Pathological phenomena that occur in the associated graded object Gr_F(A) of a filtered algebraic object (ring, module, sheaf), such as emergence of unexpected nilpotents, loss of reducedness, failure of expected dimension counts, or breakdown of functorial properties.

Principle

The passage from a filtered object to its associated graded is not exact and can separate interacting terms of different degrees; subtleties in the filtration (non separatedness, nonreducedness, torsion) create artifacts in the graded object not present in the original.

Demonstration

A reduced local ring with a filtration by an ideal can have an associated graded ring that is nonreduced: the tangent cone at a singular point may carry nilpotent elements even when the local ring has no embedded nilpotents, altering multiplicity computations.

Misapplication

Assuming invariants computed on Gr_F(A) (Hilbert series, multiplicity, depth) automatically reflect those of A without checking hypotheses like good filtration or flatness of Rees constructions.

Consequence

Incorrect geometric or algebraic conclusions: miscomputed tangent cones, wrong predictions about smoothness or multiplicity, and invalid transfer of regularity or flatness properties between filtered and graded settings.

Reversal

Well-behaved associated graded constructions occur under strong hypotheses (e.g., filtrations coming from powers of a regular parameter ideal in a regular local ring), where Gr_F preserves reducedness and other properties.

Boundary

Applies specifically to filtered-to-graded passage; does not mean the graded category itself is pathological, and excludes cases where the filtration is exhaustive, separated, and compatible with Noetherian or flat hypotheses that guarantee good behavior.

Semantic Tension

Between the associated graded as a computational simplification and its role as an exact invariant: it is a powerful probe of leading-term behavior but can mislead when lower-order interactions are essential.

Synthesis

Associated graded pathology warns that taking leading terms via a filtration can introduce artifacts—nilpotents, lost exactness, and skewed invariants—so one must verify filtration and flatness hypotheses before drawing conclusions from Gr_F.