How Children Learn Language: Stages, Theories and Evidence

Almost every typically developing child learns to speak a native language by about age five, without formal lessons. That fact is one of the most studied in developmental psychology. Researchers want to know what children hear, what they can perceive, and how much of the process depends on innate abilities versus experience.

This guide follows the usual sequence of milestones, summarizes the main theories, and separates well-replicated findings from open questions.

Before words: perception and babbling

Newborns are already tuned to speech. In the first days of life, infants prefer their mother's voice and listen longer to their native language than to a foreign one. Peter Eimas and colleagues showed in 1971 that very young infants can discriminate speech sounds that differ in ways important to adult languages, even sounds that their own language does not use.

Over the first year, perception narrows. Infants lose the ability to tell apart some sound contrasts that their language never uses and keep the ones it does. This process of perceptual tuning is well documented, and it is one reason adults often struggle to hear some foreign sound distinctions.

Babbling begins around six months. At first infants produce a wide range of sounds, including sounds from languages they have never heard. Over the following months, babbling becomes more restricted and begins to resemble the sounds in the child's environment. Deaf infants exposed to sign language also babble with their hands, which suggests that the drive to produce structured communicative output is not tied to speech alone.

First words and the vocabulary spurt

Most children say their first recognizable words around twelve months. Early vocabularies are small and often include names of familiar people, food and objects, along with social words such as "bye" or "no". Early words are often used with overextension, where a word covers a broader category than it should, such as calling every four-legged animal "dog".

Between about eighteen and twenty-four months, many children enter a vocabulary spurt, learning new words much faster than before. The exact timing varies a lot between children. Parent report data, such as the MacArthur-Bates Communicative Development Inventories, has been used to track typical ranges in many languages.

Why the spurt happens

No single explanation is accepted by everyone. Some researchers point to improved memory and categorization. Others point to growing skill at using social cues, such as where an adult is looking, to work out what a new word refers to. Most current accounts combine several factors.

Grammar: from two words to full sentences

Around eighteen to twenty-four months, children start combining words. Early two-word utterances such as "more juice" or "daddy go" show a rudimentary sense of structure. Roger Brown's influential studies of three children in the 1970s documented a steady progression from telegraphic speech, which drops small grammatical words, toward full sentences with inflections, auxiliaries and questions.

Children also make errors that reveal their hidden rules. The overregularization of past tense verbs, such as "goed" or "breaked", is a classic example. Children who say "goed" have not simply copied adults. They have extracted a general rule and applied it too broadly, then gradually learn the exceptions.

Theories of acquisition

The main theories differ on how much of language is built in. Noam Chomsky argued in the late 1950s and 1960s that the speed and uniformity of acquisition, along with the poverty of the input children hear, suggest a biological capacity for language, often called universal grammar. This view proposes innate constraints that narrow the possibilities a child considers.

Usage-based theories, associated with researchers such as Michael Tomasello, argue that children build grammar gradually from patterns in the language they hear. On this view, general learning abilities, social understanding and statistical pattern detection do most of the work. Both camps accept that children are skilled learners, but they disagree about how much of that skill is specific to language.

The role of input

Children learn more from conversations that include them directly than from speech they overhear. Studies by Catherine Snow and others have found that the amount of parent talk, the richness of vocabulary, and the amount of back-and-forth turn-taking all predict later language outcomes. Reading aloud to children also shows strong links to vocabulary growth, although the size of the effect varies across studies.

The critical period question

Is there a time after which language learning becomes much harder? The case of Genie, a girl discovered in 1970 after years of severe isolation, is often cited. She acquired some vocabulary but never fully mastered grammar. Her case is hard to interpret because her deprivation was extreme and involved many other factors.

Eric Lenneberg proposed in 1967 that a critical period for first language ends around puberty. The evidence for a sharp cutoff is less clear than it once seemed. Studies of people who learned a second language at different ages suggest that the ability to reach native-like grammar declines gradually with age rather than switching off, and that the rate of decline depends on the language and the learner.

Bilingual children

Growing up with two languages is normal and common. Bilingual children often reach the same milestones as monolingual children, although their vocabulary in each language may be smaller at a given age when measured separately. When both languages are counted together, their total vocabulary is often comparable.

The idea that bilingualism causes confusion or delay is not supported by the bulk of the research. Mixing words between languages is a normal pattern that usually decreases as children learn which speaker uses which language.

Milestones at a glance

The table below gives approximate ages for common language milestones. The ranges are wide, and a child who is somewhat behind is not necessarily a cause for concern. Persistent delays should be assessed by a clinician.

Age	Typical language milestone
0 to 3 months	Prefers speech sounds, responds to voices
6 months	Babbles with consonant-vowel strings
12 months	First recognizable words
18 to 24 months	Vocabulary spurt, two-word combinations
3 years	Sentences of three or more words, many questions
5 years	Mostly adult grammar, with some errors

What parents and caregivers can do

Research on child language points to a few practical habits. Talking with a child during daily routines, such as meals, bath time and walks, gives many chances to hear new words in context. Responding to a child's babbling or gestures with a matching word or question invites more turn-taking, which is a strong predictor of later vocabulary. Naming objects and actions while pointing at them helps children connect words with their referents. Reading picture books together, and talking about the pictures rather than only reading the text, also supports vocabulary growth.

These habits are supported by the observational and intervention studies described above, but they are not a guarantee of any specific outcome. Children differ in temperament, hearing, health and many other factors. Parents who are worried about a child's speech should raise the concern with a pediatrician. Hearing checks are an important first step, because hearing loss can affect speech development even when it is mild.

Key takeaways

Language learning starts before birth in the sense that infants already prefer speech, and it moves through predictable stages of babbling, first words and grammar.

The vocabulary spurt and early grammar both show children extracting patterns, not just copying adults.

The nature versus nurture debate is still active. Most researchers agree that both a learning capacity and rich input matter.

Conversation with children predicts language growth better than passive exposure. Bilingual children generally develop language on a normal schedule.