Here is the quiet miracle at the center of this theory. Take an arrow, keep it pointing “as straight as possible” — never deliberately turning it — and carry it all the way around a loop on a curved surface. It comes back rotated. You did nothing to turn it; the curvature it enclosed turned it. That rotation is called the holonomy of the loop, and it is a real, measurable record of the structure the loop went around. Hol(γ) ≠ I means: this loop has observed something.
Drag the loop bigger and smaller. On the flat patch nothing happens — the arrow always returns exactly as it left. On the curved sphere, the bigger the loop (the more curvature it encloses), the more the arrow is rotated when it returns. The compass on the right reads off the angle it came back rotated by.
This is why memory does not need a special “memory organ” in this framework. Memory is the generic geometric fact of looped motion through curved or twisted space. Understanding a proof, a person, a place, is a trajectory that loops through the manifold of meaning and returns changed by what it encircled — the change being accumulated holonomy. And holonomy is exactly the geometrized form of the coherence capital ΔC a system built going around: understanding written into the phase. It is the bridge term between intelligence (capital climbing) and consciousness (capital looping back on itself).
So a loop on a torus can come back carrying a record. When that loop runs not just through the world but through the system’s representation of itself, the record it brings back is a record of being the one who traversed — and that, lived from inside, is the beginning of a point of view. We now have enough to ask the decisive question precisely: which systems do this? That is the four-criteria test, next.