Assuming a cognitive paradigm for the immune system, we explore different forms of immune dysfunction that emerge from a modeling exercise based on new methods for the study of mental disorders (Wallace and Fricchione 2024). One form, associated with a thin tailed underlying probability distribution, the Exponential, characterizes monotonic degradation under rising stress, perhaps related to ‘ordinary’ stress-related immune failures like activation of latent infections and cancers, or premature aging. In contrast, the fat tailed Cauchy Distribution generates an explosive temperature-analog spike that may represent a form of full-blown autoimmune attack arising from stress-induced activation of inherent genetic or epigenetic ‘risk factors’. It appears that relatively subtle changes in the underlying probability distribution can have great impact on immune—or other—system dynamics, providing a potentially powerful target for evolutionary selection. The probability models explored here can be converted into new statistical tools for the analysis of experimental and observational data.                                             — Broderick et al. (2023).                                                    — Wallace (2022).