An ongoing investigation by Stanford University, The Max Planck Institute for Evolutionary Anthropology and the Massachusetts Institute of Technology (MIT) is trying to understand how children integrate information from different sources to learn new words.
Previous theories to explain how children learn new words
Psychologists have tried to come up with theories to explain how children understand the meaning of new words. From their studies, they note that children expect speakers to use new words when talking about unfamiliar objects. Children can also look from their past interactions with a speaker to understand what they are talking about. They can also use the present context of the conversation to know the meaning of a word.
Researchers often study these information sources in isolation from each other. However, this view does not represent the reality of the complexity of social interaction where many sources of information can be present simultaneously. This study aims to understand how children use all the information available to them at once to learn new words and their meanings.
Using a computational cognitive model
According to Michael Henry Tessler, a co-author of the study, the researchers carried out experiments to measure the children’s sensitivity to different sources of information. Afterwards, they developed a computational cognitive model which details the way all the information is combined. The model then predicted what would happen in hypothetical scenarios where all those sources are available to the children.
When they tested these hypothetical situations in the real world, they found that the cognitive model had predicted the children’s behaviour accurately.
Researchers also came to the conclusion that children ignored some information sources. They also believed that the way children integrated these information sources changed as the children got older.
To test the children’s sensitivity to information sources, researchers collected data from 148 children aged 2-5 years. They used the data they obtained from these children to create the cognitive model. For the real-world experiments, 220 children of age 2-5 years were used.
The researchers were quick to note that it is still early to come to any definite conclusions. More tests with more participants are needed.
