Not being listened to, or having their advice mistrusted, is a daily problem for teachers and parents, as well as for doctors, scientists, technologists and other experts. It is also a serious dilemma in applied social psychology (Eiser, 1986) as well as in education where educators talk persuasively with strong evidence about not taking drugs, not being racist, changing dietary life-style, and not smoking, with plenty of supporting evidence, but often to little effect. One might say that they are not listening, so as not to have to act. There are several reasons why science education should be interested in the question of risk. One is that that the risks associated with new technologies--nuclear power, cloning, and recent diseases on the farms--are all a part of what a citizen may need to know (Irwin, 1995). STS (Science Technology and Society) was started in the early 1980s in Holland, Canada, and Britain because it had come to seem obvious that students should be taught about topics that were relevant and important to everyone. STS is now clearly back in the curriculum: in England it is a part of "science for citizenship." There is another, almost unconnected reason. The concept of risk links together the cognitive and the affective domains: it advocates not only thinking but also empathising. This indicates that the topic lies in a boundary region between science and the humanities, and so could usefully become an important part of teaching, in an age when science seems too dry and mathematical to catch the imaginations of most students. In this short article the author will identify four rather different ways in which educators might conceptualise giving and accepting advice about risk. They arise out of complementary perspectives, all concerned with the social world in which advice and its reception both exist. (Contains 1 note.)