How Do We Learn to Speak?

Through the study of the language development in toddlers and the learning abilities of a robot it is possible to show how we learn languages. This is the view of Francisco Lacerda, professor of phonetics at Stockholm University and EGI customer, who is studying the learning of oral language in children – a field that reaches far beyond the domains of phonetics.

According to previous hypotheses (the nativistic theory) we are born with a pre designed language module in the brain. However, according to Francisco Lacerda and his research group the language development is rather a part of the child’s general development of cognitive and motor functions. By connecting certain sequences of sounds to other visual or sensory inputs the child learns to understand the meaning of a language sound.

The child interacts with the environment, tries to repeat parts of what the adults have said, and when the child hears a word in several different contexts, and receives confirmation that his attempts of pronunciation are correct, the child realizes that and expression is built up of different sound sequences, which are linked to specific meanings.

In order to test this knowledge in specific research models the robot Chico has been developed. The goal is that Chico shall learn to interpret and understand visual and auditory stimuli through interaction with a “caretaker”, Francisco Lacerda says. In this way it is possible to study how repetitions and variations in pronounciation and context, gradually affect Chicos possibilities to continue learning new words and their meaning.

Note: the original version of this story was published by Dagens Nyheter as an advertisement for Stockholm University.