The experiment is not validation. It is the forge. The architecture goes in soft. The data strikes it. What remains is harder than what entered.

The first ECG experiment broke the encoding. The five-beat sliding window smeared the anomaly across the normal. The second experiment broke the assumption. We was not needed. The architecture told us itself — single Geruon, 0.8 seconds, forty percent. The third experiment broke the redundancy. Phase was already in the gid chain. The arrow was carrying it twice. M1.1 had solved this before we knew it was a problem.

Each strike of the data against the architecture revealed something that was always there. The window is an aperture. The boundary is a detection. The signature carries time. None of these were designed. They were revealed by the impact. The architecture went into the forge soft. The data struck it. What came out was harder.

Validation is asking: does it work. Tempering is asking: what is it made of. The ECG experiment did not validate the architecture. It showed the architecture what it was made of — and what it was not made to do. The phase experiment did not validate the gid chain. The gid chain showed it had already carried the answer. Validation confirms what you built. Tempering removes what you did not need to build.

Every experiment in this project has been a strike. The encoding. The window. The query. The phase. The quantum. The collective. The bookshelf. Each time the data struck, something fell away. Not because it was wrong. Because it was never part of the architecture. The forge does not break good steel. It burns away what was never steel.