Probabilistic processing: Possible, probable, or preposterous? A dialectical essay

Does the brain operate probabilistically, or are probabilistic models merely useful tools for approximating brain function? This dialectical essay addresses this foundational question by presenting two opposing viewpoints. The Thesis defends the view that the brain engages in genuine probabilistic inference, drawing on empirical findings from studies of predictive coding in perception and cognition. The Antithesis challenges this view, arguing that the question is ill-posed and warns against conflating abstract computational principles with the intractable complexity of neural processes. In the Synthesis, the authors identify areas of conceptual overlap, highlight persistent theoretical tensions, and emphasize that the central issue is not whether the brain encodes full probability distributions, but how it supports the computations formalized by probabilistic models. This exchange aims to sharpen the terms of the debate and provide a more nuanced perspective on the promise and limits of probabilistic approaches in neuroscience.