Subfields of ample fields. Rational maps and definability

From MaRDI portal
Publication:961026



Abstract: Pop proved that a smooth curve C over an ample field K that has a K-rational point has |K| many K-rational points. We strengthen this result by showing that there are |K| many K-rational points that do not lie in a given proper subfield, even after applying a rational map. As a consequence we gain insight into the structure of existentially definable subsets of ample fields. In particular, we prove that a perfect ample field has no existentially definable proper infinite subfields.


F. Pop introduced the notion of large field, also known as \textit{ample} field, which subsumes several seemingly unrelated classes of fields: separably closed, pseudo algebraically closed, real closed, and Henselian valued fields. A field \(K\) is ample, if every smooth curve \(\mathcal C\) defined over \(K\) with a \(K\)-rational point has infinitely many of them. This class of fields is now important in contemporary Galois theory, has appeared in several other related topics and turns out to be richer than previously believed (see reference [Pop09] of the paper, which has now appeared as [\textit{F. Pop}, Ann. Math. (2) 172, No. 3, 2183--2195 (2010; Zbl 1220.12001)]. The main result of the paper, which generalizes a result of Pop, asserts that a smooth curve \(\mathcal C\) over an ample field \(K\) with \({\mathcal C} (K)\neq\emptyset \) has \(| K|\) many rational points \textit{that do not lie in a given proper subfield, even after applying a rational map.} The author gives several applications of this result. E.g. a perfect ample field \(K\) has no existentially \(K\)-definable proper infinite subfields (in the ring language), from which it follows that the ring of standard integers \(\mathbb Z\) cannot be diophantine over an ample field of characteristic zero.











This page was built for publication: Subfields of ample fields. Rational maps and definability

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q961026)