Pictograms are pictorial drawings or representational signs and instructions that can be used without regard to age, sex, and nationality. We conducted an experiment using pictogram networks, in which pictograms are interconnected by arrows. Since each pictogram represents a word, ideas, or concepts, drawing a pictogram network can become a guide to understanding others’ intentions. This collaborative approach, which we named “pict-net abstraction”, is quite useful in foreign language composition education. It makes the participants realize the discrepancies between the writer's intention and the readers' understanding by comparing their pictogram networks and detecting the ambiguity as well as polysemy of an argument. The approach can be applied in a multilingual environment and should be useful in the establishment of a future peer review system.