Bourbaki

Remarks

Tag 0EEN, permanent. Cite it as [Bourbaki, Tag 0EEN].

Algebra, chapter V, § 4. Algebraically closed extensions

1) For each i∈Ii \in I let cic_i be a non-zero element of KK and let fi′=cifif'_i = c_i f_i. Then it is clear that every splitting extension for the family (fi)i(f_i)_i, is also a splitting extension for the family (fi′)i∈I(f'_i)_{i \in I} and conversely. In particular, in studying splitting extensions we may limit ourselves to the case of monic polynomials.

2) Suppose that I is finite and put f=∏i∈Ifif = \prod_{i \in I} f_i. Using the uniqueness of the decomposition of a polynomial into irreducible factors in E[X]E[X] (IV, p. 13, Prop. 13), we can easily show that a splitting extension for the polynomial ff is a splitting extension for the family (fi)i(f_i)_i, and conversely. In other words, the case of a finite family may be reduced to the case of a single polynomial.

3) Let f∈K[X]f \in K[X] be a polynomial of degree ≥1\geq 1 and let EE be a splitting extension of ff. If x1,…,xnx_1, \ldots, x_n are the roots off in EE, we thus have E=K(x1,…,xn)E = K(x_1, \ldots, x_n) and [E:K][E : K] is finite (VV, p. 18, Th. 2); but it may happen that EE is distinct from the subfields K(x1),…,K(xn)K(x_1), \ldots, K(x_n) generated by a single root; this can happen even when ff is irreducible1^1. We note however that when ff is irreducible, the fields K(xi)K(x_i) all have the same degree nn over KK, and whenever EE is equal to one of them, we have [E:K]=n[E : K] = n and hence E=K(x1)=⋯=K(xn)E = K(x_1) = \cdots = K(x_n).

Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.

Cited by

Cites

Nothing.

Transcribed from the printed edition, 2003, Springer. Not translated.