A Way Diagrams Explain: Analysis Based on Consequence Matching

Diagrams are often used as a means to provide explanations as part of interactions between two agents. While common, this use of diagrams has not been widely examined in the literature. We show that diagrams are particularly useful in these situations because it is possible to observe and reason about an appropriately constructed diagram as a proxy for reasoning about the target domain. This is possible because of “consequence matching” which can occur when the logic of the diagram is designed to match the logic of the target domain in ways that are relevant to the explanation. We illustrate the relevant phenomena by way of examples using a range of common information graphics. We contrast the kind of diagrammatic explanations that we have identified with various kinds of “visual explanations” that have been explored in cognitive science, computer science and information design.