By Chia-wa Yeh, Research Coordinator and Head Teacher
Stanford psychology student Andrei Cimpian spends much of his time at Bing Nursery School. He can often be seen reading stories at snack time and pushing children on the swings. In the past four years, more than 500 children at Bing have participated in his studies in the research game rooms.
Cimpian, now in his fourth year of graduate studies, grew up in Constanta, a port city by the Black Sea, in Romania. He received a scholarship to study at Franklin and Marshall College in 1998. Majoring in psychology and philosophy, Cimpian graduated with honors in both. While at Franklin and Marshall, Cimpian conducted studies on spatial memory with adults and tool use with capuchin monkeys. When asked what he likes about studying in the United States, Cimpian said that he appreciated the opportunity to decide on a major while in college. In Romania, students decide on their majors before starting college and then take an entrance examination for the subject they’ve chosen. “I also like the focus on research and the fact that there are so many professors who are genuinely interested in and dedicated to their research and also to mentoring students,” said Cimpian.
After graduation, Cimpian received a Stanford Graduate Fellowship to study at Stanford under the guidance of Psychology Professor Ellen Markman, PhD. Now at Stanford, Cimpian’s research explores the impact of language on young children’s thought. More specifically, he focuses on a particular type of linguistic constructio—the generic sentence. “Generics are sentences that express generalizations about categories—such as, “Kittens are playful”—or individuals—such as, “John likes chocolate,” explained Cimpian. For the past several years, his research has explored a number of questions: How do generic sentences shape preschoolers’ inferences about natural categories? What effects do generic and non-generic praise have on children’s motivation? How do preschoolers determine which sentences are generic?
Cimpian is interested in generics because they are such an important means of conveying knowledge about the world. For example, generic sentences about natural categories—objects that occur naturally in the world, such as cats, dogs, trees, clouds—help children grasp common properties that apply to a category. Similarly, generics about social categories, such as occupational and ethnic groups, might also shape children’s worldview. In contrast, non-generic sentences point out features of individual objects in a category (e.g., “That kitten is playful”; “My nurse is kind”).
The first study in the series looks at kind generics. According to Carlson and Pelletier, kind generics express a property that applies to an entire kind (e.g., “Bananas are sweet”, “Kittens are playful”). Cimpian and Markman hypothesized that children who hear a property phrased generically generalize more than those who hear a property in a non-generic sentence. In this study, 3- to 5-year-old children heard either a generic or non-generic sentence regarding a property of an object. For example, Cimpian showed them a picture of a bird and said either “They are afraid of raccoons” (generic) or “It is afraid of raccoons” (non-generic). The researcher then showed the children three other pictures—a typical bird, an atypical bird, and a giraffe—and asked them if these animals are also afraid of raccoons. The results confirmed the researchers’ hypothesis. Children who heard the generic sentence were more likely to infer that other members of the same category (e.g., the other two birds) are also afraid of raccoons.
One of Cimpian’s studies with particular relevance to parents and teachers looks at individual generics.Individual generics express a property that applies across situations in an individual’s life (e.g., “John likes chocolate”). In contrast, non-generic sentences about an individual report a specific fact or event (e.g., “John had some chocolate last night.”) The goal of this study was to examine the impact of generic and non-generic praise on children’s reactions to challenges. Stanford professor Carol Dweck’s research on helplessness showed that children who receive praise about the whole person (e.g., “You’re a good boy/girl”) are more likely to feel helpless when faced with subsequent mistakes than those who receive praise on process (e.g., “You found a good way to do it.”). This result, Cimpian and Markman argue, might have been obtained because praise about the whole person is generic whereas praise about process is non-generic.
To test this hypothesis, Cimpian and Markman used generic and non-generic praise sentences that were more similar in content than those previously used by Dweck. For example, “You are a good drawer” was the generic praise whereas “You did a good job drawing” was the non-generic praise. Cimpian and Markman were interested in finding out whether children’s motivation would be influenced by this subtle linguistic cue.
In this study, children participated in role-playing with honors student Holly Arce, Stanford senior, using two small dolls. The children had their dolls play the role of a child and the researcher had her doll play the role of the teacher; each held a small piece of pipe cleaner that was used as a pretend pencil.
The researcher narrated stories involving drawing. In the first few stories, the teacher doll gave either generic or non-generic praise of the child doll’s pretend drawing. The researcher simply talked about the drawing without presenting any actual pictures. In a later story, the teacher doll pointed out some missing parts in the pretend drawing. The children then answered questions meant to uncover how they thought about the mistake story and whether they would be able to generate strategies for repairing the mistake.
At the end of the session, the researcher repeated the scenarios and provided children with positive responses and supportive comments. As predicted, the results showed that children who received non-generic praise (“You did a good job drawing”) were more motivated to repair their missteps and felt better about themselves and their skill at drawing. Cimpian hopes this study will encourage parents and teachers to be mindful of their own language when speaking to children, since sometimes even praise can backfire.
How do children distinguish between generic and non-generic sentences? Cimpian and Markman investigated whether preschoolers can determine if an ambiguous sentence is generic (e.g., “They like to play with toy cars,” uttered in the presence of a picture of two cats). In this study, Cimpian showed children a picture of two animals, for example, two cats, and told them either, “Let me tell you something about cats” or “Let me tell you something about these two cats,” followed by, “They like to play with toy cars.” Cimpian then put away the picture and asked the children to tell Mr. Elephant (a stuffed animal) what they just learned. To respond without the picture in view, the children needed to decide if the property—playing with toy cars—can be applied to cats as a category. If children responded with “Cats like to play with toy cars,” then they understood the property to apply to cats as a category. On the other hand, “The cats/these cats/those two cats like to play with toy cars” indicated that the child attributed the property only to the two cats shown in the picture. The findings from this study show that most preschoolers are able to use linguistic information (e.g., “about cats” or “about these two cats”) to determine if an ambiguous sentence is generic or non-generic.
Cimpian’s current research focuses on whether generics influence how children represent natural and social categories, as well as the properties that might apply to these categories. This research will constitute part of his dissertation and will further our understanding of the role of linguistic input in children’s development.