Concrete example first
Start with no objects: (X_0=\emptyset).
Suppose closure rules require:
- add the terminal object,
- add the recognition object (H),
- add objects needed to interpret RTL operations on what is already present.
Set (X_{n+1}=UG(X_n)). Once (X_{n+1}=X_n), closure has stabilized. That stage is the generated universe.
Formal definition
The generative closure operator [ UG:P(\mathrm{Ob}(R))\to P(\mathrm{Ob}(R)),\quad R=\mathbf{Sh}(T,J) ] is monotone and inflationary, and it encodes which objects must exist for sheaf semantics, recognition dynamics, and RTL interpretation.
GFRTU is defined as the least nontrivial fixed point of (UG) above (\emptyset): the smallest closed object set supporting the required structure.
Why it matters in GFRTU
- It enforces minimality: no extra objects are assumed.
- It turns local rules into a global universe by fixed-point iteration.
- It gives the “generative” part of Generative Fibered Recognition Trace Universe.