Guest Post by: Maithri Sivaraman, Ph.D., BCBA
Maithri Sivaraman is an Assistant Professor of Psychology and Education at Teachers College, Columbia University. She studies early language acquisition in neurotypical and neurodivergent children.
The early childhood market is overrun with toys and videos, many of them AI-powered, and most of them miss the point entirely. A recent report found that AI toys activated unexpectedly, they ignored children’s bids for interaction, used language far too complex for young children, and perhaps most importantly, disrupted the open-ended, child-led play that leads to learning and skill acquisition. Instead of fostering contingent, back-and-forth exchanges that decades of research have identified as foundational, these products replace genuine interactions with simulated stimulation and canned responses.
This is a symptom of a larger problem: a growing discord between the world we curate for children and how they actually learn. The blueprint for early learning is in contingent interactions between children and their caregivers, and apps or “optimized” toys do not provide this.
Language acquisition, at its core, is a social exercise. It unfolds through a series of moments in which a child and a caregiver respond to one another, each influencing the other’s next move. What adults think of as a simple interaction has far reaching consequences for a young child. The choreography is subtle yet powerful. A fourteen-month-old spots a brightly colored image on the wall. He doesn’t have the word for it, but he leans forward, and his mother immediately tracks his gaze: “Ohhh, you saw the butterfly! There’s the butterfly.” In that small exchange, the child experienced his communicative power. The mother, for her part, seized a fleeting moment that captured the child’s attention, used short sentences, and repeated the key word. Early interactions like this rely on a synchronized, multimodal package of caregiver behaviors: vocalizations, mutual gaze, pointing, touch, and simple words, all timed to one another (Sivaraman & Pelaez, 2026).
Central to this social dance is a specific mode of speech that caregivers adopt when addressing infants. Infant-directed speech, often called parentese, is marked by a higher pitch, elongated vowels, exaggerated intonation, a slower tempo, and simple vocabulary. Sounds are clearer, longer, more exaggerated when compared to standard speech. A parent mid-conversation with a colleague will catch their baby’s eye and shift registers instantly: “Hii sweet pea! Look at those tiny toes!” This speech style is also characterized by exaggerated articulatory gestures and social affect (Golinkoff et al., 2015). But parentese is not a lecture delivered at an infant. It is only one half of a feedback loop. Infants smile and vocalize more in response to it, and when they do, caregivers keep going. Both participants are actively influencing the interaction.
Another critical aspect of early interactions is contingent vocal imitation. When a baby coos or babbles, a responsive caregiver mirrors that sound back immediately. A six-month-old lets out a bright “aaah” and the parent returns an enthusiastic “aaah!”. Studies show that adult imitation reliably increases infant vocalizations and gradually draws those sounds closer to adult speech patterns (Pelaez et al., 2011). Imitation seems to be particularly salient, perhaps due to the high degree of correspondence between the infant’s own behavior and that of the caregiver.
In infant-caregiver interactions, both the content and the timing of a response carry independent weight. The quality of a caregiver’s response, its multimodal richness, tonal quality, and content adapted to the child’s level, control what and if an infant learns. Infant-directed speech and imitation are examples of high-quality responses to infants’ bids. But these responses must also arrive within a precise window for it to be meaningful to the child.
Research has shown that poorly timed caregiver’s responses disrupt the interactional flow that gives meaning to these exchanges. When a response is delayed, or it disregards the child’s vocalizations, we are effectively severing the feedback loop that drives vocal learning. Studies have illustrated just how sensitive infants are to breaks in contingency: when caregivers suddenly become unresponsive or respond noncontingently, infants show signs of distress immediately; withdrawing their gaze, reducing vocalizations, and attempting repeated bids to re-engage (Hirsh et al., 2014; Nadel et al., 2005). These findings underscore that in infant-caregiver interactions, as with comedy, timing is everything.
Both contingent infant-directed speech and vocal imitation are empirically validated interaction patterns that apply directly into clinical practice. Embedding these into early intervention programs results in higher vocalization rates, sometimes combined with increased echoic responses, and a facilitated transition from babbling to first words in children with language delays and developmental disabilities (Fiani et al., 2021; Gazdag & Warren, 2000; McDaniel, 2025; Sun et al., 2025). By intentionally integrating these strategies into everyday routines, ordinary moments like a diaper change, a meal, or a walk, we create additional learning opportunities.
Five Evidence-Based Considerations to Support Early Vocal Development
Below, I will outline five broad considerations borne out of the research literature for practitioners and caregivers interested in adopting these interaction strategies with infants and young children with developmental delays.
