The first four years of a person’s life are crucial for language development, but the speed at which children learn to speak varies greatly. Most babies say single words by 18 months and progress to phrases and sentences around two to three months, but others begin babbling much earlier.

For years, Harvard University psychologist Elika Bergelson has wondered what individual differences lead to such a variety of abilities. His latest observational research on this topic has yielded surprising results.

After studying 1,001 children under the age of four, she and an international team of psychologists were surprised to find no correlation between the number of sounds a child makes per day and their gender, socioeconomic status or level of multilingualism.

Instead, the experimenters found that the most talkative children in the global study were those who, combined with age and clinical factors such as prematurity or dyslexia, tended to hear more adult speech.

“Slowing language development is often associated with parents of lower socioeconomic status providing less input to their children (from a middle-class, Western-oriented perspective), leading to calls for behavioral interventions to increase this,” write Bergelson and colleagues. friends coworkers. .

“Proponents of such interventions may emphasize our connection between adult speech and child language; critics may instead emphasize our finding that socioeconomic status was not significant in either our main analysis or any other reanalysis we tried.”

More research is needed to find out which interpretation is correct.

The results of the study cannot reveal why children’s speech and adult speech are so closely intertwined, but there are several possible explanations. Children who talk more may cause adults to talk more, or a more conversational environment may encourage more conversation among children.

The second scenario suggests that getting adults to talk more to their children might be beneficial. Bergelson’s research covers 12 countries and 43 languages, as well as regional and urban areas. Voices for the study were collected using portable voice recorders placed in children with various developmental levels and abilities, ranging in age from two months to four years old.

Using machine learning to examine more than 40,000 hours of recordings, Bergelson and his colleagues examined a number of factors that could influence how often and how early a child babbles or says syllables, words or sentences.

Ultimately, the team found no definitive correlation between the number of sounds a child makes during the day and his or her gender or socioeconomic status (determined by the mother’s education level). Instead, the results show that children under the age of four produce an additional 66 sounds per hour with each year of development.

This was to be expected, since age is closely related to cognitive development, but there was another factor that showed a surprisingly strong effect: Children who heard adults talk more talked more themselves. On average, for every 100 adult sounds a child heard in an hour, that child produced 27 more sounds. This “adult speech” effect increased by 16 vocalizations each year of the child’s development.

To put the significance of this effect into perspective, children in the study with abnormal language development, such as dyslexia, produced 20 fewer sounds per hour than their peers. Every year, this difference widened by approximately 8 voices per hour. The current study only counted sounds made throughout the day; that is, it did not examine the complexity of the language children use. Consequently, socioeconomic or gender factors may still influence some elements of language development that were not taken into account in this study.

For example, children growing up in families with higher socioeconomic status may have parents who read to them more, possibly improving their vocabulary or grammar development. Bergelson acknowledges that his latest work uses a “coarse-grained” approach that can miss some fine details.

“This is an algorithmic assessment of how much speech a child is hearing or saying,” Christy told DeSmith. Harvard Gazette. “But I think this is a complementary approach to a job that is otherwise very, very time-consuming and sample-limiting.”