Songs as regulation tools: what music research says about young children's emotions

Long before a child can understand a single word of reassurance, they can be reassured by a song. Every culture we know of sings to infants, and the songs converge on the same features — slow, repetitive, higher-pitched, rhythmically regular. That universality made researchers suspicious in the best way: when parents everywhere independently invent the same tool, it is usually because the tool works. Over the past two decades, developmental psychologists — most prominently Sandra Trehub's group at the University of Toronto — have put that hunch through the lab. The results are worth knowing, because they turn "sing to your child" from folk wisdom into something closer to a usable technique.

The most universal parenting tool there is

The research distinguishes two things that often get blurred: music's ability to capture attention, and its ability to regulate arousal — the physiological gear a child's body is in. A toddler mid-meltdown doesn't have an attention problem; they have an arousal problem. The interesting question is whether songs genuinely shift that dial, or just distract. The evidence says: both, and the dial-shifting is measurable.

Singing vs speech: the nine-minute finding

In a 2016 study in Infancy, Corbeil, Trehub and Peretz played recordings to infants aged 7–10 months until the babies showed the first facial signs of distress. The recordings were either adult-directed speech, infant-directed "baby talk", or singing — in Turkish, a language none of the infants knew, so the content couldn't be doing the work. Babies listening to speech lasted around four minutes before the "cry face" appeared. Babies listening to song lasted roughly nine.¹

The authors' interpretation is important: they credit the temporal patterning of song — its steady beat, metrical structure and tempo — with inhibiting distress, possibly by giving the infant's nervous system something predictable to entrain to. Predictability regulates. It is the same principle that makes routines calming for toddlers, which we explore in our post on predictability and child anxiety.

Lullabies and playsongs: two dials, not one

Parents don't just sing one way. Laura Cirelli and colleagues had mothers sing "Twinkle, Twinkle, Little Star" to their infants repeatedly, alternating between a soothing rendition and a playful one, while measuring skin conductance — a physiological index of arousal — in both mother and baby. During soothing renditions, arousal dropped below baseline in both members of the pair as the song went on. During playful renditions, arousal held steady and infants paid more attention to mum.²

In other words, the same song functions as two different tools depending on delivery: a lullaby-style rendition is a down-regulator, a playsong an engager. And strikingly, the mother's own arousal dropped alongside her baby's — singing regulates the singer too. This is co-regulation in its most literal, measurable form, a mechanism we cover more broadly in our post on co-regulation before self-regulation.

Why the familiar song wins

If any song helps, a known song helps more. In a 2020 study in Developmental Psychology, Cirelli and Trehub compared how well different audio recordings calmed distressed 14-month-olds: familiar songs (ones sung regularly at home), unfamiliar songs, and speech. Familiar songs reduced distress most effectively — and were the only condition that got toddlers smiling again.³

This is the finding with the most practical leverage. A song that is theirs — sung at the same points in the day, attached to the same routines — becomes a conditioned safety signal the child can carry into new situations. It's also the research rationale for why each Little Narratives 5-day preparation plan pairs its daily story with a personalised song: by the time the big day arrives, the song is familiar, and familiarity is precisely what the evidence says does the regulating.

Rhythm, movement and co-regulation

Why does rhythm in particular have this power? Part of the answer is that young children appear to be built for it. Zentner and Eerola tested 120 infants aged 5–24 months and found they moved significantly more rhythmically to music and steady drumbeats than to speech — and the better their movement coordinated with the music, the more they smiled.⁴ Rhythmic engagement and positive affect travel together from the start.

A steady beat also gives adult and child a shared timing structure — you rock, pat or sway with the song, and the child's body is carried along. That shared timing is one reason songs are so good at marking transitions: the tidy-up song, the goodbye song at daycare, the brushing song. The song externalises the structure of the moment, so the child isn't relying on words alone to know what happens next — particularly useful for the transition battles we describe in our post on why toddler transitions are so hard.

The clinical evidence: music in paediatric settings

The strongest stress-test of music as a regulation tool is the hospital. A systematic review by Klassen and colleagues pooled 19 randomised controlled trials — over 1,500 children — of music used during medical and dental procedures, and found significant reductions in both pain and anxiety (standardised mean difference −0.35 overall).⁵ The effect held whether the music was delivered by a trained music therapist or simply played to the child.

If music can take a measurable edge off a needle or a dental drill, the everyday applications — drop-off, bedtime, the waiting room — are asking much less of it. For the specific challenge of needles and doctor visits, we've collected the broader preparation evidence in our guide to preparing children for doctor visits.

The honest limits

• Effects are real but modest, and measured in minutes. Singing delays distress; it does not abolish it. A song is a tool for taking the edge off and buying regulation time — not a replacement for responding to whatever the child actually needs.
• The clinical trials are methodologically mixed. The Klassen review itself flagged that most included trials were of poor methodological quality, so the pooled effect should be read as promising rather than settled.⁵
• Most lab studies involve infants, not preschoolers. The distress-delay and arousal findings come largely from babies under two. Older children's responses to music are better documented in clinical settings than in everyday tantrums, where the evidence is thinner.
• Live and responsive beats recorded and generic — probably. The co-regulation findings involve a caregiver singing, not a speaker playing. Recorded music helps (the clinical trials mostly used recordings), but a parent's own voice, adjusted in the moment, is likely the stronger version of the tool.

Using songs as regulation tools at home

None of this requires being able to sing well. Infants show no evidence of musical snobbery; they show strong evidence of preferring the voices they love, singing songs they know. The research message is almost embarrassingly simple: songs are one of the oldest regulation technologies humans have, they are free, and — used with a bit of intent — they work alongside the story-based preparation we cover in our post on narrative rehearsal as well as anything in the parenting toolkit.