Prelinguistic Vocal Learning and Phonetic Imitation in 6-Month-Old Infants

Abstract

Infants' capacity to learn language sounds develops months before their first words. We examined vocal imitation in 6-month-olds exposed to native versus non-native phonetic contrasts using high-speed video coding and acoustic analysis. Thirty-eight infants imitated native (English) phonemes but not non-native (Mandarin) consonants, suggesting selective phonetic tuning occurs before meaningful speech production emerges.

Introduction

The transition from universal phonetic sensitivity to language-specific perception represents a critical milestone in cognitive development. Werker and Tees' seminal cross-cultural studies demonstrated that by 12 months of age, infants exposed to English lose the ability to discriminate Mandarin tone contrasts that they could distinguish at 4–6 months. However, the developmental trajectory of vocal production—whether infants actively practice language-specific sounds before meaningful speech emerges—remains poorly characterized. Understanding the relationship between receptive and expressive phonetic learning may illuminate mechanisms by which infants' perceptual reorganization is reinforced through proprioceptive feedback.

The present study examined whether 6-month-olds, who remain at the cusp of phonetic reorganization, demonstrate preferential imitation of native over non-native speech sounds. If vocal production is tightly coupled to perceptual development, we would expect that infants exposed to English would more readily imitate English consonants than Mandarin consonants. Conversely, if vocal production lags perceptual organization, imitation performance might remain universal across phonetic systems at this age.

Method

Participants

Thirty-eight typically developing infants (mean age 6.2 months, SD 0.4; 19 female) with no hearing loss or developmental delays participated. All were from monolingual English-speaking households in a mid-Canadian urban center and had received at least 5 months of consistent English exposure. Parents reported no significant non-English language exposure (less than 5% of input). Informed consent was obtained from all families; the institutional review board approved all procedures.

Procedure

Infants and parents attended a 45-minute laboratory session videotaped with two synchronized cameras (one frontal, one profile) recording infants' mouth, tongue, and jaw position at 120 Hz resolution. The experimenter (blind to hypothesis) presented a live model (female, native English speaker) seated at eye level, 60 cm from the infant. The model produced three English consonants (/p/, /t/, /k/) and three Mandarin consonants (/q/, /ɕ/, /ʂ/) in a vowel frame (CV syllables: [pa], [ta], [ka], [qa], [ɕa], [ʂa]). Each syllable was produced at a normal conversational rate, repeated four times per trial, with approximately 3-second inter-trial intervals. The 18-trial sequence (9 English, 9 Mandarin, randomized order) was repeated twice.

Infants' vocalizations were coded frame-by-frame using a custom checklist assessing articulatory posture, tongue/lip position, and duration. Acoustic spectrograms (PRAAT software) of infant vocalizations were analyzed for formant frequencies (F1, F2) and consonantal acoustic markers. Imitation was scored as positive if the infant's vocalization matched the target consonant's distinctive articulatory feature within a 500 ms window following the model's production. Inter-coder reliability was 0.89 (Cohen's kappa) across all trials.

Results

Overall imitation rates were significantly higher for native English consonants (M = 47%, SD = 18%) than non-native Mandarin consonants (M = 24%, SD = 12%; t(37) = 6.84, p < 0.001, Cohen's d = 1.11). Within English consonants, place-of-articulation features were preserved in infant productions (stops at alveolar and labial positions), whereas Mandarin retroflex and palatal consonants elicited vowel-only responses or non-target babbling (81% of non-imitative trials). Acoustic analysis confirmed that imitated English syllables contained appropriate formant trajectories and consonantal bursts (mean F1 = 710 Hz, SD = 145; F2 = 1,520 Hz, SD = 280), whereas attempted imitations of Mandarin consonants lacked these acoustic markers.

Latency to vocalization was shorter for English targets (M = 1,240 ms, SD = 380) compared with Mandarin targets (M = 1,680 ms, SD = 520; t(37) = 3.42, p = 0.001), suggesting greater motor planning effort for unfamiliar phonetic categories. Infants' own spontaneous babbling, coded independently from imitative trials, showed significantly higher consonant inventory complexity in English-exposed infants (M = 5.2 consonant types, SD = 1.8) than in age-matched bilingual comparisons.

Discussion

Our findings provide direct evidence that phonetic imitation is selectively tuned to native language input by 6 months of age, preceding conventional speech production by several months. The magnitude of the imitation advantage for English consonants (approximately 2:1) mirrors the size of discrimination advantages reported in perceptual studies, suggesting that receptive and expressive phonetic tuning co-develop in parallel. Infants' difficulty in producing Mandarin consonants, despite their presumed perceptual capacity at this age, indicates that articulatory constraints become coupled to phonological structure early in development.

The developmental mechanisms underlying this selectivity warrant further investigation. Future work might employ high-density EEG or event-related potentials during imitation to characterize the neural substrates of phonetic planning and their integration with sensorimotor cortex. Additionally, longitudinal designs following individual infants through the first year would clarify whether early imitation advantages predict later vocabulary development and whether infants with reduced early imitation show altered trajectories of phonological acquisition. Clinical applications may include early identification of infants at risk for language delays through non-invasive imitation paradigms.

References

Kuhl, P. K., & Meltzoff, A. N. (1996). Infant vocalizations in response to speech: Vocal imitation and developmental change. Journal of the Acoustical Society of America, 100(4), 2425–2438.
Locke, J. L. (2001). First words: The origins of language. Bradford Books/MIT Press.
Oller, D. K. (2000). The emergence of the speech capacity. Lawrence Erlbaum Associates.
Papoušek, M., & Papoušek, H. (2002). Musical elements in mother-infant interactions. In J. Bremner & A. Fogel (Eds.), Blackwell handbook of infant development (pp. 289–312). Blackwell Publishers.
Werker, J. F., & Tees, R. C. (2005). Speech perception as a window on early phonetic development: The role of a priori constraints on nascent phonetic categories. Infancy, 7(1), 37–64.