A Developmental Stability Theory of Predictive Hierarchies: Why Some Perceptual Architectures Remain Stable Under Perturbation

The human brain is not a passive camera recording the world; it is a relentless predictor. Every moment, it constructs a model of what is happening around us, constantly checking that model against the raw data streaming in from our senses. When the data matches the prediction, the brain moves on. When there is a mismatch, it generates a signal of error, a tiny alarm telling the system to update its model. This process, known as predictive processing, happens in layers, with higher levels of the brain sending down expectations and lower levels sending up corrections. For most people, this system is remarkably stable. But for those with schizophrenia, the system can fracture. They may hear voices that are not there, hold beliefs that defy evidence, or feel that their own thoughts are being inserted by someone else. The central mystery has always been why this happens to some people and not others, and why it typically strikes in late adolescence rather than at birth.

A new theory proposes that the answer lies not in what the brain calculates, but in how its physical architecture holds up under pressure. Imagine the brain's predictive layers as a complex web of connections. If this web is too large and too tightly tangled, a small error can spiral out of control, turning a simple mismatch into a full-blown hallucination or delusion. The theory suggests that the brain's stability depends on a delicate balance between the size of this web, how strongly its parts are linked, and how loudly the brain listens to its own errors. Crucially, this balance is set during early development by the sensory experiences a child has. If a child is born without sight, the brain builds a different kind of web, one that is naturally more stable against the specific kind of collapse that leads to psychosis.

The researchers behind this theory, led by Sakina Bukhari at Monash University Malaysia, developed a mathematical framework to test this idea. They treated the brain's predictive hierarchy as a dynamic system, similar to how engineers study the stability of a bridge or a building. They asked a simple question: under what conditions does this system remain steady, and when does it tip into chaos? Their model identified a specific tipping point. The system becomes unstable when three factors multiply together to exceed a critical limit. The first factor is the dimensionality of the dominant sense. Vision, for instance, is incredibly rich and complex, involving thousands of independent details that the brain must track simultaneously. The second factor is the strength of the connections between different parts of the brain. If these connections are too loose, the system is weak; if they are too tight or too numerous, they can create feedback loops where a single error reinforces itself endlessly. The third factor is the gain, or the volume, at which the brain listens to its own error signals. If the brain turns the volume up too high on its own mistakes, it becomes hypersensitive to noise.

The theory posits that for a person to develop schizophrenia, their brain must have a specific combination of these traits: a highly complex sensory foundation (like vision), strong and numerous connections, and a moment when the volume on error signals gets turned up too high, perhaps due to changes in brain chemistry during adolescence. When these three elements align, the system loses its stability. Errors that should be dismissed as noise begin to amplify, creating false realities that feel as real as the physical world. This explains why the condition often appears in late adolescence: that is when the brain's sensory architecture has finished its major construction phase, and when the chemical systems that regulate error signals undergo significant maturation.

One of the most striking predictions of this theory comes from looking at people who are born blind. The researchers point to a specific type of blindness called congenital cortical blindness, where the eyes are healthy but the part of the brain that processes vision never develops. In a large study of nearly 470,000 children in Western Australia, not a single child with this condition developed schizophrenia, whereas children who lost their sight later in life did not share this protection. The theory explains this by noting that these children never built the high-dimensional visual architecture that makes the brain vulnerable. Without the massive, complex web of visual predictions, the brain's structure remains simpler and more robust. Even if their error signals get turned up high later in life, the system does not have the necessary complexity to spiral into the kind of instability seen in schizophrenia. This is not a general shield against all mental illness; the theory suggests these individuals are still susceptible to other conditions like depression or anxiety, but they are specifically spared from the unique instability that causes psychosis.

The model also sheds light on why the symptoms of schizophrenia are so varied. When the system becomes unstable, it does not collapse all at once. Instead, specific parts of the network begin to drift. If the instability affects the parts of the brain responsible for organizing thoughts and words, the result is disorganized speech. If it affects the parts that decide where a sound or image is coming from, the person may hear a voice that feels like it is outside their head. The theory provides a way to distinguish between these experiences. A hallucination, for example, is not just a false signal; it is a false signal that the brain has incorrectly tagged as coming from the outside world. This happens when the brain's internal prediction of its own actions fails to cancel out the signal, leaving the brain to conclude that the voice must have an external source.

The researchers also used their model to explain the effects of psychedelic drugs. These substances are known to temporarily loosen the brain's grip on its own predictions, making the world feel more fluid and strange. The theory suggests that these drugs work by lowering the brain's confidence in its own prior beliefs, which effectively turns up the volume on incoming sensory errors. This pushes the system closer to the instability threshold. However, because the underlying structure of the brain remains unchanged, the effect is temporary and reversible. In contrast, for someone with a brain that has already developed a fragile architecture, this same mechanism can trigger a permanent shift into psychosis.

The theory does not claim to have solved the mystery of schizophrenia completely. It is a framework that connects molecular biology, brain circuitry, and human experience into a single story. It suggests that the risk of developing the condition is not just about bad genes or bad chemistry, but about how those factors interact with the brain's physical structure during its most critical years of growth. The author is careful to note that while their model fits existing data, such as the protective effect of congenital blindness and the timing of the illness, many of its specific predictions have yet to be tested in the lab. They propose new experiments to measure the stability of the brain's connections in people at risk, looking for the specific signs of a system that is teetering on the edge.

Ultimately, this work reframes the question of mental illness. Instead of asking what is broken in the brain, it asks how the brain's design holds up when the pressure is applied. It suggests that the brain is a system that can be beautifully stable or dangerously fragile, depending on the path it took to get there. For those born without sight, the path was different, and the result was a system that, by chance, avoided the specific trap that catches so many others. This insight offers a new way to think about prevention and treatment, suggesting that the key to stability might lie in understanding and preserving the delicate architecture of the brain's predictive networks.