|My presentation is based on an ethnographic study of the work of contemporary dancers during the improvisation project. My research project about dance improvisation was inspired by John Law's idea and practice of studying order through the ordering practices (1994). Through the course of my fieldwork I came to appreciate – apart from all excellent sociological job done by "my" dancers themselves – Harold Garfinkel's (2002) repertoire to study instructed action. On the ground of this project and these theoretical inspirations I would like to speak about two issues: (1) Instructed action is one where agency (and authority) seems to be clearly localized, yet it is not. To control the number and sort of actors involved demands unending flow of hard work. If this is the case, then what can we, in the Study of Religion, learn as a lesson in studying "religious" norms? What agency is connected with them? (2) Some say that the Study of Religions is in crisis (Horyna 2011), and I agree. To take the fact of crisis for granted is, however, different then to concur with an identification of its cause. I ascribe it to the loss of contact with the development of theory end method in neighbouring social sciences. Particularly disastrous I see the loss of contact with decades of developments in qualitative social research in general, and Science and Technology Studies in particular. Instead of that Academic Study of Religion is wasting time and energy in discussions with theology and in seeking redemption in "hard" science. With reference to this diagnosis I would like to demonstrate what the Study of Religion can learn from Actor-Network Theory and Ethnomethodology – at least with reference to questions of agency.