The Signal of Abstraction: Conceptual Presence, Counterfactual Richness, and the Calibration of Mathematical Understanding

The same mathematical object can feel abstract and hollow to a novice yet concrete and almost tangible to an expert. Most accounts treat concreteness and abstraction either as properties of the objects themselves or as stages on a developmental ladder. This article develops a third position, in which the felt abstractness of a concept is a metacognitive signal rather than a feature of the concept and rather than prediction error itself. Drawing on relational accounts of mathematical abstraction, predictive processing, research on processing fluency and metacognitive experience, and first-person reports of working mathematicians, I argue that conceptual presence, the feeling that a concept is a real and manipulable thing, is produced by the counterfactual richness of the generative model a learner holds: a concept feels concrete when the learner can cheaply, consistently, and abundantly simulate what would follow if the concept were perturbed, combined, or viewed from another representation. Felt abstractness is the signal that this counterfactual generative capacity is impoverished or held with low confidence. The signal is useful but imperfect. A calibration analysis distinguishes genuine mastery from false fluency, including the illusion of explanatory depth and pseudo-structural symbol manipulation, and identifies social interaction as an external calibration device. A worked trajectory through quotient groups shows how the theorytracks development without overgeneralizing, because unresolved frontier problems remain legitimately abstract even for experts. The article closes with competitor accounts and testable predictions that separate counterfactual richness from familiarity, error, and imagery, offering a candidate empirical program.