of a dialogue, is carefully regularized, providing precise, workable models of the language used in the community (Ward, 1971).

Detailed ethnographic research studies have shown striking differences in how adults and children interact verbally. Because of the prevalence of the use of questions in classrooms, one particularly important difference is how people treat questions and answers. One classic study, a comparison between the questioning behavior of white middle-class teachers in their own homes and the home question interaction of their working-class African-American pupils, showed dramatic differences (Heath, 1981, 1983). The middle-class mothers began the questioning game almost from birth and well before a child could be expected to answer. For example, a mother questions her 8-week-old infant, “You want your teddy bear?” and responds for the child, “Yes, you want your bear” (see Box 4.6 above). These rituals set the stage for a general reliance on questioning and pseudo-questioning interactions that serve a variety of social functions. Children exposed to these interaction patterns seem compelled to provide an answer and are quite happy to provide information that they know perfectly well an adult already possesses.

Such “known-answer” questions, where the interrogator has the information being requested, occur frequently in classroom dialogues (Mehan, 1979). Teachers routinely call on children to answer questions that serve to display and practice their knowledge, rather than to provide information that the teacher does not know. Similarly, in middle-class homes, known-answer questions predominate. For example, in one 48-hour period, almost half the utterances (48% of 215) addressed to 27-month-old Missy were questions; of these questions, almost half (46%) were known-answer questions (Heath, 1981, 1983).

In general, questions played a less central role in the home social interaction patterns of the African-American children; in particular, there was a notable lack of known-answer rituals (Heath, 1981, 1983). The verbal interactions served a different function, and they were embedded within different communicative and interpersonal contexts. Common questioning forms were analogy, story-starting, and accusatory; these forms rarely occurred in the white homes. For example, the African-American children were commonly asked to engage in the sophisticated use of metaphors by responding to questions that asked for analogical comparisons. The children were more likely to be asked “What’s that like?” or “Who’s he acting like?” rather than “What’s that?” Such questions reflected the African-American adults’ assumptions that preschool children are adept at noting likenesses between things, assumptions that are also revealed in speech forms other than questioning,