On division algebras having the same maximal subfields. (Q981666)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 5729741
Language Label Description Also known as
default for all languages
No label defined
    English
    On division algebras having the same maximal subfields.
    scientific article; zbMATH DE number 5729741

      Statements

      On division algebras having the same maximal subfields. (English)
      0 references
      0 references
      0 references
      2 July 2010
      0 references
      Amitsur proved that the family of splitting fields of a division algebra determines (and is determined by) the subgroup generated by the algebra in the Brauer group. The subgroup generated by a division algebra in the Brauer group determines the maximal subfields, so it is natural to consider the following property of a field \(K\), denoted here (!): the maximal subfields of a central division algebra over \(K\) determine the subgroup generated by the algebra in the Brauer group. Let (*) denote the restriction of this property to \textit{quaternion} division algebras. All fields below are of characteristic not \(2\). The property (*) is related to arithmetic subgroups in absolutely almost simple algebraic groups of type \(A_1\). For example, the fact that number fields have (*) implies that every Zariski-dense subgroup which is weakly commensurable to an arithmetic Zariski-dense subgroup, is commensurable to it. In fact, (*) holds for any algebraic extension of \(\mathbb{Q}\), but infinitely generated fields for which it does not hold are known. The authors prove that (*) is a stable property: if the property holds for a field \(K\), then it also holds for \(K(x)\). The proof builds on Faddeev's exact sequence via a careful analysis of ramification, and implies a recent theorem that \(k(x_1,\dots,x_n)\) has (*) for every number field \(k\). The authors also prove that if \(k\) is finite (of odd characteristic) or totally imaginary with a single diadic place, then \(k(x_1,\dots,x_n)\) satisfies the restriction of (!) to algebras of exponent \(2\) (which clearly implies (*)). Somewhat relaxing the condition, they show that if \(K\) is algebraic over \(\mathbb{Q}\) with a finitely generated absolute Galois group and \(f\) is an absolutely irreducible polynomial over \(K\) such that at least one of \(\deg f\), \(\deg_xf\) or \(\deg_yf\) is odd, then the field \(\widehat K=K(x,y|f)\) of \(K\)-rational functions on the affine curve \(f(x,y)=0\) satisfies (*'): the number of isomorphism classes of quaternion division algebras over \(\widehat K\) with the same maximal subfields is bounded.
      0 references
      0 references
      division algebras
      0 references
      quaternion algebras
      0 references
      maximal subfields
      0 references
      weak isomorphisms
      0 references
      fields of rational functions
      0 references
      Brauer groups
      0 references
      central simple algebras
      0 references
      splitting fields
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references
      0 references