The Self as the Brain's Highest-Level Story About Itself
Before the narrator, there was only the telling.
The sense of being a 'self' is not a given — it is the brain's most abstract model, a learned compression of experience. Like any model, it can lose its grip, revealing a selfless awareness that neuroscience and contemplative traditions both recognize.
The Translation
AI-assisted summaryFamiliar terms
The self, from a predictive processing standpoint, is the deepest node in the brain's Generative hierarchy — the most abstract, most compressed model the system maintains. It is not a metaphysical given but a learned prior, assembled over developmental time as the infant gradually differentiates its own agency from the external world. The recognition that the caregiver is a separate intentional agent is a pivotal moment in this construction: it implies, by symmetry, that there must be a corresponding locus of agency on this side of the sensory boundary. From that seed, the full narrative self elaborates.
Because the self-model is a prior like any other, it is subject to the same precision-weighting dynamics that govern all perceptual inference. Psychedelics, which are thought to flatten the precision landscape of high-level priors, reliably attenuate or dissolve the sense of selfhood. Long-term contemplative practice achieves something similar through sustained attentional retraining. In both cases, the phenomenological report converges: awareness persists, but the feeling of being a bounded subject — a someone to whom experience is happening — drops away.
This convergence is significant. From the free energy framework, non-dual awareness is not an exotic spiritual state but the natural residuum when the highest-level compression in the generative model loses its Precision Weighting. What meditators call 'awareness without content' is simply the system's ongoing inference minus its self-referential wrapper. The self was never the ground of experience — it was the last and most persuasive story layered on top of it.