To a “hanging” little child, adults all over the world speak with greater emphasis, with higher tones, and also with much more variation in pitch than they would to an adult. And when they start singing especially for that child, people all over the world sing much flatter in tone, lower in tone and also with much less rhythm than in songs for adults.
This is shown by a large-scale comparative ‘baby talk’ study of 21 different cultures from all continents (from hunter-gatherer groups to residents of large American cities). It’s this week published in Nature Human Behavior. In the study, an average of 15 parents per culture were asked how they would sing and talk if their baby were ‘fuzzy‘ (‘hanging’, ‘difficult’, ‘weepy’) was also recorded as they sang a random song to adults and how they told something to the adult experimenter – all in their own language. In 90 percent of the admissions, the young child, up to the age of two, was actually present.
Bambinese
The research by a large team of linguists and anthropologists was initiated by Courtney Hilton and Samuel Mehr (both Harvard University). Mehr is also known for a large study of the common characteristics of lullabies and dance songs worldwide. This new study was started because it is often assumed that all over the world the language adults use to talk to small children is about the same (baby talk, parent language, ‘mammian’, ‘motherese‘, ‘infant-directed-speech‘, or as the French say: ‘mamanaise‘, Italian: ‘bambinese‘ or ‘linguaggio bambinesco‘). But that has never been properly researched, according to Mehr and Hilton. At most, disparate recordings from different cultures were compared, never looking at how normal speech or songs for adults sounded in the same subjects. With acoustic analysis of 1,615 recordings of 410 test subjects from 21 different cultures, this has now all been done.
Mehr and Hilton see the common features of baby speech and the baby songs as a clear indication that these ‘child communication tools’ are based on general human patterns, which even correspond to patterns in the animal world. The higher pitch in children’s speech, for example, occurs in a lot of animal communication as a sign of friendliness and openness – perfect for putting human children at ease as well. It contrasts with the usually lower tones of alarm cries or threatening calls.
A selection of the speech and vocal recordings was sent through an international (English-speaking) citizen science-website also presented to a large internet panel of nearly seventy thousand participants from 187 different countries. In a kind of quiz, the participants had to indicate whether the sound fragment they heard was intended for a child or not. This was correctly guessed significantly more often than would have been expected by chance. There were also few differences worldwide, although participants guessed songs and speech in a language related to their own somewhat more often, probably because they could more or less understand those words.
No nursery rhymes
According to the researchers, the fact that child-directed speech and child-directed singing appear to have opposite properties (a lot of emphasis versus flatness), may be due to the different purpose of the action. For with speech people would ‘fussy infant’ wanting to distract from his or her discomfort: the greater variation in tone and rhythm of speech serves to attract attention. And singing is used in the case of a wailing baby to lull the baby to sleep: hence the flat tone and the slight variation. The differences may also have been amplified because the subjects were instructed to talk to, or sing to, a ‘clingy baby’. For happy babies, people might start singing differently.
The researchers also note that children’s singing and speech were not equally popular in all cultures studied. In the Quechuan/Aymaran culture in Mexico it turned out that there were no children’s songs in their own language, even though the people there knew Spanish-language children’s songs. And with Bayaka pygmies in Congo, songs for small children are particularly popular, but again very little is spoken to children under two.
It is precisely these cultural differences in the use of ‘baby talk’ that another researcher, Alejandrina Cristia (Université Paris Sciences et Lettres), published a few months ago. a great review article in Developmental Science. Her conclusion, based on 29 previous studies from around the world, is that children under the age of two in an urban environment are much more likely to be spoken to with baby speech than comparable children in agricultural or hunter-gatherer societies. Toddlers in Chicago, for example, are more than ten times more likely to be spoken to by adults than their peers in a Mayan village in Yucatan. And that speeds up language acquisition: in the big city, the small children have a much larger vocabulary than those in the Mayan village.
Biggest concern
Cristia explains the differences in children’s speech between urban and rural areas, among other things on the basis of the different goals that parents can have for their upbringing. In some cultures, it is more important for the children to become an active member of the community than for them to develop individually. It is also important there to learn the correct forms of address in different social situations. Parents therefore do not find it necessary to talk one-on-one with the small children, but they do ensure that the children can hear the conversations of others. And more broadly, the biggest concern in pre-industrial farming societies is the survival of children, Cristia said. Language and rapid cognitive growth are then not a priority.
In a more modern and urban environment there is less child mortality and the children are much more cognitively stimulated so that they do better in school and will get better jobs. And because parents in the city have fewer children, there is also more time available for each individual child.
Also read about language acquisition: Language is coming (2004)