Lemma 10.136.17. Let R \to S, S \to S' be ring maps.
If R \to S and S \to S' are syntomic, then R \to S' is syntomic.
If R \to S and S \to S' are relative global complete intersections, then R \to S' is a relative global complete intersection.
Lemma 10.136.17. Let R \to S, S \to S' be ring maps.
If R \to S and S \to S' are syntomic, then R \to S' is syntomic.
If R \to S and S \to S' are relative global complete intersections, then R \to S' is a relative global complete intersection.
Proof. Proof of (2). Say R \to S and S \to S' are relative global complete intersections and we have presentations S = R[x_1, \ldots , x_ n]/(f_1, \ldots , f_ c) and S' = S[y_1, \ldots , y_ m]/(h_1, \ldots , h_ d) as in Definition 10.136.5. Then
for some lifts h_ j' \in R[x_1, \ldots , x_ n, y_1, \ldots , y_ m] of the h_ j. Hence it suffices to bound the dimensions of the fibre rings. Thus we may assume R = k is a field. In this case we see that we have a ring, namely S, which is of finite type over k and equidimensional of dimension n - c, and a finite type ring map S \to S' all of whose nonempty fibre rings are equidimensional of dimension m - d. Then, by Lemma 10.112.6 for example applied to localizations at maximal ideals of S', we see that \dim (S') \leq n - c + m - d as desired.
We will reduce part (1) to part (2). Assume R \to S and S \to S' are syntomic. Let \mathfrak q' \subset S be a prime ideal lying over \mathfrak q \subset S. By Lemma 10.136.15 there exists a g' \in S', g' \not\in \mathfrak q' such that S \to S'_{g'} is a relative global complete intersection. Similarly, we find g \in S, g \not\in \mathfrak q such that R \to S_ g is a relative global complete intersection. By Lemma 10.136.9 the ring map S_ g \to S_{gg'} is a relative global complete intersection. By part (2) we see that R \to S_{gg'} is a relative global complete intersection and gg' \not\in \mathfrak q'. Since \mathfrak q' was arbitrary combining Lemmas 10.136.15 and 10.136.4 we see that R \to S' is syntomic (this also uses that the spectrum of S' is quasi-compact, see Lemma 10.17.8). \square
Comments (3)
Comment #6686 by WhatJiaranEatsTonight on
Comment #6687 by WhatJiaranEatsTonight on
Comment #6893 by Johan on
There are also: