Even though the sounds of speech are very much dampened in the uterine environment, studies based on fluctuations in fetal heart rates have shown that fetuses can distinguish a new syllable when it arises in the middle of a repeated familiar word. The same methods have also shown that a fetus can distinguish a male voice from a female one.

Its mother’s voice is the one that a fetus hears best, because this voice reaches its ears from two different pathways: through the air, and through the mother’s body. It is therefore unsurprising that observed variations in the frequency of sucking behaviour in newborns have shown that even only 12 hours after birth, a baby prefers its mother’s voice to those of other women. But no studies have yet shown that a newborn prefers its father’s voice to those of other men.

In another interesting set of studies, babies’ mothers and women who were not their mothers read aloud to them when they were only a few days old. The results showed that a baby prefers hearing its mother read a passage that she had also read aloud during the six weeks before its birth rather than a passage that she had never read aloud before. More surprisingly, the baby still prefers the more familiar text even if the woman reading is not its mother! This shows that the baby has also retained a memory of linguistic elements that it heard before its birth, and that it is elements such as the melody and the rhythm of language — what we call “prosody” — that have attracted its attention the most.

As the baby grows older, it will begin processing language in a less global way and focus more and more on the various sounds that it encounters. Before it is one year old, it will gradually lose certain abilities, such as the ability to recognize certain sounds and accents that it does not experience in its mother tongue.