14 Feb 2024

Conceptual graphs as understanding

Students struggle more in classes where they have less background knowledge of the subject. Research to expand an area of knowledge is significantly harder than learning. Careful metaphors are powerful tools to guide someone building an understanding of a new domain.

All three of these phenomena can be partially-but-usefully explained with a single, straightforward model. Say understanding is a conceptual graph; individual concepts densely linked by clear relationships. Building understanding, then, consists of adding nodes and relationships to your personal graph for the domain.

In a class where you don’t have sufficient background knowledge, your graph is sparse. You’re introduced to new concepts, but you don’t have a structure in which to place them, or other concepts with which to link them. You’re effectively lobbing nodes into an empty space.

When you’re doing research at the edge of a domain, you presumably have the background knowledge – but you’re trying to identify novel concepts, and you only have relationships coming from one angle. If learning is the process of filling-in a known network, you have the benefit of relationships coming from many directions to help you triangulate the concept. In research, you’ve only got potential connections coming from one side, the current body of knowledge.

Metaphors can be thought of as overlaying different conceptual networks. You have an incomplete graph for the domain you’re trying to learn, and you have a more developed one for the metaphor’s domain; if you can find touchpoints – nodes that show the same sorts of relationships in each graph – then that can help you learn the new domain by showing you where there might be new concepts beyond those you’ve already integrated.