In this paper, we examine some of the tenets of the current conceptualisations of communicative competence. Drawing on the empirical data collected in linguistically diverse university classrooms, we show that meaning-making in social interaction is considerably more complex and fluid than is envisaged in theoretical models of communicative competence. We suggest that curriculum and assessment frameworks such as the Common European Framework of Reference for Languages (CEFR), based on existing notions of communicative competence, do not adequately capture the agentive and contingent nature of co-constructed meaning-making and meaning-taking in social interaction, particularly in situations where high levels of ethnolinguistic diversity are fast becoming a norm. We suggest that for benchmarking curriculum and assessment frameworks such as the CEFR to be relevant, they need to be empirically investigated and theoretically critiqued on a regular and systematic basis.