Predictive Coding Corollary Discharge and the Self Other Boundary.pdf
One of the most fundamental tasks of a nervous system is to distinguish what is happening inside the organism from what is happening outside it. Without this distinction, thoughts can feel like voices, intentions can feel like external control, and the sense of a coherent self begins to fray. Modern computational neuroscience explains how the brain normally maintains this boundary—and what happens when the mechanism fails—through the linked concepts of predictive coding, corollary discharge, and source monitoring.
This article provides the foundational neuroscience that underpins the RBSI account of psychosis-spectrum experiences. The language is kept accessible while remaining faithful to the peer-reviewed literature.
The dominant contemporary view is that the brain does not passively receive sensory data and then interpret it. Instead, it continuously generates predictions about the causes of its sensory inputs and updates those predictions when they are wrong. This framework is known as predictive coding (or, in its more encompassing form, active inference).
Karl Friston’s free-energy principle formalizes the idea. Any self-organizing system that persists over time must minimize the long-term average of surprise (or free energy). It does so by maintaining an internal generative model of the world and acting to make that model’s predictions come true, or by updating the model when predictions fail. Perception, action, and learning are all expressions of the same underlying imperative.
In hierarchical predictive coding, higher levels of the brain generate predictions that are sent downward to lower levels. Lower levels compare those predictions with incoming sensory data and send the mismatch—the prediction error—back upward. The system’s goal is to minimize these prediction errors over time. Crucially, not every prediction error is treated equally. The brain assigns a precision (a kind of confidence weighting) to both its predictions and its sensory data. High-precision prediction errors drive strong updates; low-precision ones are largely ignored.
A specialized and particularly important form of prediction is corollary discharge (also called efference copy). Whenever the brain issues a motor command—whether to move a limb or to generate inner speech—it simultaneously sends a predictive copy of that command to the relevant sensory areas. This copy allows the brain to anticipate the sensory consequences of its own actions and to dampen or “cancel” them.
In the auditory domain the process is especially clear. When we speak out loud, or even when we silently imagine speaking, a corollary-discharge signal travels from speech-production regions to auditory cortex. In healthy individuals this signal reduces the auditory cortical response to the self-generated sound or thought. The reduction serves as a neural tag: “This event was caused by me.”
Recent EEG research has made the mechanism visible. When healthy participants silently imagine a syllable that matches an external sound, their auditory N1 brainwave response is suppressed. People currently experiencing auditory verbal hallucinations often show the opposite pattern: an enhanced rather than suppressed response. Their brains treat the self-generated signal as more surprising, not less. This is direct evidence that the predictive tagging of inner speech has failed.
Corollary discharge is one concrete implementation of a broader cognitive ability called source monitoring—the capacity to determine the origin of a mental event. Was this thought internally generated or did it come from the external world? Did I say this or only imagine saying it? Did this memory originate in perception or in imagination?
Decades of research show that people with schizophrenia-spectrum conditions frequently exhibit source-monitoring deficits. They are more likely than healthy controls to misattribute internally generated material to an external source. This bias appears across experimental paradigms and is particularly pronounced for auditory verbal material. The deficit is not simply poor memory; it is a specific difficulty in binding an experience to its correct origin.
When corollary discharge and source monitoring function normally, the boundary between self and other remains intact. Inner speech is recognized as inner; intentions are recognized as one’s own; bodily sensations are experienced as belonging to the body.
When the predictive tagging fails, the same internal events arrive without the usual “I caused this” label. Because the brain’s default assumption in the absence of a clear self-tag is often that the event is externally caused, the result is the classic phenomenology of psychosis: