Bourbaki

Theorem 1

Tag 06KH, permanent. Cite it as [Bourbaki, Tag 06KH].

Algebra, chapter V, § 7. Separable algebraic extensions

— Let E be an extension of K. Then the following conditions are equivalent: a) E possesses a primitive element; b) there exist only a finite number of subextensions of E. These conditions are satisfied when E is a separable extension of finite degree. Suppose first that EE possesses a primitive element xx, and let ff be the minimal polynomial of xx over K. For each monic polynomial g∈E[X]g \in E[X] dividing ff in E[X]E[X] denote by EgE_g the subextension of EE generated by the coefficients of g. Since the possible polynomials g are finite in number (if ff splits in E[X]E[X] into a product of r monic irreducible polynomials, the number is bounded by 2r2^r), the subextensions EgE_g are finite in number. To prove b) it is therefore enough to show that every subextension L of EE is one of the EgE_g. Now if LL is a subextension of E, then we have L[x]=EL[x] = E; if g is the minimal polynomial of xx over L, we have [E:L]=deg⁡(g)[E : L] = \deg(g). Besides, g is a divisor off in L[X]L[X], hence in E[X]E[X]; so we have Eg⊂LE_g \subset L and E=Eg[x]E = E_g[x]. Since g(x)=0g(x) = 0, we have [E:Eg]≤deg⁡(g)[E : E_g] \leq \deg(g), therefore [E:Eg]≤[E:L][E : E_g] \leq [E : L] and so L=EgL = E_g as we wished to show.

Next we observe that Condition b) implies that the extension E is of finite degree: by Remark 2 of V, p. 18 it suffices to prove that it is algebraic; now if z is an element of EE transcendental over KK then the subextensions K(zn),n∈NK(z^n), n \in \mathbf{N} are pairwise distinct.

To show that b) ⇒\Rightarrow a) we now distinguish two cases: A) If the field KK is finite, the field EE is a vector space of finite dimension over K and hence is a finite set. Therefore 1^1 (V, p. 78, Lemma 1) there exists an element xx of EE generating the multiplicative group of EE, and we have E=K[x]E = K[x].

B) Suppose now that the field KK is infinite. If b) holds, the extension EE is of finite degree, so b) can also be expressed by saying that EE possesses only a finite number of subalgebras. This being so, the implication b)⇒a)b) \Rightarrow a) is a consequence of the following more general proposition (for which the hypothesis that the field KK be infinite is indispensable, cf. V, p. 153, Ex. 5 of § 7):

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

Cited by

Cites

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