Infants progress through predictable phonological stages—cooing (2-3 months), babbling (4-6 months), canonical babbling (6-10 months), variegated babbling—gradually refining motor control and acoustic feedback for phoneme production. Exposure to native language phonemes during sensitive periods (first year) shapes speech perception and production abilities, with non-native phoneme discrimination declining after 12 months.
Listen to audio/video samples of infant vocalizations at different ages; compare early undifferentiated sounds to increasingly controlled phonetic productions and native language-specific patterns.
Babbling is not random vocalization; it shows structured motor learning of articulatory patterns. Children's phonological errors are not laziness but reflect learning of motor sequences and rule abstraction from the language they hear.
From your study of language acquisition, you know that infants enter the world as universal phoneticians—capable of discriminating the phonemic contrasts of any human language. This broad sensitivity does not last. During the first year of life, a process of perceptual narrowing occurs: the infant's speech perception system tunes itself to the phoneme categories of the ambient language, becoming better at distinguishing native contrasts and worse at distinguishing non-native ones. Japanese infants, for example, are initially as capable as English infants at distinguishing /r/ from /l/, but lose this ability by 12 months if they are raised in a Japanese-speaking environment. This is a direct product of the sensitive period for phonological development you studied: the language system is primed for input, and the input it receives sculpts the perceptual categories it will use for life.
Production lags behind perception by several months and follows a predictable sequence. At 2–3 months, infants produce cooing—vowel-like sounds made at the back of the mouth with little articulatory control. By 4–6 months, marginal babbling appears: more consonant-like sounds in less organized sequences. The landmark transition occurs around 6–10 months with the onset of canonical babbling—repeated consonant-vowel syllables like "ba-ba-ba" or "ma-ma-ma." This stage is significant because it reflects genuine motor learning: the infant is practicing the rhythmic jaw and lip movements that underlie syllable production. Deaf infants who have sufficient auditory input before this window begin canonical babbling on schedule; those with severe hearing loss do not, demonstrating how acoustic feedback drives the motor learning process.
After canonical babbling, infants enter variegated babbling, in which the syllable sequences become more varied and begin to approximate the prosodic contours—rhythm and intonation—of the native language. Parents often comment that late-stage babbling sounds like "conversation without words." This is not an accident: the infant is practicing the melodic structure of language before they have mastered its sound categories. Phonological errors in early word production (like "wabbit" for "rabbit" or "pasketti" for "spaghetti") are not random laziness—they reflect systematic phonological simplification strategies the child applies because producing complex consonant clusters or less familiar phonemes exceeds current motor capacity. The child has already perceived the correct form; the production system simply has not caught up.
The entire arc of phonological development illustrates a general principle in developmental psychology: perception leads production, and input during sensitive periods shapes both. The infant is not passively recording language but actively building a phonological grammar from statistical regularities in the input—which sounds appear together, which contrasts are contrastive versus allophonic, which sequences occur at syllable boundaries. By the time a child produces their first words, they already have a sophisticated implicit phonological system that will guide how they perceive and produce speech for the rest of their lives.