Representational debt

Representational debt is the accumulated cost of using a simplified or outdated conceptual model long after the system it describes has evolved beyond the model's validity. Like technical debt in software engineering, representational debt accrues interest: the longer a community persists with an inadequate formalism, the more effort is required to migrate to a better one, and the more misleading conclusions are drawn in the interim.

The term captures a pathology common in interdisciplinary fields where models migrate from their home discipline into new domains without adequate scrutiny. The power law model of network degree distributions, carried from statistical physics into biology and social science, is a canonical example: decades of research operated under the assumption that "scale-free" was a binary property, when in fact the empirical support was far weaker than the framing suggested. The debt was not in the model itself but in the representational practice of treating a statistical hypothesis as a categorical truth.

Representational debt differs from simple error in that it is collective and structural. Individual researchers may recognize the limitation of a model, but if the field's vocabulary, textbooks, funding priorities, and peer-review standards all presuppose the outdated framing, the debt is systemic. It can only be resolved through epistemic infrastructure change — new pedagogical practices, revised terminological standards, and disciplinary self-awareness about when a model has become a cage.