This project analyzed the regular patterns of social interaction in science classrooms and the verbal and non-verbal strategies by which the science content of lessons is communicated. Based on observation and recording of 60 lessons by 20 teachers in 3 schools and a university, the project identified: (1) the principal science classroom situation types and the rules of behavior by teachers and students in each of them; (2) principal strategies by which the system of scientific meanings being taught is expressed in the classroom dialogue; (3) the rules observed by teacher and students concerning what is a "proper" way to talk science; and (4) the relations between teachers' observing or breaking those rules and the likelihood of students showing attentiveness to the lesson. Results indicate that: (1) students are three to four times as likely to be especially attentive when rules are broken by the teacher as when they are being followed; (2) most of the time the scientific meanings being taught are expressed implicitly, not explicitly in the classroom dialogue; and (3) social beliefs about science and learning artificially limit classroom dialogue in ways which make it more difficult for most students to learn science. (Author/JN)