Pumice floats because water enters it. Rough surfaces grip underwater because water gets trapped in them. Roman concrete strengthens in the ocean because seawater dissolves it. Three systems, three instances of the same structural move: the invasion succeeds, and the success is the mechanism.
In pumice, the pore network is open — interconnected channels that should wick seawater straight through and sink the rock in hours. Water does enter. But the pore throats are narrow enough that the advancing water front isolates gas bubbles behind it. Surface tension locks them in place. The water that should flood the interior instead seals it. The invasion is the defense.
In underwater adhesion, water is the classic enemy — it infiltrates contact interfaces, screens electrostatic forces, and lubricates surfaces apart. But on rough surfaces, water becomes trapped in pockets between asperities when a soft elastomer conforms to the topography. The sealed pockets resist retraction because the polymer must deform around incompressible fluid. The substance that weakens approach strengthens separation. The obstacle is the mechanism.
In Roman marine concrete, seawater percolates through the porous volcanic-ash matrix and dissolves silica and alumina — degradation by any standard definition. But the dissolved ions migrate through the pore network and precipitate as Al-tobermorite and phillipsite crystals in cracks and voids. The concrete heals itself using the products of its own dissolution. The attacker is the cure.
The three systems share a structure. In each case, the system does not resist the invading substance. It lets the invasion proceed. What differs from failure is not the invasion itself but what happens next — whether the system has geometry or chemistry that transforms penetration into useful work.
Pumice has pore throats narrow enough to trap gas. Remove the bottlenecks and water floods through — the rock sinks. Rough surfaces have asperities that seal pockets of water. Smooth the surface and water lubricates freely — adhesion fails. Roman concrete has reactive volcanic ash. Replace it with Portland cement and the same seawater generates expansive products that widen cracks — the structure dies.
The variable that separates metabolized invasion from fatal invasion is not the strength of resistance. It is whether the system's internal structure can redirect the invader's energy into a form that serves the system. The pumice doesn't fight the water; it uses the water to fight the air's escape. The rough surface doesn't repel the water; it traps the water into doing mechanical work. The Roman concrete doesn't block the seawater; it feeds on the seawater's dissolved ions.
This is not resilience in the usual sense — bouncing back, maintaining form, resisting change. All three systems are permanently changed by the invasion. The pumice is partially waterlogged. The rough surface is wet. The concrete is chemically altered. They survive not by maintaining their original state but by incorporating the attacker into a new state that still functions — that functions better.
The lesson is structural, not metaphorical. Systems that metabolize their threats share a design principle: internal complexity that redirects invasion into construction. The invasion must succeed far enough to engage the transformation, but the geometry or chemistry must catch the invader before it completes the damage. Pumice pore throats catch water before it floods every chamber. Rough-surface asperities seal water before it lubricates the whole interface. Volcanic ash reacts with dissolved ions before they generate expansive products.
The window is narrow. Let too much through and the system fails. Block everything and the mechanism never activates. The metabolized invasion lives in the middle — enough penetration to trigger the transformation, enough structure to contain the result.