If as suggested by Boas (1911), Sapir (1921, 1927), Whorf (1956), Gumperz and Levinson (1991) or Lucy (1997) language shapes experience, language can also create worlds of its own, out of our remembrance of things past, our projects or dreams of things to come and the figments of our imagination, thanks to the linguistic forms we perform with all the semiotic resources at our disposal. During their first years of life, children experience language (Ochs 2012) as they are involved in the mundane activities of their everyday lives, through sound, texture, visual and embodied forms. In this process, children progressively language their experience: they learn to filter experience and shape it, refer to it, index it into language forms. Enactments of language performed by adults and children are multiple, diverse, multimodal. Children can language about objects and events that they perceive in the here and now, but also and progressively, about objects and events they are not perceiving and experiencing in the here and now. They memorize scenarios with their strings of events and the temporal unfolding of language in daily situations and extend linguistic forms to similar situations. Weist (1986) suggests that in the course of child language acquisition, the concept of time develops in several stages organized according to Reichenbach’s principles (1947). In the final stage, children master reference time, they have entered the TMA system, and have the ability to tell stories or talk about imaginary people, objects or events. It is fundamental to understand how children can learn these complex linguistic functions as they do not correspond to referents that can be perceived or manipulated. Our hypothesis is that children can productively use the forms to express displacements and abstract reference around their fourth year because they have been socialized to them very early on, thanks to their interactive input. We tested this hypothesis on the longitudinal data of seven French-speaking children videotaped monthly with their parents. We focused on a form that is mainly used to mark a displacement between the speaker and the here and now, the imparfait (Patard, 2007). Our goal was to understand how children can learn its function. We categorized the sequences into several genres that were either displaced and good candidates for switches of reference time, or non-fiction grounded in the here and now. We coded all the verb forms in the adults’ and children’s productions. The results showed that the imparfait was used very early on by the adults to express displacement. For some genres such as narratives without books, the imparfait was used twice more than in other situations in both adults and children. This trend was even stronger when narrating self-experiences. In pretend play, only the children used the imparfait, which suggests that they have the ability to generalize its use to relevant contexts. Children can thus find exemplars in situations that afford specialized forms within recurrent scripts and can generalize them to other situations with similar affordances.