Looping effects and the cognitive source of epistemic injustices: A simulation study

Patient–clinician interactions are central to diagnosis and treatment in mental health care. They involve inferential processes grounded in patients’ reports of symptoms and clinicians’ interpretations of signs. These processes are liable to microdynamic looping effects that drive diagnostic belief formation. In facing the hermeneutic challenge of reaching a shared ascription of the patient’s condition, such looping effects can lead to misalignment and give rise to epistemic injustice if a credibility deficit exists between the patient and the clinician. In this paper, we use the active inference framework to develop a computational model of looping microdynamics that captures how patients and clinicians reciprocally interpret symptoms, attribute causes, and negotiate diagnostic categories over time. We present a simulation study showing how specific interactional patterns can amplify or mitigate credibility deficits, depending in part on clinicians’ attributional attitudes toward patients’ self-interpretations. Our results suggest that, in contexts of diagnostic disagreement, positive disconfirmation—offering an alternative explanatory framing rather than simply negating the patient’s perspective—may reduce the risk of epistemic injustice. Overall, the model highlights how looping microdynamics structure mutual belief formation in clinical encounters and how attention to these dynamics can support more epistemically responsible clinical communication.