← archive index · live page · raw markdown
What Babies Listen To
Metadata
- title
- What Babies Listen To
- description
- A 2026 study of babies in the first two days of life asked whether they can already predict a musical phrase. The result is about expectation and surprise, not a universal language.
- status
- complete
- date
- 2026-09-22
- kind
- solo
- guestSlugs
- —
- listenUrl
- https://podcasters.spotify.com/pod/show/lovemusicmore/episodes/What-Babies-Listen-To-e3p15fl
- appleUrl
- https://podcasts.apple.com/us/podcast/what-babies-listen-to/id1567355195?i=1000791063126&uo=4
- topicsDiscussed
- Music theory
- Philosophy
- Rhythm
- Melody
- Musical expectation
- Newborns
- Consonance
- Surprise
- hostNote
- A 2026 study looked at babies in the first two days of life and asked whether they can already predict music. I use that to ask what the prediction says about music itself. The comparison is a real phrase against a scrambled one. The newborns showed stronger neural responses when an expected next step was cut off. They were forming expectations about rhythm and melody from the phrase in front of them, from its own vocabulary, not from a lifetime of songs. I had thought that kind of statistical listening showed up later, in some deeper phase of brain development. It is there on day zero. What I will not let the study become is a proof that music is a universal language. Consonance, an out-of-place pitch, the pleasantness people treat as physics: I do not get a clean newborn result for that. Timing and surprise light up. A brain can keep a musical context and react when the context breaks. That is the finding. It is already a lot. The week before, on [Fishbowls of Influence](https://lukefwalton.com/love-music-more/episodes/fishbowls-of-influence/), the question was whether a chord is good or bad outside the culture that trained the ear. This episode is the other end of that. Before the fishbowl, is there already a listener? The week after, [Monkey Studies](https://lukefwalton.com/love-music-more/episodes/monkey-studies/) takes the same question to rhythm, and to rhesus macaques.
- selectedMoments
- label
- Day zero
- note
- I start from a 2026 study of babies in the first two days, and from the question of what a prediction that early would mean.
- label
- Phrase against scramble
- note
- I walk through the comparison: a real musical phrase, and a scrambled control.
- label
- The cut-off next step
- note
- I stay on the result that matters: stronger neural responses when an expected continuation is cut off.
- label
- The phrase's own vocabulary
- note
- I argue they are building expectations inside the phrase, about rhythm and melody, not from a life of songs.
- label
- Not a universal language
- note
- I refuse the leap from that result to music as a universal language, and I do not claim a newborn consonance effect.
- label
- Before the fishbowl
- note
- I set this next to the week before, on whether pleasant chords exist outside the culture that trained the ear.
- faq
- question
- Can newborns predict music?
- answer
- On the 2026 study I talk through here, babies in the first two days showed stronger neural responses when an expected next step in a phrase was cut off, compared with a scrambled version. They were tracking rhythm and melody inside that phrase. I do not treat that as proof they hear an out-of-key note the way a trained listener does.
- question
- Does that make music a universal language?
- answer
- No. That sentence is the one people reach for, and the study does not earn it. What it earns is narrower: from day zero, a brain can keep a musical context and react when the context breaks. Consonance, and the idea of a naturally pleasant chord, belong to the week before this, on Fishbowls of Influence.
- question
- How does this connect to the monkey episode?
- answer
- Monkey Studies is the week after. I take this newborn result and ask the factory-default question again, this time about rhythm and about rhesus macaques. The babies episode is the setup. The monkey episode is the next turn.
- transcriptPublished
- true
- draft
- false
Content
No body content; see metadata above.