Bourbaki

Exercise 23, § 1

Commutative Algebra, IX, p. 47

Let AA be a ring and II the set of negative integers. Let mm be an integer ⩾1\geqslant 1 and [a0,...,am−1][a_0, ..., a_{m-1}] an element of Wm(A)W_m(A). Associate to it the element (bi)i∈I(b_i)_{i \in I} of AIA^I defined by

{bi=ai+m−1for 1−m⩽i⩽0bi=0for i⩽−m. \left\{ \begin{array}{ll} b_i = a_{i+m-1} & \text{for } 1-m \leqslant i \leqslant 0 \\ b_i = 0 & \text{for } i \leqslant -m . \end{array} \right.

Thus Wm(A)W_m(A) is identified with a subset of A1A^1. a) For every integer n⩾0n \geqslant 0, the mapping Vn:Wn(A)→Wn+1(A)V_n : W_n(A) \to W_{n+1}(A) induced by the shift is a group homomorphism. The identifications of the groups Wn(A)W_n(A) with subsets of A1A^1 are compatible with the mappings VnV_n and make it possible to identify the group lim⁡Wn(A)→\lim \overrightarrow{W_n(A)}, the inductive limit of the groups Wn(A)W_n(A) with respect to the VnV_n, with the subset CWu(A)CW^u(A) of A1A^1 formed by the elements all but a finite number of whose components are zero, elements which shall be called unipotent Witt covectors. By transport of structure, one obtains on this subset a group structure. b) For a=(ai)i∈Ia = (a_i)_{i \in I} and b=(bi)i∈Ib = (b_i)_{i \in I} in CWu(A)CW^u(A), one has

a+b=(ci)i∈Iwherec−n=Sm(a−m−n,...,a−n−1,a−n;b−m−n,...,b−n−1,b−n) a + b = (c_i)_{i \in I} \quad \text{where} \quad c_{-n} = S_m(a_{-m-n}, ..., a_{-n-1}, a_{-n}; b_{-m-n}, ..., b_{-n-1}, b_{-n})
for every integer mm sufficiently large. c) For every ring homomorphism φ:A→B\varphi : A \to B, the mapping CWu(φ):CWu(A)→CWu(B)CW^u(\varphi) : CW^u(A) \to CW^u(B) defined by CWu(φ)(ai)i∈I=(φ(ai))i∈ICW^u(\varphi)(a_i)_{i \in I} = (\varphi(a_i))_{i \in I} is a group homomorphism.

Solution

No solution has been written yet.

Machine translation of the English, by gpt-5.4. Not checked by a person.