Introduction
Most of our folk theories of self predate the neuroscience. The gap is starting to close, and the picture that emerges is more fluid — and more workable — than the picture it replaces.
Why This Matters
A more accurate model of the self reorients therapy, education, and everyday self-knowledge. It also reduces the shame we attach to states we do not yet know how to name.
The Science
Barrett (2017), in Social Cognitive and Affective Neuroscience, formalised the theory of constructed emotion as an active-inference model, treating emotional experience as the brain’s best interpretation of interoceptive signals given prior learning — not the readout of a fixed circuit. Seth (2013), in Trends in Cognitive Sciences, extended this logic to selfhood itself, arguing that the felt sense of being a self is a predictive inference about the body, with body ownership and subjective feeling arising from the brain’s modelling of its own physiological state.
Current Research
Contemporary work continues to test predictive-processing models across depression, anxiety, chronic pain, addiction, and psychosis, refining a framework that treats symptoms as failures of inference rather than failures of the person.
Practical Implications
For clinicians: teach patients that their inferences are updatable, because they are. For anyone: the self you experience is a model, and models can be revised.
Common Misconceptions
**"I am my personality."** You are a nervous system running a model of yourself. **"If the self is constructed, it isn’t real."** Constructed does not mean unreal. Buildings are constructed.
Key Takeaways
- Emotion and selfhood are best modelled as active inference.
- What we experience as fixed traits are stable patterns of prediction.
- The self is a model — and models can be revised.