Lead
One afternoon, your child sits on the front step putting on shoes and announces, without particular prompting: "I'm a big kid now." No younger sibling has arrived. No one seems to have told them to say this. Yet from that day, they begin to refer to themselves a little differently.
Then, the same week, they want you to brush their teeth. At drop-off, they demand to be carried. Progress and regression, alternating without obvious pattern.
Ages 5–6 mark a period when children begin to describe "who I am" with more distinct outlines than before. It is a turning point in a long developmental line that began with recognizing oneself in a mirror as an infant. This article traces that line through the classical research, to understand what it means for a child's self-concept to take shape.
A long line beginning in front of a mirror
The most frequently cited starting point in self-recognition research is the mirror self-recognition test: a developmental test assessing whether an infant recognizes their own reflection, indicating an emerging sense of self devised by Lewis and Brooks-Gunn (1979) [1]. The procedure is simple: while the child is not watching, a small mark is placed on their nose. Then a mirror is introduced. A child who has achieved self-recognition will reach for their own nose — not for the reflection's nose. This behavior becomes reliable at around 18–24 months, and virtually all typically developing children show it by 24 months [1].
What emerges in that moment is recognition of the most fundamental self — the bodily self, the physical "I." Howe and Courage argued that it is this bodily self (they called it the "cognitive self") whose establishment is the precondition for autobiographical memory [2]. Before a child has a stable self as a reference point, events cannot be organized as things that happened to me, and they are not consolidated into retrievable long-term memory. This reasoning has been offered as one account of infantile amnesia: the normal inability of adults to retrieve autobiographical memories from before approximately age 3 — why most adults cannot reliably recall events from before ages 3–4 [2].
Over the years between that first mirror moment and ages 5–6, children layer onto the bodily self a mental self (preferences, abilities, traits) and a social self (how others see me). Susan Harter's self-system theory — developed across her career and synthesized in her 1999 book — traces this layering through the distinction between the "I-self" (the self as agent, the knower) and the "Me-self" (the self as known object) [3]. Both evolve in parallel across early childhood.
What changes in self-description at ages 5–6
Ask a child ages 2–4 to describe themselves, and the answers tend to cluster around external, observable features: "I like drawing." "I have red shoes." "I run fast." Harter characterized this as concrete, observable self-description: the young child does not yet understand the self as a bundle of abstract traits [3].
By ages 5–6, this begins to shift. Children start offering descriptions that operate at a slightly higher level of abstraction: "I'm good at drawing." "I'm kind to my friends." "I'm a little shy." Rather than listing behaviors, they begin attributing characteristics [3]. At the same time, the vocabulary of social comparison enters: "I'm better at this than Yui" or "I can't do that as well as she can." Social comparison increases markedly from the late preschool years through middle childhood, and its appearance in self-description has been repeatedly documented [3].
Running in parallel is the organization of autobiographical memory. Nelson and Fivush (2004), in a major theoretical paper in Psychological Review, proposed that autobiographical memory does not emerge as a single capacity but develops gradually across the preschool years through the interaction of several components: basic memory ability, language, memory talk: caregiver–child conversations about shared past events that help structure autobiographical memory with caregivers, time understanding, and self-other understanding [4]. Talking with a caregiver about a shared past event — "What did we do at grandma's house for the holidays?" — and doing so repeatedly transforms the event into a narrative with temporal structure, a story about something that happened to me. Ages 5–6 are a period when the vocabulary and structure for this kind of narrative become considerably richer [4].
Seen through this lens, the announcement "I'm not a baby anymore" is not just a growth declaration. It is an act of self-organization: the child is distinguishing a past self (the baby they were) from a present self (the big kid they are now), placing temporal value on the latter. It is a piece of the larger work of assembling a self-concept with a timeline.
Regression is not failure — it is part of the structure
Here is a distinction worth making carefully. The day a child declares "I'm a big kid" and the night they cry for a parent to stay — these are not contradictory.
Harter's research shows that preschool-age self-evaluation is highly unstable, varying considerably with context [3]. Young children also tend to evaluate themselves in an unrealistically positive direction — overconfidence is typical in early childhood, not a character flaw. The instability and the positive bias together characterize the period before a more stable and realistic self-evaluation consolidates in middle childhood [3].
This means that a child who declares independence on Tuesday and collapses at the gate on Wednesday is not going backward. What you are seeing is a self-image that has not yet hardened — still soft, still responsive to situational demands. And this softness is precisely what will allow gradual, smooth change over the years ahead.
The phrase "But you're a big kid now" — said as encouragement — can carry more weight than intended when directed at a self-concept that is still in formation. A child who does not yet have a stable self-image cannot simply will themselves into one; being pushed toward it prematurely can reduce the safety of regressing, and paradoxically, children who cannot safely regress tend to have fewer opportunities to renarrate themselves forward.
What parents can do: vocabulary and record-keeping
Two things distill most of what the research suggests for parents of five- and six-year-olds.
The first is not narrowing the child's self-descriptive vocabulary. Rather than fixing a characteristic — "you're such a shy one" — try connecting behavior to context: "You concentrate so hard when you're drawing." "You really want a hug when you're tired." This stays closer to the concrete, observable level of self-description that Harter found is developmentally appropriate for this age [3], without prematurely locking the child into a single trait label.
The second is keeping accessible materials for revisiting the past. As Nelson and Fivush showed, autobiographical memory is organized through conversation — through memory talk [4]. When a child says "I want to see pictures of me when I was a baby," having that record within easy reach increases the density and richness of the narrative that can be built around it. A chronological record in something like Memori is, in the immediate term, a memory aid for the parent. In the longer term, it is material for a child's own self-narrative — the archive from which they will eventually piece together the story of who they were, and therefore who they are.
And holding space for regression matters. The coexistence of "I'm a big kid" and "carry me" in the same week is a sign that the self-concept is still forming — still soft enough to grow into the shape it needs to take.
Summary
"I'm not a baby anymore" arrives as a declaration, but it rests on a long developmental line that begins with mirror self-recognition in infancy [1,2]. By ages 5–6, the self-concept is moving from concrete self-descriptions toward more abstract trait attributions, from context-dependent self-evaluation toward greater stability, and absorbing social comparison while doing so [3]. In parallel, autobiographical memory is gaining temporal structure through caregiver-guided memory talk [4].
Declaration and regression are not contradictions. They are both characteristic of this period — which is why it is also the period when words that will never be said again get said, and moments that will never repeat themselves are happening right now. Keeping a record of them is the quietest form of care.
References
- Lewis M, Brooks-Gunn J. Social Cognition and the Acquisition of Self. New York: Plenum Press; 1979.
- Howe ML, Courage ML. The emergence and early development of autobiographical memory. Psychol Rev. 1997;104(3):499–523. doi:10.1037/0033-295X.104.3.499. PMID: 9243962.
- Harter S. The Construction of the Self: A Developmental Perspective. New York: Guilford Press; 1999.
- Nelson K, Fivush R. The emergence of autobiographical memory: a social cultural developmental theory. Psychol Rev. 2004;111(2):486–511. doi:10.1037/0033-295X.111.2.486. PMID: 15065919.
- Howe ML, Courage ML. On resolving the enigma of infantile amnesia. Psychol Bull. 1993;113(2):305–326. doi:10.1037/0033-2909.113.2.305.
- Fivush R, Nelson K. Culture and language in the emergence of autobiographical memory. Psychol Sci. 2004;15(9):573–577. doi:10.1111/j.0956-7976.2004.00722.x.