinate.

For Erikson, toilet training matters for a different reason. It is where a child’s control over its own body first becomes a social transaction.

What the child takes away is a stance towards its own agency, meaning its sense of being able to act for itself. It is not a lodged quantity of drive energy.

See Erikson vs. Freud for the full comparison.

References

Birmingham, R. S., Bub, K. L., & Vaughn, B. E. (2017). Parenting in infancy and self-regulation in preschool: An investigation of the role of attachment history. Attachment & Human Development, 19(2), 107–129. https://doi.org/10.1080/14616734.2016.1259335

Bradshaw, E. L., Duineveld, J. J., Conigrave, J. H., Steward, B. A., Ferber, K. A., Joussemet, M., Parker, P. D., & Ryan, R. M. (2025). Disentangling autonomy-supportive and psychologically controlling parenting: A meta-analysis of self-determination theory’s dual process model across cultures. American Psychologist, 80(6), 879–895. https://doi.org/10.1037/amp0001389

Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. Plenum Press.

Deci, E. L., & Ryan, R. M. (1987). The support of autonomy and the control of behavior. Journal of Personality and Social Psychology, 53(6), 1024–1037. https://doi.org/10.1037/0022-3514.53.6.1024

Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01

Erikson, E. H. (1950). Childhood and society. Norton.

Erikson, E. H. (1963). Childhood and society (2nd ed.). Norton.

Erikson, E. H. (1964). Insight and responsibility. Norton.

Freud, S. (1905). Three essays on the theory of sexuality (Standard Edition, Vol. 7, pp. 123–246). Hogarth Press.

Freud, S. (1908). Character and anal erotism (Standard Edition, Vol. 9, pp. 167–176). Hogarth Press.

Kim, J., Bendel-Stenzel, L. C., & Kochanska, G. (2026). Examining self-regulation as a heterogeneous construct: A nuanced approach to understanding its relational antecedents. Attachment & Human Development, 28(3), 292–311. https://doi.org/10.1080/14616734.2026.2655283

Klein, M. R., Lengua, L. J., Thompson, S. F., Moran, L., Ruberry, E. J., Kiff, C., & Zalewski, M. (2018). Bidirectional relations between temperament and parenting predicting preschool-age children’s adjustment. Journal of Clinical Child & Adolescent Psychology, 47(sup1), S113–S126. https://doi.org/10.1080/15374416.2016.1169537

Malone, J. C., Liu, S. R., Vaillant, G. E., Rentz, D. M., & Waldinger, R. J. (2016). Midlife Eriksonian psychosocial development: Setting the stage for late-life cognitive and emotional health. Developmental Psychology, 52(3), 496–508. https://doi.org/10.1037/a0039875

Ochse, R., & Plug, C. (1986). Cross-cultural investigation of the validity of Erikson’s theory of personality development. Journal of Personality and Social Psychology, 50(6), 1240–1252. https://doi.org/10.1037/0022-3514.50.6.1240

Rosenthal, D. A., Gurney, R. M., & Moore, S. M. (1981). From trust to intimacy: A new inventory for examining Erikson’s stages of psychosocial development. Journal of Youth and Adolescence, 10(6), 525–537. https://doi.org/10.1007/BF02087944

Vasquez, A. C., Patall, E. A., Fong, C. J., Corrigan, A. S., & Pine, L. (2016). Parent autonomy support, academic achievement, and psychosocial functioning: A meta-analysis of research. Educational Psychology Review, 28(3), 605–644. https://doi.org/10.1007/s10648-015-9329-z

Vogel-Scibilia, S. E., McNulty, K. C., Baxter, B., Miller, S., Dine, M., & Frese, F. J. (2009). The recovery process utilizing Erikson’s stages of human development. Community Mental Health Journal, 45(6), 405–414. https://doi.org/10.1007/s10597-009-9189-4

Wang, S., & Gai, X. (2024). Bidirectional relationship between positive parenting behavior and children’s self-regulation: A three-wave longitudinal study. Behavioral Sciences, 14(1), Article 38. https://doi.org/10.3390/bs14010038