> We exist in a reality that is inherently chaotic. We can approximate but never reach. But our approximations help us move around, because as we can see the ch

> We exist in a reality that is inherently chaotic. We can approximate but never reach. But our approximations help us move around, because as we can see the chaos also has some kind of structure Yes, obviously models can be useful approximations of reality. I'm not rejecting all models, I'm rejecting the models which do a poor job of approximating reality. But if they're too inaccurate, they aren't useful. The point here is that the coherent, singular "self" model of reality is too inaccurate to be useful any more. Neuroscience has demonstrated the idea of a coherent, singular self to be wrong. When we're asking questions like at the beginning of the article ("Did Anna Karenina really want to die?") the singular, coherent model of self isn't useful. This shouldn't be taken as a criticism of those who believed in the singular self before neuroscience. They created hypotheses, and the experiments which tested those hypotheses and resulted in our current understanding of the self would not have happened without their work. We can remember them as history. But we can't treat these disproven hypotheses as equally valid to more nuanced understandings of the self, out of some sort of nostalgia for philosophy. This isn't a rejection of the practice of modeling reality, it's a rejection of the models which our understanding has grown beyond, in favor of more accurate models.