Voice 1
Thank you for joining us for Spotlight. I’m Liz Waid.
Voice 2
And I’m Joshua Leo. Spotlight uses a special English method of broadcasting. It is easier for people to understand, no matter where in the world they live.
Voice 1
When Owen was born, his mother, Janice, took many pictures of him. Her camera was always busy! She took pictures of Owen wearing funny hats. She took pictures of Owen taking a bath. She took pictures of Owen at the play area, in the car, with other babies, and with the family’s cats. Sometimes Janice also recorded Owen with a moving picture, or video, camera. She watched her video recordings over and over. She wanted to remember how Owen looked and acted as a baby. She wanted to keep these memories for a very long time. When Owen grows older, he can also watch these videos of himself as a baby.
Voice 2
Professor Deb Roy also takes pictures and video of his baby, John.* But Professor Roy has a different reason. Professor Roy is trying to discover how human babies learn language. He wants to discover how a baby’s environment shapes human development and learning.
Voice 1
Today’s Spotlight is on the Human Speechome Project.
Voice 2
The Roy family lives in Massachusetts, in the United States. Professor Roy works with computers. He tries to build computers that can learn to communicate in human-like ways. This means he first observes how humans communicate. Then he tries to build a computer model of their behaviour. Building computer models helps him and his research team to understand how people learn to communicate.
Voice 1
In August 2005, Professor Roy and his wife had their first baby - a boy. Professor Roy thought that this happy event could also be an interesting chance to learn. He could observe a human baby learning language from the very beginning. So, when his son John was born, Professor Roy began ‘The Human Speechome Project’.
Voice 2
Speechome is not a real word. But it combines the words ‘speech’ and ‘home.’ That is because Professor Roy will observe his son learning speech in his home. He expects that John will develop language skills normally - like any other child. And Professor Roy will use his observations to make a model of the way people learn language.
Voice 1
No one really knows how human babies learn language. Experts do know that communication is an important part of human development. And they know that babies learn and understand their first language quickly. Children listen to the speech of their parents, brothers and sisters, and other family and community members. They try to make sounds of their own. They even try to form the words they hear. But many experts agree that listening is not the only way that children learn a language. So, how do babies learn to speak? That is the question that the Human Speechome Project is trying to answer.
Voice 2
The Human Speechome Project will try to record baby John as he learns new words. How will John’s learning be recorded? Well, Professor Roy has set up microphones and video recording devices all around his house. He put eleven [11] cameras in different rooms. These cameras will record videos of people who enter or leave the room. He also put up fourteen [14] microphones in rooms around his house. These microphones will record all of the sounds in his house.
Voice 1
The cameras and microphones record data information for twelve [12] to fourteen [14] hours every day. They record three hundred and sixty five [365] days a year. The recorded information goes to a main storage computer in Professor Roy’s house. After a few days, Professor Roy’s research team takes the stored information to their laboratory. There, several computers and human researchers all study it.
Voice 2
There is a lot of information. So, computers do much of the work. Some computers look for simple repeated processes or events that happen during the day. For example, the computers can recognize simple actions like cleaning the dishes after a meal. Researchers can see if these everyday, normal events affect language learning. And the computers keep the information organized. Researchers can then spend more time studying important events like parent and child play time or feedings.
Voice 1
Some computers simply show how and where people move through Professor Roy’s house. But other computers are able to find repeated speech. Repeated speech patterns could be very important in the process of learning language.
Voice 2
As baby John grows older he will begin to form words. What will John’s first words be? Who used those words when he was present? Where were they when the person used those words? What was John doing when he learned the words? Researchers will be able to look through their computer records for the answers to these questions.
Voice 1
With all this information, Professor Roy wants to build another, final, computer as a model. He will enter the collected information into the computer. Then the computer will model John’s experiences as a baby. Hopefully, the computer will ‘learn’ by ‘hearing’ and ‘seeing’ the same sounds and images that John saw as a baby. The computer will learn words and sentence structure just like John did. And hopefully, the computer will be able to show Professor Roy and his research team the process of how John, or any baby, learns language.
Voice 2
Professor Roy plans to record his son until he is three [3] years old. A healthy human baby can usually speak his first language by this age.
Voice 1
Language is an important part of every person’s life. Without language, people could not communicate with each other. So, Professor Roy’s work with language is very valuable. Success in the Human Speechome Project could lead to better treatment for people who have trouble learning language. Or it could lead to computer programs that can learn to speak.
Voice 2
Observing a baby learning language offers wonderful opportunities. But it also presents questions. Is it acceptable for Professor Roy to study his son’s development like this?
Voice 1
And if the project succeeds, what will Professor Roy and his research team learn from the information they collect? Will they begin to understand more about how people learn language? Will they be able to teach computers to learn a new language? For now, no one really knows the answers to these questions. But in the future, people may begin to understand more about how people learn language.
|