Tseltal children learn language with minimal child-directed speech

Picture: Lead transcription assistant Rebeca Guzmán López plays with her niece at home 

Many Western caregivers consider it important to talk to young children directly and frequently. When they do so, they speak in a way that they believe is comforting, interesting, or easy for the child to understand. However, talking directly to young children is not the norm around the world. “In many places children have been reported to learn their home languages without much child-directed speech at all”, says Dr Marisa Casillas, lead author of the study. So for these communities, what role does child-directed speech really play in language development?

In collaboration with linguistic anthropologists Dr Penelope Brown and Professor Stephen Levinson, Casillas investigated language development in a rural Tseltal Mayan community in southern Mexico. Brown has conducted research in this community since 1971 and has long noted that Tseltal caregivers raise their children to pay attention to what is going on around them. She has also observed that most speech to children comes in brief bursts, is focused on actions (rather than objects), and tends to be initiated by the children themselves. “We wanted to see how much child-directed speech children typically hear while learning Tseltal. If the answer is ‘not much’, it opens up the question of how else they are managing to learn language”, explains Casillas.

To measure their language environments, ten Tseltal children aged three and younger were fitted with wearable audio recorders and photo cameras, so that the researchers could collect naturalistic recordings of at-home speech during most of the waking day (approximately ten hours). These recordings were transcribed with the help of a native speaker. For each child, the investigators transcribed one hour's worth of short clips sampled from the day-long recording, and made notes about who was talking to whom. In addition, the children’s own utterances were transcribed and categorised according to their linguistic maturity (for example, cooing vs. babbling vs. recognisable words).

The children heard about three and a half minutes of directed speech per hour, which is about one third of what North American children are estimated to hear. In contrast, there was an enormous amount of speech directed to other people that the children could have listened to (about twenty-one minutes per hour). Importantly, this relative lack of child-directed speech did not lead to a language delay; the Tseltal children started babbling and speaking at the same age as Western children do. For example, children were producing words by age one and short sentences by age two. How do they do it? It is possible that children learn from the speech around them, even though it is not directed at them. But the authors also discuss a more intriguing option: the limited quantity of child-directed speech may simply have been sufficient. Children typically heard more and ‘richer’ directed speech in the morning, when it accompanied routine activities such as mealtime. This is comparable to the peaks in Western parents’ directed speech, which are also coupled to activities (such as storytime). As Casillas explains: “The results do not mean that child-directed speech doesn't matter, only that there is more to the picture—timing, context, and other access points to language may all be crucial for language learning, even in Western contexts”.

“By better understanding the different routes to language taken by children across human societies, we can shed light not only on the multiple cognitive tools available for learning language but also how those tools are flexibly used to adapt to the linguistic niche in which each child is raised”, concludes Casillas.

Publication

Casillas, M., Brown, P. & Levinson, S. Early language experience in a Tseltal Mayan village. Child Development

Link to paper