Even a superficial review of the literature on collegiate mathematics education reform reveals a common thread: the imperative nature of the college mathematics student's active participation in the learning of mathematics. Active participation in the classroom may take the form of working cooperatively with classmates in small groups, spending time in the computer lab using class-related software, presenting problems or concepts on the board to classmates, or other activities germane to the subject bestdes passively listening to a lecture presented by the instructor. There is a wide range of lectures styles, but the stle to which I refer is one characterized by little expectation of student engagement besides perhaps following the logic of the lesson. Because active learning in the mathematics classroom is advocated by several professional groups ([1], [4], [5], [6], [8]), it is certainly reasonable for mathematics instructors to reflect on their rationale for maintaining a lecture style of teaching, even part of the time. During the 1995-1996 academic year, that is exactly what I did. I was teaching a section of reformed calculus, with a class size of forty. I had planned on extensive use of cooperative groupwork, and had accordingly divided the class into ten groups of four people each. I attempted to make the groups heterogeneous (in gender and ability) by collecting quantitative measures of the students' previous mathematical performance. Several times during the first two sveeks of class, I set aside time for group work. However, due to the physical layout of the classroom (five rows of auditorium seating with twelve seats per row and an aisle up the middle) my students were having a difficult time actually communicating in their groups. Classtime initially devoted to guided discovery in the groups degenerated into work time for forty individuals, certainly not the vision I had foreseen when planning earlier that summer. To restructure this time, I tried to view my entire class as a large cooperative group, with me as the group leader. Instead of breaking the class into four-person groups as before, I would present examples of concepts we were currently studying, and pose open-ended questions to the class. For example, while developing a geometrical approach to Newton's method for finding the roots of an equation, I asked my students which solution they thought the graphical process would yield when the initial guess for the x-value was directly between two of the roots. Was there a pattern that allowed us to predict the general result? Another time wheh discussing inverse functions, we did an impromptu in-class exploration of which Itnear functions, and later, which general functions, are their own inverses. Such discussions involved much conjecturing, arglling, and scratchpaper seatwork, individually and in pairs, among the students, and were opportunities for algebraic, geometric, and numerical exploration. l Because of my class's general enthusiasm and willingness to participate, If felt we had developed a classroom atmosphere free of criticism, and open to conjecturing. The students often presented problems for each other, furthering their own mathematical communication. Although I was somewhat disappointed not to be able to watch my students develop cooperative skills, I felt that we had made the