Abstract Relational Reasoning
Understanding something new by relating it to something familiar, e.g., seeing that a statistical prediction interval is abstractly similar to a fishing net, is fundamental to how humans think, from expert scientific reasoning to classroom learning. We are investigating the cognitive and neural bases of abstract relational reasoning, with a particular focus on how people understand abstract similarities between things that seem different on the surface. The most valuable similarities are the ones that reveal hidden connections between things that seem different (e.g., the connection between the structure of the atom and the structure of the solar system). These abstract, hidden similarities have been recognized by effective thinkers, from Kepler to Einstein to Steve Jobs, as the fundamental basis for understanding and teaching complex, novel concepts. Our research has initiated the development of a new area of “semantic distance” research in relational reasoning. Semantic distance research addresses the ways in which cognitive and neural processes of relational reasoning change as surface-level differences increase (i.e. as the connections become more and more abstract).
Our research has shown that relational integration, a core process for identifying similarity in relational reasoning, is cognitively and neurally distinct from other components of the relational reasoning process (Green, Kraemer et al., 2006; Green, Fugelsang et al., 2006; Green, Fugelsang et al., 2008). This work focused on analogical reasoning (e.g., “nose is to scent as antenna is to signal”) because analogical reasoning is the most common form of relational reasoning, and its component processes are relatively well understood. Using functional magnetic resonance imaging (fMRI), this research identified a region of left frontopolar cortex that appears to play an important role in relational integration (Green, Kraemer et al., 2006; Shamosh et al., 2008). In a concurrent line of work, we have developed the Micro-Category Account of Analogy, which empirically explains analogical reasoning as the product of multiple categorization processes (Green, Fugelsang et al., 2008). Subsequent research extended these findings to demonstrate abstract analogical reasoning in children on the autism spectrum (Green et al., 2014), including analogical connections between social situations (Green et al., 2017). A new project in this research area is exploring analogical connections across sensory/perceptual modalities.