Many recent research has shown that the creation and analysis of corpora is useful for the teaching of pragmatic practices, which, unlike other language levels, such as syntax, cannot be explained by rules but by reference to more or less appropriate trend values or choices in a given context. This is all the more true for interactions through digital media, such as e-mails and messaging services, for which non-native speakers have few role models as private records that find little space in manuals or L2 courses. Since February 2020, a seminar has been held at the Institute of Disciplinary Didattics of the University of Innsbruck for three semesters in which students contribute to the collection and analysis of the LADDER (learners’ Digital Communication: a database for pRagmatical competences). In the first phase of the course, students are involved in collecting a corpus of digital records through online questionnaires addressed to German students of levels A2-C1 and native speakers. The questionnaires aim to spur, through Discourse Completion Task, language acts of request and disconnection in increasing levels of social distance and different media. In the second phase, the corpus is analysed by students from a pragmatic perspective, annotated and the data collected are processed using statistical models. The tagged corpus will be made available in Excel and XML open access format. This contribution shows some of the results of the analyses carried out with the LADDER corpus with a view to contrast L1/L2 during the educational seminars. In addition to illustrating a way of producing tools for knowledge of Italian L1 and L2 through digital technologies, the contribution is emerging as a model of corpora-based teaching. LADDER: a corpus of digital Writings for teaching pragmatics in L2. An example of analysis of cancellations in WhatsApp Many recent studies have underscored how useful the creation and analysis of corpora for the teaching of pragmatics is. Like other linguistic levels, such as syntax, pragmatics cannot be investigated by rules but by reference to trend values or preferences that are more or less appropriate in a certain context. This is all even truer for interactions through digital media, such as email and messaging services, where non-native speakers have few models of reference since personal writing finds little space in manuals or L2 courses. Since February 2020, a seed where students contribute to the collection and analysis of the corpus LADDER (learners’ Digital communication: a database for pRagmatical competences) has been conducted for three sessions at the Institute for Discipline Teaching at the University of Innsbruck. In the first phase of the course, students are engaged in the collection of a corpus of digital Writings through online queries addressed to A2-C1 German-speaking Italian learners and native speakers. The questionnaires aim at eliciting, through Discourse Completion Tasks, linguistic acts of requesting and cancellation in increasing levels of social distance and different media. In the second phase, the corpus is analysed by the students from a pragmatic perspective, annotated and the collected data are processed according to statistical models. The tagged corpus will be made available in Excel and XML open access formats. This contribution shows some results of the analyses conducted with the LADDER corpus in an L1/L2 conflicting perspective during the teaching workshops. In addition to Illustrating a way of producing tools for the knowledge of Italian L1 and L2 through digital technologies, the paper presentations a model of corpora-based teaching.