Note
On E-infinity- and DAlg-Descendability
Article date: August 30, 2026.
Abstract
For a map of -rings, its th finite Amitsur stage is
Antieau and Stefanich asked five questions about multiplicative retractions from these finite stages. We answer all five.
For every prime and every , the restriction map
is descendable but not -descendable, and hence is not -descendable; for the map is the identity. Thus ordinary and -descendability differ even for cochains on finite cyclic -groups.
Complexification is -descendable: the unit
admits a retraction in . The units at stages , , and do not split in , and the unit at stage does not split even as an --algebra map. Thus the least -splitting stage is either or .
Let be a Noetherian -algebra of finite Krull dimension, put , and let be a faithfully flat map of discrete -algebras. Without any finite-type hypothesis on , the unit admits a retraction in derived commutative -algebras
specializing to gives a retraction at stage . The proof uses the first two natural layers of a free derived symmetric algebra, a two-term induced resolution over the Frobenius skew-polynomial ring , and integral flattening.