Lead
"Yuta do it." A two-year-old refers to herself by name. From an adult's vantage point this sounds oddly formal, but in childcare settings and homes with toddlers it is entirely ordinary. And then, somewhere in the approach to age three, it shifts — "I do it," "me too" — and the third-person reference quietly disappears.
The shift is not accidental and it is not a sign of incomplete discipline. It marks a developmental stage in which self-concept and first-person pronouns are becoming aligned — a point at which the child's ability to see herself from the outside is transitioning to a stable view from the inside. This phenomenon receives surprisingly little attention in parenting guides, despite being one of the better-documented transitions in developmental psychology.
What "third-person self-reference" actually is
The behavior is called "third-person self-reference" in the developmental literature. Researchers treat it as a functional stage in developing self-concept — not a passing error or a gap in language learning, but a form that works within the child's current cognitive architecture.
Mirror self-recognition research illuminates the mechanism. Lewis and Ramsay (2004) followed 66 infants longitudinally at 15, 18, and 21 months, observing mirror self-recognition, first-person pronoun use, and pretend play at each visit [1]. The results were clear: infants who demonstrated visual self-recognition in the mirror used significantly more first-person pronouns ("I," "me") and showed more advanced pretend play than infants who had not yet reached self-recognition [1]. These three abilities tended to emerge together around 18 months. The authors described the pattern as reflecting the emergence of "a metarepresentation: a mental representation of one's own thoughts, perspective, or identity — being able to think about oneself as an object of thought of the self" in the second year [1].
The implication: a child who refers to herself as "Yuta" is not using language incorrectly. She may simply not yet have a stable internal first-person vantage point to draw on. The name "Yuta" is a known referent — she has heard it applied to herself repeatedly. "I" is something else: a word that points inward to a self that is still solidifying.
Why first-person pronouns are genuinely hard to acquire
The difficulty of learning "I" is not arbitrary. Unlike nouns such as "chair" or "dog" — which point to the same object regardless of who is speaking — first-person pronouns are what linguists call shifting reference: "I" means the current speaker, so the same word refers to a different person in every conversation. Understanding "I" requires grasping that the perspective of the speaker is what the word tracks, not a fixed external referent. To use "I" correctly, a child must distinguish her own perspective from other people's, and understand that those perspectives shift with who is speaking.
Tomasello (1995) argued that the stable acquisition of first-person pronouns is tightly linked to joint attention: the shared focus of two people on the same object or event, a key building block of early social communication — the ability to coordinate attention with another person toward a shared object or event — and to understanding other people as intentional agents [2]. First-person pronoun use is a language problem and a social cognition problem simultaneously. A child who calls herself "Yuta" when she means "I" may be expressing something closer to "the agent in this situation" than to the first-person pronoun as adults use it.
Budwig's (1989) analysis of young children's self-reference showed that children divide into two patterns in the transition period: some rely primarily on their own name in subject position, while others use multiple self-reference forms [3]. The variation is in the trajectory, not the destination.
What changes between two and three
Third-person self-reference is most frequent around age two. As the first half of the third year unfolds, first-person pronouns stabilize and the name-based reference fades.
Lewis and Ramsay (2004) describe the 18-to-24-month window as the period in which the metarepresentation of the self undergoes significant development [1]. Once the internal first-person perspective is stable, "I" has a fixed thing to point to, and the child uses it consistently. The transition happens without explicit correction from adults — which is itself a developmental point worth noting. You cannot accelerate it by drilling pronouns; you do not need to, because the child's emerging self-concept drives the shift on its own timeline.
There is no evidence that rushing this transition — correcting every third-person instance — confers developmental benefit. The corrective impulse is understandable, but from the child's perspective, the name-based reference is doing adequate communicative work. The upgrade to first-person pronouns will happen as the underlying cognitive capacity arrives.
When to pay attention
The ordinary picture is third-person self-reference declining through the third year as first-person pronouns consolidate. A different pattern is worth bringing to a professional's attention.
If a child past age three continues to refer to herself exclusively by name and rarely or never uses first-person pronouns, or if there are pronoun reversals — using "you" when meaning "I," for instance — a consultation with a speech-language pathologist or developmental specialist is reasonable. Pronoun reversal: using "you" when meaning "I," or vice versa — studied as one possible feature in autism spectrum communication patterns has been studied as one feature of language use in autism spectrum disorder; Charney (1980) proposed a social-communicative explanation, framing it as difficulty tracking shifting discourse roles [4]. The presence of this one feature is not diagnostic — many children with ASD show typical pronoun development, and pronoun reversal can occur in typical development — but as part of a broader pattern of atypical social communication, it is relevant.
Raising the question with a clinician is appropriate if you observe it as part of a broader cluster: limited eye contact, reduced joint attention, restricted interests, or delayed language more generally. This is not a list to apply anxiously to every child who calls herself "Yuta" at age two — it is context for when the third-year shift does not arrive.
Recording the transition
This developmental moment has a quality that makes it worth documenting. The first appearance of "I," used in a context where it clearly refers to the child herself, is a marker of the metarepresentational shift Lewis and Ramsay describe. It tends to emerge in small, unannounced moments — a request, a complaint, a bid for a turn.
If you note in a parenting log the month when first-person pronouns began appearing consistently, and in what kinds of utterances, you have a lightweight developmental record. Looking back at that record from age four or five — when the same child is fluently saying "I think," "I want," "I remember" — closes the loop on one of the less obvious but cognitively significant transitions of the toddler years.
Summary
A toddler who calls herself by name is operating within a developmental stage, not correcting wrongly or speaking carelessly. Lewis and Ramsay (2004) showed that mirror self-recognition, first-person pronouns, and pretend play emerge together around 18 months as expressions of an emerging metarepresentation of the self [1]. The difficulty of first-person pronouns is genuine — they require tracking shifting speaker perspective, which demands social cognitive capacities that are still developing [2]. Through the third year, as that internal perspective stabilizes, first-person pronouns take over and name-based self-reference fades without parental intervention.
Observation serves this stage better than correction.
References
- Lewis M, Ramsay D. Development of self-recognition, personal pronoun use, and pretend play during the 2nd year. Child Dev. 2004;75(6):1821–1831. doi:10.1111/j.1467-8624.2004.00819.x. PMID: 15566382.
- Tomasello M. Joint attention as social cognition. In: Moore C, Dunham PJ, editors. Joint Attention: Its Origins and Role in Development. Lawrence Erlbaum Associates; 1995. p. 103–130.
- Budwig N. The linguistic marking of agentivity and control in child language. J Child Lang. 1989;16(2):263–284. doi:10.1017/S0305000900010400. PMID: 2793635.
- Charney R. Pronoun errors in autistic children: Support for a social explanation. Br J Disord Commun. 1980;15(1):39–43. doi:10.3109/13682828009011492. PMID: 7426055.
- Diesendruck G, Shatz M. Two-year-olds' recognition of hierarchies: evidence from their interpretation of the semantic properties of novel words. Child Dev. 2001;72(5):1675–1681. doi:10.1111/1467-8624.00368. PMID: 11699014.