- Timing matters: Contingent interaction depends on immediacy. Responding within a few seconds of a vocalization has been shown to be key for subsequent increases in children’s vocalizations.
- Use the full package: A vocal response alone does not represent the full picture. Pairing a vocal response with a warm smile, a gentle touch, or an animated expression creates richer, more meaningful interactions. Human connection is multimodal, and layering these cues together amplifies their effect.
- Wait, watch, and follow the child’s lead: Rather than steering toward a planned agenda, start at the child’s current developmental level and let them direct the interaction. Following a child’s lead can be harder than it sounds. Infants do not always loudly announce what has captured their attention. Learning to read non-vocal cues is a skill, and like any parent knows, it takes practice and can be learned and honed. Try a free video library for models of infant vocal behaviors.
- Read the child’s signals: Every infant is different, and every day is different. One child might delight in infant-directed speech delivered in a high-pitched, exaggerated voice, while another may find it overstimulating and respond better to a soothing smile and soft touch instead. Observing the child’s responses closely and adjusting accordingly is key.
- Stay adaptive to keep the loop going: A continued back-and-forth can be built by beginning with imitation of the child’s sounds, then gradually introducing new vocal models for them to explore. Crucially, we remain adaptive throughout this process. What works for a young infant, having their sounds mirrored back, may feel aversive to an older child and lead to a severed connection. Caregivers learn to continuously recalibrate their responses to stay just ahead of the child’s current level, using the child’s own responses as a guide.
The interactions described here, a mother tracking her toddler’s gaze, mirroring a baby’s vocalizations, a caregiver shifting into the lilting rhythm of parentese, are not extraordinary events. They are the ordinary fabric of early caregiving. What decades of research have revealed is that these unremarkable moments carry remarkable weight in facilitating learning and speech development in a young child’s life. For parents, caregivers, and clinicians, the implications are clear. Early interactions are a powerful lever for early language development. Ultimately, the willingness to follow an infant’s lead combined with careful attention to the timing and quality of contingent adult responses translate into a solid foundation in early childhood.
References
Fiani, T., Izquierdo, S. M., & Jones, E. A. (2021). Effects of mother’s imitation on speech sounds in infants with Down syndrome. Research in Developmental Disabilities, 119, 104118. https://doi.org/10.1016/j.ridd.2021.104118
Gazdag, G., & Warren, S. F. (2000). Effects of adult contingent imitation on development of young children’s vocal imitation. Journal of Early Intervention, 23(1), 24–35. https://doi.org/10.1177/10538151000230010701
Golinkoff, R. M., Can, D. D., Soderstrom, M., & Hirsh-Pasek, K. (2015). (Baby)talk to me: The social context of infant-directed speech and its effects on early language acquisition. Current Directions in Psychological Science, 24(5), 339–344. https://doi.org/10.1177/0963721415595345
Hirsh, J. L., Stockwell, F., & Walker, D. (2014). The effects of contingent caregiver imitation of infant vocalizations: A comparison of multiple caregivers. The Analysis of Verbal Behavior, 30(1), 20–28. https://doi.org/10.1007/s40616-014-0008-9
McDaniel, J. (2025). Effects of a contingent responses intervention on the quantity and quality of vocalizations of preschool children with autism spectrum disorder. Journal of Autism and Developmental Disorders, 55(4), 1187–1202. https://doi.org/10.1007/s10803-024-06279-5
Nadel, J., Soussignan, R., Canet, P., Libert, G., & Gérardin, P. (2005). Two-month-old infants of depressed mothers show mild, delayed and persistent change in emotional state after non-contingent interaction. Infant Behavior and Development, 28(4), 418-425.
Pelaez, M., Virués-Ortega, J., & Gewirtz, J. L. (2011). Contingent and noncontingent reinforcement with maternal vocal imitation and motherese speech: Effects on infant vocalizations. European Journal of Behavior Analysis, 12(1), 277–287. https://doi.org/10.1080/15021149.2011.11434370
Sivaraman, M., & Pelaez, M. (2026). Contingency drives children’s vocal behavior. Journal of Applied Behavior Analysis, 59(3), e70067. https://doi.org/10.1002/jaba.70067
Sun, T., Sivaraman, M., & Sun, Y. (2025). Repeat after you: Contingent vocal imitation increases children’s vocalizations and orienting responses. Journal of Applied Behavior Analysis, 58(4), 852–864. https://doi.org/10.1002/jaba.70036