The complete surviving hatan corpus contains only hatandane in Luke 1:71 and hatandans in Luke 6:27. In Luke 1, fijandam supplies enemies while hatandane supplies those hating; in Luke 6, hatandans supplies enemies while fijandam supplies those hating. Both Codex Argenteus pages are directly inspected. The lexical swap—and the class-III versus class-I dispute—cannot be represented by one timeless translation arrow.

Project Wulfila lists 42 candidates under fijan and 26 under the enemy noun fijands. Their intersection is exactly 21 surface tokens. Ayora Estevan explains the collision: *fi(j)and- had lexicalized as enemy before Gothic, while Gothic still possessed the productive present participle one who hates. Case and syntax resolve some passages; the remaining 21 stay public as noun-versus-participle ambiguity.

Project Wulfila lists 26 candidates under the Gothic enemy noun and 42 under the verb fijan; exactly 21 surface tokens belong to both inventories. Ayora Estevan explains why: Proto-Germanic had already lexicalized one who hates as enemy, while Gothic retained the productive participle. The five noun-only tokens are directly inspected; the 21-token overlap remains public instead of being counted twice as certainty.