Web edition: November 8, 2007
Spinning a good yarn may seem to have little to do with mathematics, but a new study suggests otherwise. Preschoolers who tell stories that include many different perspectives do better in math two years later than those who stick to one simple perspective. The researchers believe that the study may highlight a deep connection between mathematical ability and narrative skills.
Daniela O'Neill and her colleagues at the University of Waterloo in Ontario gave 3- and 4-year-olds the picture book Frog Goes to Dinner by Mercer Mayer. The wordless book shows a frog hopping around a restaurant, causing mayhem. The researchers asked the kids to tell the story to a puppet who'd never seen the book and measured the sophistication of each child's story. They also gave the children a test of general verbal ability. Two years later, they gave the children a test of academic achievement in a variety of areas.
The scientists found that narrative ability in preschool was a good predictor of a child's later performance in mathematics. Simple measures, such as sentence length and diversity of vocabulary, had little relevance, however. The most important factor had to do with a child's ability to switch perspectives in the stories. For example, one child told the story as if the frog were the only character in the story, while another discussed the internal states of secondary characters, with comments such as, "The waiter was mad when the frog jumped in the soup."
"These aspects of storytelling are tapping an ability to think in a more flexible way," O'Neill says. "The kids are keeping track of relationships, talking about who did what to whom." This is the same kind of mental agility that is necessary for mathematics, she says.
O'Neill became interested in the connection between math and storytelling through the work of Keith Devlin, a mathematician at Stanford University. In his book The Math Gene, Devlin argues that language arose when humans acquired the ability to visualize complex relationships among different objects when the objects themselves are not in view. The ability to do mathematics arises from that same ability to manipulate abstractions.
"A mathematician is someone who approaches mathematics as a soap opera," Devlin says. "The only difference is that the characters in mathematics are mathematical abstractions rather than abstractions from real life people." Therefore, he says, it makes sense that children who have mastered narrative skills would develop a facility for thinking mathematically.
O'Neill and her colleagues found that storytelling ability specifically predicted later mathematical skills rather than overall academic success. The children who told sophisticated stories did not, on average, score better in reading, spelling, or general knowledge. Conversely, a test of preschoolers' general verbal ability, rather than specific narrative skill, did not predict later mathematics scores.
Devlin argues that arithmetic skills have little connection with general mathematical skills. To be good at arithmetic primarily requires being able to follow an algorithm, he says, whereas mathematics in general requires much more sophisticated pattern recognition and logical reasoning. So O'Neill and Shilpi Majumder of St. Francis Xavier University in Nova Scotia have now extended the research in an effort to identify which kinds of mathematical ability are specifically related to narrative skills.
They also wanted to tease out the most important types of narrative skill, since O'Neill's initial research suggested that only certain kinds of narrative skills are relevant to mathematics. The most common way that testers assess children's narrative skills is to give them a set of picture cards that tell a story and ask the children to put the cards in order. O'Neill and Majumder suspected that this ability has little relevance for mathematics.
So they tested children another way as well, by giving them picture cards depicting three simple stories and asking them which two were most similar. In one of the sets of cards, for example, two of the stories were about someone drinking and one was about someone eating. This would require children to abstract the themes from the stories to identify the similarities.
O'Neill and Majumder then assessed the children's specific arithmetic skills and more general mathematical skills, in line with Devlin's distinction. For arithmetic, they asked the children to compute simple sums and differences. For mathematics, they asked them to demonstrate logical reasoning ability by replicating patterns. For example, the researchers presented colored blocks in a pattern such as red, yellow, blue, red, yellow, blue. They then gave the children a collection of toys, say, giraffes, elephants, and rhinoceroses, and asked them to place the toys in a similar pattern. One correct arrangement, for example, would be giraffe, elephant, rhinoceros, giraffe, elephant, rhinoceros.
As the researchers had suspected, the ability to recognize similar stories predicted mathematical skill but not arithmetic skill. The ability to put the story-telling picture cards in the correct order, however, did not predict either math or arithmetic skills.
"Often abstraction seems like a really high-level thing that you do later on in education, but the implication of this is that you can start very early on, in preschool," Majumder says. "If kids are able to pick out themes in stories and they have practice in it, that's going to help them with math."
If you would like to comment on this article, please see the blog version.