The initial hypothesis examined in this paper is that Hungarian preschoolers assign to sentences containing two numerical quantifiers and a distributivity marker the same isomorphic distributive interpretation as Hungarian adults do. This hypothesis is partially refuted by Experiment 1, a truth value judgement task, and Experiment 2, a forced choice task, which show that children can access distributive readings, however, they tend to accept both isomorphic and inverse scope. Experiment 3, an act-out task, demonstrates that if there are no strong pragmatic cues to enforce a distributive interpretation, children’s primary interpretation is the collective reading. This leads us to the formulation of a new hypothesis: if the default reading of a doubly quantified sentence for preschoolers is the collective interpretation, in line with scope economy, then a distributive reading always represents the revision of the collective interpretation. This is confirmed by Experiment 4, showing that inverse answers have an increased reaction time. The new hypothesis can explain the lack of isomorphism in children’s interpretation of distributive scope as follows: since the distributive reading is dissociated from the linear flow of speech, the linear order of the two quantifiers does not necessarily determine scope order; children can base relative scope on the hierarchy of grammatical functions, on pragmatic cues, etc.