Research on learning within CA has in recent years contributed to new insights about the social organization of learning in everyday interactions, inside and outside classrooms (Melander, 2009; Lee, 2010; Sahlström, 2011; Tanner, 2014). This research reveals how learning as social actions gets actively accomplished as participants successively change their epistemic stance to a simultaneously construed learning content. In doing this, participants coordinate their actions within participation frameworks, using both verbal and non-verbal resources such as language, bodily stance and material structures (Goodwin 2007; 2013). In this presentation I aim to further explore the use of material texts as resources in such participation frame-works. Even though the use of texts in people’s everyday lives has been well described within other fields of research, CA provides analytical tools that could contribute to deeper understandings of the role of texts in human interaction. Here I explore how texts come to play part in the organization of learning in teacher-student interactions in classrooms. Analyzed data comes from a larger classroom video-ethnography on learning and literacy practices during the middle years, i.e. students aged 10 to 12 years, in two different classrooms as they work with individual assignments in subjects Swedish and Geography. I particularly focus on so-called desk-interactions, i.e. situations when students work individually at their desks while the teacher moves around to support and supervise them. The results show how students in these situations are expected primarily to solve problems as independently as possible, and to regulate their work in relation to work-instructions and other texts. Textbooks, pictures, the blackboard or the students’ own texts are mainly used to motivate a request for help, or to remind each other of where to find information to solve an assignment. Thus, text references in desk-interaction get an indexical function, as they are used to coordinate the variety of texts middleyear students encounter in different situations. The possibilities for shared reading and discussions on text comprehension are on the other hand shown to be limited in desk-interaction. A conclusion is that the use of texts comes to work more as resources to organize learning and instruction in deskinteraction, than being construed as a learning content per se.