Quantifying Conceptual Flexibility in a Compositional Network Model

A single concept can manifest in many varied forms, depending on the context in which it is activated. That is, concepts appear to be flexible rather than static. Here we implement a compositional model of conceptual knowledge in which basic-level concepts are represented as graph theoretical networks, with the specific goal of quantifying conceptual flexibility. We collect within-concept statistics using online participants, construct network models, and validate these models in a classification analysis. We then extract network measures and find that network diversity and core-periphery structure correspond to conceptual flexibility and stability, respectively. These results suggest that a compositional network model can be used to extract formal measures that are interpretable and useful in the study of conceptual knowledge.