Systematic teaching of encyclopaedic articles is not common. Even if they do use them or refer to them extensively, or invite pupils to refer to them for oral presentations, teachers are not familiar with the genre, and even less so with the participatory digital encyclopaedic article (Renaud, 2020; Rouet, 2012). Yet today's knowledge texts are increasingly "screen writings" (Souchier, Candel, Gomez-Mejia, 2019).  There is no such thing as screen writing without "architexts". These represent all the digital tools (software, interfaces, applications, etc.) that make it possible and format it, but they also include all the paratext devices that anticipate the reading of these writings and mark out their meaning. Exploiting these devices to create meaning is not self-evident, and requires specific support. Are these new contexts and their cognitive modes of use far removed from the teacher's traditional tools? What is the best way to support teachers in transforming their tools? How should this support be progressed

Our research is in line with Vygotsky's questioning of teaching tools. If pupils' performance is to be improved, teaching tools will have to be better adapted to the obstacles posed by all learning. What are the most appropriate tools for systematic teaching of reading comprehension of diverse and varied texts, including 'screen texts'?

We compare the work of 6 teachers from 8H and 9H on the participative digital encyclopaedia article. We observe the variations and constants in the representations of primary (8H=N3) and secondary (9H=N3) teachers, their planning and their implementation. Our comparison is conducted in three stages. In the first stage, we observe the interviews between the teachers and the researchers, the purpose of which is to plan a teaching sequence. We carried out a thematic categorical analysis on two subjects: the supposed difficulties identified by the teachers for their respective pupils, on the one hand, and the tasks devised to overcome the obstacle, on the other. Secondly, we compared the sequences carried out. Using the model for analysing sequences reduced to a synopsis, developed by GRAFE (Schneuwly and Dolz, 2009; Ronveaux and Schneuwly, 2018), we describe the movements of the object being taught. The synopsis of the sequence gives us information about what is actually taught through the mediation of a sequence of tasks. From the order of these tasks, which are supposed to get the pupils active, we identify the elements taught in the participatory digital encyclopaedic text. Thirdly, we compare the pupils' productions in the last task. We observe how the pupils interpret the instructions and which elements of the text they have used to solve the problem posed by the task.

These three stages of comparison should provide us with information on how teachers adapt their tools for teaching the reading of digital encyclopaedic articles according to their pupils' levels and assumed degree of disciplinarity. What do we expect from the three stages of the comparison? (1) When they first read the encyclopaedic text, the teachers formulate difficulties that they attribute to a group of presumed pupils (l'archiélève, Ronveaux, 2014), then, after discussions with colleagues and the first task proposals, they transform these difficulties into calculated obstacles in the mediation of a task. We think that, at this point, the obstacle outlines an area of close development. This zone will differ according to the primary and secondary levels. (2) The series of tasks planned will be similar in primary and secondary schools, but there will be additions to the plans of secondary school teachers who, in order to discipline the pupils, will want to invite them to rewrite fragments or produce an article. (3) The sequences used at primary and secondary level will be different. Conceptions of the text at primary level are more flexible and less oriented towards a "perfecting" finished product. At secondary level, the final writing task is more akin to a composed commentary, in preparation for the literary dissertation.

Franck, O. (2017). A la recherche de l'archiélève lecteur à travers l'analyse du geste de planification : rôle des élèves dans les modifications de séquences d'enseignement [Doctoral dissertation, University of Geneva]. https://doi.org/13097/archive- ouverte/unige:92154 Renaud, J. (2020). Quelles cibles didactiques viser dans l’enseignement de la lecture documentaire sur support numérique au cycle 3 ? Repères, 61, 223-242. Ronveaux, C. (2014). L'archi-élève lecteur en progression entre tâche, activité et performance de lecture. In B. Daunay (Ed.). Didactique du français: du côté des élèves. Comprendre les discours et les pratiques des apprenants (pp. 123-138). De Boeck Supérieur. Ronveaux, C. and Schneuwly, B. (2018). Lire des textes réputés littéraires : disciplination et sédimentation. Enquête au fil des degrés scolaires en Suisse romande. Peter Lang. Schneuwly, B. and Dolz, J. (2009, eds.). Des objets enseignés en classe de français. Le travail de l'enseignant sur la rédaction de textes argumentatifs et sur la subordonnée relative. Presses universitaires de Rennes. Rouet, J.-F. (2012). Ce que l’usage d’internet nous apprend sur la lecture et son apprentissage. Le français aujourd’hui, 178, 55-64. Souchier, E., Candel, E., Gomez-Mejia, G. (2019). Digital as writing. Theories and methods of analysis. Armand Colin.