A review of the current literature shows that by the age of two and a half (and probably earlier), children have already acquired a rich working knowledge of human intentionality and goal-directed action (Stein & Albro 1997: 7; Mandler 1998). The paper focuses on the ways in which children use this knowledge to tell stories from pictures. The story is the description of the actions performed by animate actors. We distinguish the main actors (protagonists in the narrative line) and the background actors (participants in the narrative field) who can observe and interpret what is going on in the main action. So the narrative text contains not only the action presented by the story-teller (landscape of action) but also how this action is interpreted by the story characters (landscape of consciousness). They are all thinking minds who can think similarly or differently about the plot. And the narrator uses characters’ minds to produce different representations of the story (Bokus 1998, 2000). The narrator can confront one interpretation with another, and a) makes choices of the “true” representation of the main action (in doing this the child plays the role of the omniscient and omnipresent story-teller who is directly in touch with the ontology of the story), or b) presents a possible but not a certain story reality (the listener is not told how things are but rather how they seem to be). Therefore we can speak about the interplay of the narrator’s mind and the minds of story characters in a kind of internal narrator’s dialogue. The storyteller creates different minds and alternative ways of interpreting the main action. Also shown are examples of such inter-mind phenomena in the stories told by preschool children.