Increasing public knowledge of science is a widely recognized goal, but what that knowledge might consist of is rarely unpacked. Existing measures of science literacy tend to focus on textbook knowledge of science. Yet constructing a meaningful list of facts, even facts in application, is not only difficult but less than satisfying as an indicator of what people actually know--or need to know--as citizens. Revisiting this concept from a more sociological perspective yields a rather different concept that is here termed "critical science literacy." The implications of this reconsideration for thinking about citizen perspectives, for the work of science journalists and other professional interpreters, and for the design and evaluation of new approaches to science communication are explored.