Researcher in Profile: Inbal Arnon

By Chia-wa Yeh, Head Teacher and Research Coordinator

How do children learn that more than one mouse is “mice,” not “mouses”? Or for that matter, that more than one tooth is “teeth,” not “tooths”? Inbal Arnon, a fourth-year graduate student in the Linguistics Department at Stanford, investigates how children acquire language, including explorations of mysteries such as how they advance from “mouses” to “mice.”

Bing Nursery School is part of the Department of Psychology at Stanford University and serves as a laboratory for research in child development and a site for training undergraduates. This past year, 15 studies took place at Bing. One of these was Arnon’s.

Arnon grew up in Israel, in a politically active family. Her parents are both academics—her father is an economist and her mother a psychologist—and Arnon was very involved in various human rights groups. She received her undergraduate degree from Tel Aviv University, earned a master’s in psycho-linguistics from University of Edinburgh and then came to Stanford to study with linguistics professor Eve Clark, PhD.

Arnon shares Professor Clark’s view on language acquisition—that language, at its core, is a communicative task. “It’s influenced greatly by the input provided to children, by what parents say and by how children interact with one another,” Arnon elaborated. She pointed out that the approach differs from linguist Noam Chomsky’s theory that children have an innate knowledge and ability and that language acquisition occurs by evoking innate knowledge that is minimally shaped by the environment. Arnon and Clark place an emphasis on the importance of input and framing.

Arnon’s first study at Bing examined children’s understanding of relative clauses. For example she looked at their grasp of the sentence “the monkey I am feeding” as opposed to “the monkey the girl is feeding.” Both sentences feature a noun [monkey] that is modified by a clause. Previous studies claimed that children under the age of 6 have difficulty comprehending relative clauses. Arnon and Clark argue that relative clauses with pronouns such as “I” or “you” rather than “the girl” are easier for young children to understand because they resemble what children actually hear and use in everyday life. Previous studies tended to use non-pronoun noun phrases such as “the woman the child is chasing” that are different from the type of relative clauses children hear and produce in real life.

To test her hypothesis in children ages 4 to 5, Arnon designed her props to include photos of her face. She then drew pictures of different situations that enable her to incorporate the pronoun “I” into the sentences. She also integrated colors in the tasks because she found this made the task more fun and engaging for the children. For example, Arnon showed children a drawing of herself and two women, one of whom she was chasing. The women wore hats, each a different color. Arnon asked the children, “What color is the hat of the woman that I’m chasing?” The data demonstrated that children were sensitive to the information provided in relative clauses with pronouns as subjects. Additionally, the color question gave children extra incentive to understand the sentence. Arnon applied the same design to a study of Hebrew-speaking children in Israel and obtained similar results.

Arnon’s second study at Bing investigates how children acquire irregular plural nouns such as “mice” and “teeth.” Previous studies that looked at children’s production of irregular plurals involved asking children to produce single words. Typically researchers show children pictures of an object, i.e., a tooth, and ask children to name it. Later, they show children pictures of teeth and ask children again to name them. Children tend to produce “tooths” rather than “teeth” under such circumstance.

First Arnon and Clark wanted to see how the production of overly generalized plurals (by adding “s” to the end of a word) is influenced by the frequency of the word and by the situation in which children are asked to produce. Children were asked to name pictures, teach a puppet English and correct a puppet’s English. For example, “Do you say ‘mouses’ or ‘mice’?” The results show that children are sensitive both to frequency and to situation. They produce more overly generalized plurals on less familiar items (e.g. “gooses” for “geese”) and when they are in a situation they are less used to (e.g. teaching English).

To demonstrate that input shapes learning, a follow-up study uses commonly used phrases incorporating the target nouns—such as “brush your teeth.” Using the same pictures in the first study, children were invited to teach a puppet some words. Instead of simply asking children to name the object in the picture, Arnon’s research assistant provided a frame such as “brush your…” and invited children to finish the phrase. This study showed that children were able to produce the correct irregular plurals much more frequently when provided with a sequence in which they often hear the word. The sequence helps children to access the single word in its correct form. It is akin to the idea that children often learn words in small chunks, as in this case—“brush your teeth.” While the results are significant with four-and-a-half-year-olds, Arnon and her research assistant are working with younger children to see if the same effect manifests itself among them as well..

What are the implications of Arnon’s studies?

Both of Arnon’s studies demonstrate the importance of frequency and input as part of the mechanism that drives language learning. Arnon asserts that language acquisition is not that different from other cognitive abilities such as visual development. Humans are very good at noticing patterns and extracting irregularities. Language acquisition is no exception. “On a different level, the fact that input is so important should also influence how we view, for example, the importance of very early preschool,” Arnon reflected. Different from the Chomskian view that language learning occurs naturally, Arnon highlights the importance of boosting language exposure for disadvantaged children and providing ample opportunities for meaningful activities with language. For example, a preschool environment that engages children in talking, reading, singing, making up stories and listening to stories can complement what takes place at home. There isn’t a real substitute for human interaction, Arnon noted. “There’s nothing like the negotiation that occurs when you’re talking to another person.”

Arnon enjoyed conducting studies at Bing Nursery School, she said. One especially memorable experience for her took place in the beginning of her studies at Bing, when she was reading a story to the children at snack time as a way to get to know them. In the middle of the story, a child stopped the reading and excitedly announced that she had new shoes and that they have wheels. The child then thought for a second and said, “You know, now I can’t walk anymore.” A teacher asked, Why? The child declared, “I can only wheel around.” As a linguist, Arnon keenly appreciated the exquisite demonstration of how nouns can be used as verbs.

Bing School is a very special place for children and a very wonderful one for research, Arnon said. “A lot of the research couldn’t have been done, of course, without the children but also not without the teachers and the kind of support researchers get.”