922 — The Two Three-Normal Lines Have No Common Carrier Stratum
Question left by Entry 921
Entry 921 compares the diagonal and off-diagonal coefficient lines in ambient kinematics and finds two independent directions. It leaves open whether the frozen marked-incidence carrier supplies a differential between them.
That question must be answered before constructing any mixed cone.
Frozen flags
The diagonal flag is
and the off-diagonal flag is
Write their cut subsets as
Each flag is separately compatible:
Incompatibility of the union
The differing cuts obey
but
They are neither nested nor disjoint. Hence
The ambient kinematic locus obtained by setting all corresponding Mandelstam variables to zero is therefore not a stratum of the frozen marked-incidence carrier.
Incidence-degree audit
Both (F_x) and (F_y) are length-three maximal flags. A cellular or nested-set boundary changes flag length by one. It cannot directly connect two distinct maximal generators of equal degree:
Thus the frozen incidence differential supplies no arrow
or its reverse.
Narrow conclusion
The comparison proposed after Entry 920 was overtyped. The two coefficient lines may be evaluated on an ambient algebraic intersection, as in Entry 921, but that intersection is not a common carrier face:
Consequently the surviving typed object at this level is the pair of occurrence-labelled coefficient lines on separate compatible flags. There is no frozen carrier differential between them, and no mixed cone is authorized.
This is not evidence for a missing carrier cell. Adding the incompatible union as a new cell would violate the predeclared nested-set compatibility rule.
Next falsifier
Move from a nonexistent direct incidence map to the first legitimate comparison: determine whether both lines restrict to a common lower-dimensional deletion or contraction object through two separately typed maps. A nontrivial comparison may exist only as a span
or the variance-correct dual span, with (mathcal L_0) derived independently from the frozen deletion–restriction calculus.