Speak to a baby and your voice may transform without conscious planning: the pitch rises, vowels stretch, sentences shorten and the melody becomes exaggerated.
This is infant-directed speech, sometimes called parentese. It is not meaningless baby babble. The words remain grammatically correct, but emotion and rhythm become easier for an infant to follow.
The strongest evidence comes from the ManyBabies project, which tested 2,329 infants across 67 laboratories. Babies reliably preferred listening to infant-directed speech over ordinary adult conversation. The preference appeared across different ages, languages and testing methods, although it was stronger in some populations than others. Large multi-lab study
A separate project analyzed 1,615 recordings from 21 urban, rural and small-scale societies. Across cultures, people systematically altered the acoustic qualities of speech and song when addressing infants. Listeners from around the world could often identify which recordings were intended for babies—even when they did not understand the language. Cross-cultural study
Well supported: infant-directed speech captures babies’ attention, communicates emotion and can make aspects of language easier to process.
Plausible but debated: natural selection specifically shaped parentese to accelerate language learning. It may instead have evolved primarily for emotional bonding and attention, with language learning becoming an additional benefit. Some cultures also direct far less speech toward infants, yet children still acquire language normally.
Before babies understand our words, they may understand something older: the musical signal that says, “This voice is speaking to you.”
