In any communicative interaction, a speaker’s language competence is embedded in the dy-namics of mutually oriented and jointly coordinated courses of activities: each contribution is not only formatted linguistically; it is also articulated with those of the interlocutor(s), it is grounded in the sequential organization of mutual activities and it is inscribed in the manage-ment of turn-taking. As a consequence, engaging in communicative interaction presupposes on the part of the participants the capacity to deal with these different layers of talk and their co-ordination. Interaction-oriented research, and most centrally research emanating from conversation analy-sis, has documented different facets of the complexity of talk-in-interaction. The same line of research has generated an important body of studies that have empirically documented the specifics of second language talk-in-interaction (for second/foreign language acquisition and use see the collection of studies in the Modern Language Journal, 2004: 88/4 and in Gardner & Wagner, 2004; for earlier investigations see de Pietro, Mathey & Py 1989; Dausendschön-Gay & Krafft, 1994, inter alia). However, a dialogic model of language as well as an empirically vali-dated definition of interactional competence remain to be elaborated. In particular, the field of second language acquisition (SLA) research is still lacking an understanding of the details of interactional development over time (cf. Hall, 2004; Kasper, 2004; Markee, 2008; Wagner, 2004). Also, while there is an increasing call across Europe, within the field of language educa-tion, for communicatively oriented evaluation scales for L2 competence, the interactional di-mensions of communicative competence remain underspecified. Finally, despite rich and in-tense investigations on how to test oral capacities (Bachmann, 1990; Lazarton, 2002, inter alia), the domain of language testing is still struggling to generate operationalizable criteria allowing us to assess interactional competence. In the quoted fields of investigation, there is an urgent call for fundamental research to be car-ried out on the nature and the development of interactional competence in a second language (L2) as well as for methodological solutions allowing us to track L2 interactional development over time. Today, however, systematic empirical research into second language interactional competence is at its very initial stage. The present project sets out to address these concerns in the domain of instructed and non-instructed second language learning. Two contrastive sets of audio and video-taped data are analyzed, one comprising classroom interactions, the other spontaneous every-day conversa-tions. The first set of data, already available, consists of a corpus of classroom interactions involving lower intermediate and advanced learners of French L2 at secondary I and secondary II level in German-speaking Switzerland (cross-sectional design); the second set, yet to be col-lected, consists of a series of dinner table conversations recorded across 10 months by 10 par-ticipating Swiss-German L1 speakers sojourning in a francophone region (longitudinal design). The project proposes to study, within each of these datasets, a series of recurrent communica-tive micro-activities (which we will call “actional microcosms”), such as disagreeing, introducing a story, opening or closing a conversational episode. Analyses will be carried out onto how the way these micro-activities are organized by the learners changes over time. Changes over time will be documented in each of the datasets on behalf of a series of analytic parameters, includ-ing turn-taking, topic organization and linguistic as well as multimodal resources used by the learners. Cross-contextual analysis (i.e. comparison between the results for the two datasets) will be carried out in order to shed light on context-dependent as well as more context-neutral traits of interactional development. This set-up is designed to (a) identify and empirically validate an initial set of components of interactional competence that are fine-grained enough for capturing speakers L2 capacities more precisely than the general categories such as ‘participation’ or the ability to ‘handle social exchanges’ currently used in language testing and SLA research. (b) provide empirically based insights into the development of interactional competence over time, and (c) offer an initial analysis of how linguistic (e.g. lexicon, morphosyntax), sequential (e.g. turn-taking) and social abilities (e.g. exhibiting alignment or shared orientation) interact in the development of interactional competence. It is anticipated that results of the project enrich our understanding of how language as a so-cial phenomenon works and how it is acquired. In particular, and as regards fundamental re-search, the project is expected to contribute to current efforts toward a specification of what L2 interactional competence is, and how it can be observed. In addition, the project develops and puts to test a concrete methodology for investigating interactional competence over time. Re-sults are also anticipated to be relevant for applied research and feed into current develop-ments in language testing as to the search for operational descriptors for the assessment of interactional competence.