Pre-Galois theory (Q2046755)

The inverse Galois problem (IGP\(/k\)) over a field \(k\) asks whether every finite group is (isomorphic to) a Galois group over \(k\). The related regular inverse Galois problem (RIGP\(/k\)) over \(k\) asks whether every finite group \(G\) is a regular Galois group over \(k\), i.e., whether \(G\) is isomorphic to the Galois group \(\mathcal{G}(F/k(T))\) of a Galois (field) extension \(F\) of the rational function field \(k(T)\), which is \(k\)-regular (that is, \(k\) is algebraically closed in \(F\)). The latter question is conjectured to have an affirmative answer, for all fields; and if this holds for a Hilbertian field \(k\), then (IGP\(/k\)) also has an affirmative answer over \(k\). Generally, (RIGP) is widely open; its answer is affirmative, for the class of large (the same as ample) fields introduced in [\textit{F. Pop}, Ann. Math. (2) 144, No. 1, 1--34 (1996; Zbl 0862.12003); Ann. Math. (2) 172, No. 3, 2183--2195 (2010; Zbl 1220.12001)], and [\textit{P. Srinivasan}, Isr. J. Math. 234, No. 2, 769--776 (2019; Zbl 1466.12002)]. A field extension \(E/k\) is said to be potentially Galois (with Galois group \(G\)) if there is a finite field extension \(L/k\) such that \(E \otimes_k L\) is a Galois extension of \(L\) (with \(\mathcal{G}((E_k L)/L)\) isomorphic to \(G\)); when this holds, \(E/k\) is necessarily separable. We say that \(E/k\) is pre-Galois if \(L\) can be chosen to be a Galois extension of \(k\). The class of pre-Galois extensions of \(k(T)\) contains all geometrically Galois extensions of \(k(T)\), that is, every extension that becomes Galois and remains of the same degree over \(\bar k(T)\), where \(\bar k\) is an algebraic closure of \(k\). The paper under review develops a pre-Galois theory. Series of results show how the new Galois properties compare to each other. For example, it is proved that: a finite separable field extension need not be potentially Galois; a potentially Galois extension need not be pre-Galois; there is a pre-Galois extension with nonisomorphic pre-Galois groups (however, the uniqueness of the pre-Galois group of a pre-Galois extension is guaranteed if it is a finite simple group); a pre-Galois \(k\)-regular extension of \(k(T)\) need not be geometrically Galois. The first main result of the paper under review concerns an arbitrary field \(k\) and a finite separable extension \(E/k\) with Galois closure \(\widehat E/k\), together with: the (possibly empty) set \(\mathcal{L}\) of those overfields \(L\) of \(k\) such that \(E/k\) is potentially Galois over \(L\); the subset \(\mathcal{L}_m\) of minimal fields in \(\mathcal{L}\); the set \(\mathcal{G}\) of complements of \(\mathcal{G}(\widehat E/E)\) in \(\mathcal{G}(\widehat E/k)\); a canonically defined mapping \(\mathbb{G}: \mathcal{L} \to \mathcal{G}\). It contains several assertions including the following: (a) if \(L \in \mathcal{L}_m\), then \(EL = \widehat E\) and \(E/k\) is potentially Galois over \(L\) with group \(\mathbb{G}(L) = \mathcal{G}(\widehat E/L)\); (b) \(\mathbb{G}\) induces a bijection \(\mathcal{L}_m \to \mathcal{G}\), whose inverse is also defined. Thus the developed Galois theory includes a Galois correspondence. It investigates the corresponding variants of the inverse Galois theory. Specifically, the authors prove that the following statements are equivalent: every finite group is a Galois group over \(k\); every finite group is a pre-Galois group over \(k\). They also find necessary and sufficient conditions that: (i) \(E/k\) is potentially Galois; (ii) \(E/k\) is pre-Galois. The rest of the paper concentrates on function fields, especially, on geometrically Galois extensions. Classical results about specializing Galois extensions of function fields are extended to the pre-Galois context. One of the main results on this topic gives a partial answer to the open question of whether a given Galois extension \(\mathcal{E}/k\) can be obtained as a specialization of a geometric Galois extension of \(k(T)\). In addition, it is proved that every finite group is a homomorphic image of a geometric Galois group over \(k\). These results are further strengthened in case the field \(k\) is large.