Publication date: Aug 14, 2026
Diagnosis is indispensable to medicine, yet diagnostic labels often acquire more authority than their evidence can bear. They come to function as if they named natural entities that explain, authorize, and stabilize clinical action. This article proposes a pragmatic criterion for evaluating that ontological surplus: the Conservativity Test. The test asks which clinical, institutional, explanatory, and self-interpretive consequences follow from treating a diagnostic label as a disease entity, but would not already follow from a label-free base description of symptoms, signs, impairments, risks, mechanisms, values, and lived experience. It does not presuppose that diagnoses are either natural kinds or mere social constructions. Rather, it distinguishes conservative shorthand, justified non-conservativity, and reifying non-conservativity. To avoid circularity, the article develops a minimal-ontological account of base description and connects it with pluralist realism and causal-cluster accounts of kinds. Situating the test within the epistemology of diagnosis, it treats diagnostic labels as revisable causal-explanatory hypotheses and spells out why unjustified ontological surplus carries epistemic, clinical, ethical, and institutional costs. It then identifies four modes of diagnostic reification: explanatory, threshold, nominal-institutional, and looping. Prediabetes, ADHD, and long COVID serve as stress tests for the framework. Phenomenologically, the test protects the lived complexity of illness against both objectifying reduction and epistemic dismissal. Ethically, it offers a participatory audit for clinicians, guideline panels, researchers, and patient communities deciding when diagnostic names should coordinate care, justify intervention, or carry stronger ontological commitments.
Semantics
| Type | Source | Name |
|---|---|---|
| disease | MESH | Prediabetes |
| disease | MESH | ADHD |
| disease | MESH | long COVID |