LessWrong argues mental causation isn't needed for consciousness

In a recent LessWrong post titled 'Mental causation is not load-bearing,' jessicata reexamines the philosophical problem of mental causation—whether mental states, like beliefs or qualia, can causally affect the physical world. The author argues that the classic epistemic challenges to epiphenomenalism (e.g., why evolution produced consciousness if it has no physical effects, or why mental reports correlate with experiences) can be resolved without relying on mental causation. Instead, the key is 'intelligible supervenience': the idea that mental facts are not just brute correlates of physical facts, but can be understood as logically or functionally grounded in them—similar to how a virtual machine’s state supervenes on hardware. This kind of supervenience, says jessicata, removes the mystery from the mind-body relationship, without needing causal loops.

The post contrasts two views: epiphenomenalist functionalism and Russellian monism. Epiphenomenalist functionalism accepts that mental states are functional roles (e.g., memory as information storage/retrieval) but denies they have causal powers—a position that seems to violate evolutionary explanations. Russellian monism, on the other hand, posits that the intrinsic nature of physical entities includes proto-mental properties, offering a more parsimonious picture. jessicata suggests that while mental causation is often seen as the natural solution to epiphenomenalism's woes, it's not actually load-bearing in the argument; what matters is whether the mental-physical relationship is intelligible. The post has sparked discussion on LessWrong, touching on AI consciousness, panpsychism, and functionalism's implications for building minds.